May 30, 2025 • 56 mins
Join Amy Kemp of Glossier as she takes Ger and Gordon out on the CX waves, where service agents help navigate customers through anything from gentle breezes to force 10 gales.
You'll find out whether AI is more than just a wizmo gizmo and hear how Glossier is using it to genuinely enhance the customer experience. But fear not, the tech isn't driving the boat. A rich and meaningful customer service experience still requires a human hand on the tiller, if you're looking to use CX to drive customer satisfaction and loyalty.
Beware, though. There are hidden rocks beneath the CX. If you don't want your service agents to founder, or be mauled by the salty sea dogs in the shipping department, they'll need a deep understanding of how your business works. Amy explains how experience of the entire business helps CX agents provide customer solutions without generating internal problems.
Learn about CX as a differentiator, CX as a post-purchase experience, a pre-delivery experience and post-delivery experience. Find out what to do when CX is out of your control, as well as how Amy sent away sheets to be dyed another colour and received a dress back.
And look out! There are pirates ahead! Twice!
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:01):
This is Functional & Fabulous, the omnichannel podcast where we unbox tales of online retail and digital transformation.
In this episode, celebrities lose their temper...
There was an incident where a celeb accused me of calling the paparazzi on them.
You'll hear horrifying stories that will curdle your blood...
(00:25):
That makes it sound very dramatic.
Ger has another bloody poem...
I like to spout poetry from time to time.
It takes care of the mundane to give you time for the meaningful.
Gordon explains what it's like to listen to Ger's poems...
Oh, you hurt me.
And oh, you hurt me again.
And he suggests a punishment even worse than listening to Ger's poems...
(00:50):
I actually think that everybody should be forced to work either in a contact centre.
100 per cent.
Or as a CX agent.
Ger gets his revenge...
He's laughing at me now.
I can't do it.
You're going to have to do it now.
You're going to have to do it,
I can't do it.
And there's another sodding pirate joke...
(01:13):
Hardy-har-har-harr. So what's your second pirate joke?
Two.
Two sodding pirate jokes.
This episode of Functional & Fabulous is brought to you with pride by StudioForty9, retail ecommerce experts, omnichannel growth consultants and cut-through performance marketing specialists.
StudioForty9, where your digital retail success is built.
(01:37):
Hello and you're very welcome to another episode of Functional & Fabulous.
Today we're joined by Amy Kemp, Director of Omnichannel Customer Experience at Glossier, a powerhouse beauty brand based in New York City.
Amy's career has been a little funky, to use her own words.
She's gone from NYC restaurants to med school startups, to leading CX at Glossier, and has tackled everything from panicked students to supply chain chaos on the journey.
(02:03):
Now she's making AI work for customers, using gizmos to handle Wizmos, and discovering that pick-pack satisfaction is very real.
We're talking good AI, bad AI, and what the hell AI, retail and digital, and why CX is the best job nobody notices until something goes wrong.
Amy Kemp, welcome to Functional & Fabulous.
Thank you for having me.
(02:25):
So Amy, what would you say is the important aspect of working in New York City restaurant scene as an ideal place to start to learn about handling customer experience issues?
You know, somebody who is hungry is really a particular customer.
And I think if you can learn the ins and outs of, you know, somebody who's maybe, like, on the edge of a nervous breakdown because they haven't eaten, making them feel settled and excited, you can probably tackle anything.
(02:52):
I feel like it's way more dramatic as well.
Like, the drama must be real in a New York City...
You know, my hunger is so immense.
I'm going to have a nervous breakdown if you don't bring me my spaghetti immediately.
Immediately.
I will also share, again, setting the scene.
I'm, you know, I'm 23.
Indie sleaze is still popular.
(03:12):
I was living in a basement, you know, really trying to figure out my life.
Working in the, you know, West Village in New York, which is a pretty chi-chi part of town, if you will.
Lots of celebs were coming in, you know, and I will share,
there was an incident where a celeb accused me of calling the paparazzi on them while they were eating at the restaurant.
(03:33):
And there I am, you know, fighting back tears, being, like, I don't even know how to do that.
Honest to God.
I think Robert De Niro is really reaching, you know?
I mean, seriously, what does he think?
Is he that important?
If you get to the other side of that, you can kind of take on anything.
So I can well imagine.
I mean, like, there's a number of different crucibles that we talked about.
(03:54):
You have the New York restaurant scene, panicking medical students.
The wire cut pick that effectively turned a small mattress company you worked for into a viral sensation.
So you've had a good few kind of interesting peaks of having to handle irritated people, I would say, over your career path.
(04:15):
I would agree with that.
And also, you know, managing my own feelings in this crazy world.
I'm probably the toughest customer out there.
I don't think that gets talked about enough.
That's true.
Yeah, 100 per cent.
It's totally real.
So tell us a little bit about Glossier.
How did it get started as a business?
(04:36):
And where is it today?
So Glossier was founded by Emily Weiss about 11 years ago at this point.
She was working predominantly in fashion.
She was going to the backstage of runway shows, working with models, working with Vogue, and started kind of getting very curious about all of these people who work in beauty, create all these beautiful looks.
(04:59):
What's in their kind of top shelf?
What's in their medicine cabinet at home that allows them to make people look so fabulous all the time?
And she started a blog called Into the Gloss that was primarily focused on kind of exploring what's in people's medicines cabinets that have these really interesting beauty backgrounds.
And after doing this for a number of years, you know, she understood, like, hey, there's kind of a gap in the market of some really amazing products, right?
(05:25):
Like, all of these people on the runway are using foundation, but they're first sheering it out with moisturiser.
Like, we have the opportunity to really develop something, like, in that space that fills that hole.
And she launched Glossier, which is, you know, the French dossier, but with gloss, Glossier.
And the brand really kind of took off and here we are still going strong.
(05:48):
I mean, it's an absolutely gorgeous website.
It looks like an absolutely beautiful product.
What's the clientele like?
I think there isn't any, like, one particular Glossier customer.
We, a couple, when I joined, we relaunched our Soho flagship store.
And sometimes when we open new stores, people will wait in, like, line for an insane number of hours to get in for the first time.
(06:14):
And we were walking down the line, kind of talking to people and hearing their stories.
And there was a woman who I think was probably in her seventies who told us, like, I am going to wear Glossier products until I'm in ashes.
Which I thought was so funny.
But because, like, our motto really is skincare first, make-up second, I think we offer a large catalogue of products where, like, anyone can really find their journey.
(06:40):
And within that journey for the customer, you're the director of Omnichannel Experience.
So what is your role in helping with that customer's journey?
Yeah, I think, so I have, I would think that I have kind of two big functions within my job.
One is, like, basically everything customer service kind of flows into me.
(07:02):
So really understanding, like, okay, you know, we're always going to have those questions, like, where is my package?
What's going on, et cetera, et cetera?
Like, how can we make sure we really have the right team in place or responding to people?
We have the right process and policy to support our customer.
So kind of that whole customer service piece.
And then the other function of my job is really understanding kind of our customer more globally and, like, what we need to do to continue iterating on our experiences holistically to keep our customer engaged, to keep that retention, that loyalty, et cetera.
(07:34):
And to me, that's really kind of one of my favourite parts of the job is, like, spending all this time in customer data, getting these clear insights and then basically going to other parts of the company and, like, trying to influence them to be, like, you know what would be really cool? Is if we did this instead of that.
And this is why.
(07:54):
So yeah, I would say, like, I'm an influencer.
I love that.
So it's kind of split, isn't it?
It's make customers happy and then make changes to keep them happy and keep them.
Yeah.
And, you know, I think in our, like, modern ecommerce world, pretty much every brand has competition, right?
And I think more and more brands are really competing on this idea of, like, loyalty and retention or, like, how are you distinct and how do you have a relationship with your customer where they're going to continue to, like, pick you versus somebody else?
(08:27):
Recording in progress.
"Recording in progress" - versus kind of somebody else in the space.
And I think so much of customer experience is about that, is about, like, really listening to your customer, like, hearing where we can still deliver and, like, making those changes to really build those meaningful relationships.
Yeah.
Do you think, like, the role of technology then plays a big part in this?
(08:50):
Because on one hand, you've got the commerce experience now, which is fairly leveled across most retailers.
And then you have the experience component where, okay, this is the battleground where we're going to compete.
That makes it sound very dramatic.
But if you're entering in that battleground, is there a role that technology needs to play and what role do you think it plays?
(09:14):
First of all, I went to performing arts high school, so I love the drama.
You can keep it coming through the entire conversation.
But absolutely.
I was thinking about kind of this morning, you know, as, like, a customer experience team, like, most teams in ecommerce, right?
We have, like, a couple, like, governing KPIs that we're, you know, responsible for delivering to the brand.
(09:37):
And one of those big KPIs is FRT first response time, right?
Like, how quickly do we get back to that customer when they reach out?
Like, I'm sure we've all had that experience of, you know, somebody, I don't know, charged us incorrectly.
And, like, we need to talk immediately.
Like, I need to get refunded.
And so I think first response time is really a critical part of that customer journey.
(09:58):
And I think holistically, we've kind of come at that problem really from, like, a staffing POV of looking at, like, okay, we know that we get more inbound tickets -
you know, we call people who reach out to us tickets - on Mondays.
So, like, we're going to staff for, like, this on Mondays and then this on Tuesday, and we'll put this team here and this team here.
(10:18):
And I think where the industry is going, which makes a lot of sense to me, is instead of looking at this from, like, a big holistic staffing POV is really looking at kind of the different reasons why people reach out to your brand.
And then kind of making the right processes and protocols based on contact reason.
And so, for example, you know, to give, like, an obvious one that I always give, you know, what we call, like, order changes.
(10:43):
So cancel requests and, like, address changes.
If somebody realises, okay, I've already said I'm moving, right?
It is, I bet you three to five times in this process, I'm going to deliver something and send it to my old apartment.
And I'm going to realise the second I look at that confirmation email, and I'm going to panic and cold sweat and email that brand immediately.
(11:07):
And, you know, at Glossier, we have a one-hour cancellation window.
If I can connect with someone in that window, great.
They can, you know, update my address.
It's fine.
If I can't, it's going to be a really bad experience.
So that to me feels like a great use case for AI, right?
Like, AI responds almost immediately.
So it's going to guarantee that we can connect with that person, perform that action, cancel that order, whatever it is, that again, is going to, like, build that customer loyalty.
(11:36):
And then I think I say this, like, kind of nested in the overall conversation of first response time, because we can still hit that first response time kind of on average by prioritising the tickets where speed really, really matters.
And then where it's less important, you know, we can follow different processes while still, like, holistically really ensuring that we're hitting those KPIs.
(12:01):
That's really interesting that you mentioned AI, because I think we're probably going to talk about that quite a bit.
One of the things that I think is probably important for us to flag earlier on is the difference between ticket deflection and actually providing meaningful resolution for customers using AI.
(12:24):
Because the temptation is, I'm going to deflect as many tickets as I can from agents.
I'm going to reduce the amount of headcount and the amount of costs that I've got.
And we're going to tick this big box to say, yes, we've done this and we're using AI.
But unless the resolution is meaningful, then you're just causing the brand a longer term problem, right?
(12:46):
I could not agree more.
I don't think I'm being hyperbolic when I say probably once or twice a day, I get an email from some type of AI sales company saying, I can promise you 100 per cent deflection.
And for my POV, I'm, like, I'm sure I could deliver 100 per cent deflection to myself and then I wouldn't have a business anymore, right?
(13:06):
Because why would someone bother reaching out to us when they're never going to connect with a human agent?
So I think what we have to really look at things, I try to make it all decisions kind of on a triangle.
So really thinking about the customer experience, but also the business experience, like, what's the right move for the experience?
So the business, excuse me, and then also kind of through an employee experience.
(13:30):
And when we can use that triangle to really look at kind of decision-making, generally we're doing something right.
Brilliant.
So when you think about an AI, I'm just going to obsess over the cancellation example for one second.
The AI has to not only be able to respond to the customer, but it has to be able to complete the action of cancellation, or the action of an address change.
(13:55):
And I think for people who haven't experienced this on the business side, this will appear quite revolutionary.
I'm assuming that's what happens with you guys.
The AI just, it can handle the process as well as the conversation.
Yeah, definitely.
And I think AI, depending on how well you train your AI, to use a kind of goofy term, it can also escalate things, right?
(14:20):
It can be that first line of defence.
And if something truly goofy is going on, it can be, like, this is outside of my realm of expertise.
Let's make sure we're connecting you with the right person.
Yeah, and again, that is fundamentally important.
But it deals with the customer because you've got, okay, here's the urgent thing that's very simple.
And that's now resolved.
Customer very, very happy.
(14:42):
Customer's got a bit more complexity, needs to talk to a real person.
And that person then needs to be empowered to, okay, we can fix this.
From anyone thinking about this, because I'm sure there'll be retailers listening, going, oh, this sounds, this is fantastic.
How do you even start on that journey of implementation?
Yeah, so for us, I think Glossier was a pretty early adopter of AI.
(15:06):
It predates me joining the brand.
So sadly, I can't report, like, what year they brought that on as a resource.
But in 2023, for kind of a series of reasons, we took a step back and looked at our AI strategy holistically.
I mean, obviously there've been a lot of advancements in AI over recent years.
And we really started by kind of obsessing over the reasons, like, the contact reasons, the reasons why customers, like, reach out to us in the first place.
(15:33):
And I think that, and not only why people reach out to us, like, when do we see X versus Y, right?
Like, we know if we have a promotion, we have an exciting product launch.
You know, probably, like, seven to 12 days later, that's when we're going to get a huge influx of, like, where's my package tickets, right?
And so really understanding kind of what our customer's journey is, and then working backwards again, looking at the use cases in which having again, that incredibly fast first response time is most important.
(16:07):
Got it.
So you have that.
And then how do you then deal with the more complex things?
Is that a case of, okay, we're going to route, and then we're going to train our, we're going to route the query over towards our agents, and we're going to train our agents, or do you then double down on your, how can we train the AI to do this better?
I think obviously with AI, like, you're constantly giving it feedback.
(16:31):
So in some ways, like, when we do see maybe something that wasn't perfectly handled, or we see something being escalated that we think that the AI can handle, like, we're going to give feedback and we're going to iterate and improve, but predominantly, like, pass it to a human agent, right?
I think for me, if the AI is no longer providing the value that we think the AI should be providing, which again is that incredibly fast response time or the ability to do that small action that has a downstream impact, why then try to fit, what's that expression - like, a square peg into a round hole?
(17:03):
Like, that's when I think we should definitely pass it to a human agent.
And when you're looking at the different types of tickets that you've got coming in, I'm assuming that 80 or 90 per cent of the tickets can be classified in very few ways.
Is that right?
Do you categorise that way?
Definitely.
I think we definitely see patterns and trends.
(17:25):
And then also there are only so many reasons, like, why someone's going to reach out to an ecommerce brand.
You know, it's so...
So Wizmo tickets, where's my package?
Or we sometimes call them package status at Glossier, so that's what I mean.
Obviously, that's a huge reason why people reach out to us.
And I don't think, I think sometimes the ways that, like, AI technology companies or technology companies in general, like, talk about package status tickets is, like, a little bit maybe unkind to customers.
(17:58):
You know, there's this idea, like, oh, you know, most brands use Shopify.
We use Shopify.
You know, Shopify has statuses that don't, you know, denote, like, oh, this order has been shipped.
This hasn't been, et cetera.
And so the AI promises, like, oh, well, we can go into Shopify.
We can, like, understand the order status and, like, we can pass it back to the customer.
Done.
(18:19):
Pretty much everyone buys everything online, right?
Like, everybody on a computer, like, probably has some working knowledge of how to understand what's going on on a tracking link.
I have a friend who works for the Home Shopping Network.
I don't know if you have, like, the equivalent?
(18:41):
Mhmm.
Okay, great. So could not be further away from ecommerce, you know, in terms of, like, her day-to-day job.
But recently made a joke.
She was, like, oh, you know, you're in trouble when you land on that, like, Narvar tracking page.
And I was, like, how do you know what Narvar is?
Right?
(19:01):
Like, I think most people have, like, a pretty good fundamental idea of, like, how this part of ecommerce works.
So if a customer is taking the time to reach out to you and say, where's my package?
What's going on?
It's probably not as simple as, like, it shipped.
And so for us finding kind of an AI partner who could help us with some of these queries, which again, we think if someone's scared about where their package is, a fast response really matters here.
(19:29):
It was having an AI partner who could actually open a tracking link and parse that information, like, a human and understand, okay, yep, this got lost somewhere in Norway.
And therefore, like, I understand that with in mind, like, let's take the next action step versus, like, well, it shipped. AI out.
(19:50):
Yeah, we all know how to check the status of a package.
But how do you have that contextual awareness of, this package is stuck?
Because, hey, that status hasn't moved for four days and it should have done.
What do you do next?
And that's pretty impressive because the standard, oh, yeah, we've got an AI response is,
(20:13):
here's your status.
Here's your status, order processing.
It sounds, like, as well, it sounds, like, it's important that you ring fence what the AI does and knows about, we'll say, and what it doesn't.
And you have to be pretty clear about that, you know.
So, you know, do you typically train the AIs with a lot of general knowledge or is it really, okay, here's what you know about and then after that, you just hand it over?
(20:39):
Yeah, so when we were looking at AI partners and we met with a lot of different people in the field, we wanted to go with a partner that makes, like, a specific AI bot for each contact reason.
And the reason why that was important to us is because we could basically give each bot, like, the amount of information it needs to do one very specific function and nothing else, right?
(21:04):
Reducing that chance for the AI to potentially hallucinate and, like, give the customer information that they don't really need or is not helpful.
I think, so that was important to us from, like, a decision lens.
Yeah, that's quite clever because I mean, there must be a huge, well, there must be a large temptation to just say, okay, we'll dump the entire knowledge base on the AI and then figure it all out, but no.
(21:25):
I have worked at Glossier for over two years and I would say, like, once a quarter, we're, like, this is the quarter we're going to go through our knowledge,
make sure everything is up to date, and we have never done that task.
And we will eventually, because I'm an optimist, but it just, to me, there was a lot of inherent danger, I think, in just giving the AI a ton of information that I can't even confidently say is up to date.
(21:52):
Yeah, keeping your knowledge base up to date... and these things just grow and grow and grow and you change process because, like, one of the fundamental parts of your job is changing the process.
So then is every single article in the knowledge base that references that process getting updated and how do you manage that?
If anyone can tell me, I would love to know.
(22:16):
Excellent, so we'll put that call out if anybody knows how to do that without having a small army of content editors, then we would all like to know, I think.
Answers on a postcard.
What coverage would you say would you, of your tickets, are, like, entirely handled, we'll say, to the satisfaction of customers by AI?
(22:36):
Would you know?
A lady never tells her secret.
A lady never tells.
A lady never tells - 82 per cent.
Okay.
100 per cent.
100 per cent ticket deflection.
Do you know, actually, on a slightly separate
but also a wondering note,
because I've often wondered, in Ireland
it's unusual for ecommerce merchants
(22:57):
to have support,
so chat support after, let's say, 5pm, 5.30pm,
that kind of a way,
but most people are actually buying in the evenings,
and I've always thought it would be
a really useful experiment to test
to see how having that additional support
in the evenings,
when people are on the stores and buying,
(23:17):
how useful that is?
From a customer experience point of view, have you experimented in that?
Would your staff be on after hours or in antisocial hours and things like that?
So I will say I think we're lucky
in that we span,
our team spans three time zones,
so we have some members in the UK,
(23:38):
some EST and then PST,
so we do have wide coverage on that front,
but I don't think we have full force
kind of explored it in the way that you're imagining,
but it's definitely something that we think about
in terms of you can't do it all,
and you can only really do
(23:58):
one or two things well at a time,
and so I think eventually we'll get there
on the priority matrix.
So, okay, so you've done quite a few things.
It sounds like you've done a lot more than a lot of businesses have in the area of trying to approach customer satisfaction via AI.
What else, what's next for Glossier?
(24:19):
What other things are you experimenting with?
That's a good question.
I think, right now,
we have really spent a lot of time
delving into kind of customer data
from a number of different lenses,
you know, NPS, CSAT,
like, ticket data, et cetera, et cetera, et cetera.
(24:40):
And have spent a lot of time
kind of pulling out themes
and also kind of mapping,
you know, obviously people give us feedback
and I, like, to kind of run that against
data regarding if and when
people return to shop with us again.
Because talking about prioritisation again,
there are 900 things
(25:00):
that I want to do in any given day,
but in order to make the right decisions,
what is the thing that is really going to
hurt our relationship with the customer?
And not to bring this full circle,
but I think one of the things
that the data shows very clearly,
we don't get a lot of,
like, in terms of NPS, let's say,
(25:21):
we don't get a lot of detractors
who say, like, customer experience
is the reason that I'm not coming,
or is the reason I'm unhappy
with this experience,
but they have one of the highest instances
of being kind of, like, spurned
and then not wanting to come back
and kind of shop with us again.
And I think at first, when I saw that, I was kind of scratching my head, being, like, have we done something that's really that horrible?
(25:45):
But it didn't take me long to really think, well, you're only interacting with customer service, right?
It's, like, something else has already gone wrong.
And then to have that experience and then to also not have the experience you want kind of with your CX, you know, with the customer service piece, like, yeah, you're probably not going to shop with this brand again.
(26:07):
And so I think a big focus is really on, like, CX as, like, retention.
That's an interesting one.
So, yeah, so basically you're looking at what the experience is and rather than, as you say, rather than measuring just NPS and CSAT, you're measuring what interactions they had and whether they came back and bought again or not.
(26:28):
Makes an awful lot of sense.
I think that goes beyond that and the point of, like, it being double bad.
It's, like, you're not, nobody ever calls you to say, hey, just calling to say everything's fantastic.
Just calling to say it's brilliant.
Love it.
They're calling, well, not calling or they're interacting with you because something awful is, well, not awful, but it's probably a bit dramatic.
(26:57):
Their package might be late and they're, like, why is my package stuck?
Or, like, why have you double-billed my credit card?
That could be pretty awful for somebody.
Not that double-billing of credit cards happens that often, but then to not be able to either connect with somebody and not be able to, or not be able to get a resolution to your satisfaction, then why would you go back to that brand?
(27:21):
Because you're, like, oh, you hurt me. And, oh, you hurt me again.
Something went wrong and also something could not be fixed.
Yeah, exactly.
So that's a super-valid point of, like, what is the role of CX?
It's not a cost centre.
We are fixing the problems that are caused by other things within our business.
(27:45):
And I don't even know it's necessarily problems with the business.
It's just, there's only so much we can really control.
I think so much of, like, when you interact with a customer who clearly is feeling quite frazzled, it's because something scary that they can't control is happening.
(28:08):
And so, yeah, it's really, I think, our value to make people feel seen and heard.
That's also a really valid point.
It reminds me, I was once asked a question, I'll pose the same question to you.
And it comes back to carriers.
I seem to be obsessed with carriers at the moment.
But if your carrier loses a package or if your carrier is not hitting their delivery promise, whose responsibility is that?
(28:36):
Is that the carrier's responsibility or is it the brand's responsibility?
Because very often it's outside of the brand's control.
Well, I might just adjust that a tiny bit.
I know that you're, I know that's a totally valid question, but I think in the mind of the customer, who are they holding responsible?
And yeah, who the customer holds responsible, I think is, I think that's obvious.
(29:01):
I think that's quite an obvious answer to that question.
So does the other answer matter then?
But what is the reality?
Tell us, Amy Kemp.
No, I totally agree.
Like, you know, I think no one's going to kind of look at the experience cohesively and say, like, well, at this point it really became the carrier's responsibility.
(29:23):
It's, like, no, it's their experience shopping with your brand, right?
And that really ends even after the package is delivered, right?
And they open it up and the order is as expected.
And even sometimes it goes further, right?
In which, like, I don't know, an item is missing and then they have that experience with, like, CX, et cetera, et cetera.
So yeah, you have to, like, look at it as a holistic experience.
(29:46):
I think you need to look at it holistically.
It can be very challenging in terms of level of influence.
And one of the things I started to see, which is quite interesting, again, on the carrier front and feedback in various tickets is why do you use X carrier?
It's a question that then gets asked.
(30:11):
So assuming that customers are not carrier-aware... I'm not quite sure where I'm going with this, by the way, just carrier bashing, probably.
Yeah, we definitely see that a great deal.
And there's no pattern, right?
It's, like, we send something in the US via USPS and it's, like, oh, I hate USPS.
And then we send something to someone else via DHL and they're, like, oh, I hate DHL.
(30:34):
Like, you know, like, it's just people have different experiences, different carriers and we can't really control for that.
It kind of comes down to if you've had one bad experience with one carrier that stays with you.
And the point that you've just raised of if you have a bad experience with the brand, you ain't coming back or you'd certainly go back under duress.
(30:56):
And tell us, Amy, do you get on the pick line much?
Do you get to the coalface?
Are you answering the tickets occasionally?
Or are you on a pedestal?
I pontificate over tickets.
Or do you actually participate in the tickets?
I try to do about an hour of tickets per week.
(31:19):
Sometimes it's more, sometimes it's less.
And to be honest with you, I'm actually, like, not a very good CX agent.
There are people who have a lot more talent at it than I do.
I will share somewhat recently, we had a launch that sold out very quickly, much more quickly than we anticipated.
(31:40):
And one of the tickets I worked at was a customer kind of expressing their frustration of, like, I really wanted to buy this.
And I couldn't because it was sold out.
And I think I probably had been having a long day.
And I was, like, I'm going to really take a moment and connect with this customer and really give them a peek behind the curtain.
No one's more disappointed than we are that we sell out too quickly.
(32:05):
And I'm so sorry about this.
I think I even sent an email, maybe, like, if I had a time machine... I really was on this journey.
And then they responded back to me very quickly with something along the lines of, get your stuff together.
And it made me realise, I am most effective when I am really looking at this kind of puzzle holistically and understanding what are the levers that we need to change as a business to again, really drive that loyalty and retention.
(32:38):
And then again, how can I influence those priorities across the business, in a way that makes sense with, again, like, our customer experience, our business needs, and also our employee experience.
I am never going to make every single person happy.
Mm-hmm.
Even though you're empathising and you're offering the sneak peek and all the rest of it.
(32:59):
In reality, they just didn't get what they wanted.
Everybody else got it.
They have enormous FOMO and they don't care.
And you need to get your shit together, is what I'd imagine was actually in that email, right?
It's interesting you touch on employee experience there.
I always think it's very much kind of a left behind thought or concern in businesses.
(33:21):
But, you know, amongst all other people who can most affect what's going on in your business, the employee satisfaction and keeping employees happy and the fact that employees are enjoying, let's say, to the most that they can, their work environment and so on is important and is going to have a huge benefit to your business.
(33:42):
How does Glossier think about that?
What are you guys doing in that regard?
Yeah, I mean, I kind of, to go back to AI for a second, I think one thing that's been very exciting about AI is, you know, in some ways, it helps with those kind of, like, very quick, kind of, like, immediate tickets that are probably not, you know, not the most, maybe, like, memorable customer interaction someone's going to have in a day.
(34:09):
And it frees up some of that, like, time and capacity for those moments in which you really need to spend time connecting with someone, right?
Or, like, somebody had, you know, a bad experience.
We need to do something kind of unique again to, like, rebuild that trust.
That's when the work is meaningful.
I think we say this all the time, but I'll just say it again here.
(34:29):
Like, being a CX agent is an incredibly hard job.
It's really hard.
People are, you know, the anger is coming towards you and we're not really in a position where we can say, well, it's not my fault.
Like, it's, you should talk to logistics, you know, because again, for the customer, right, it's, like, one cohesive brand experience.
(34:50):
And so I think really taking those moments to carve out times for, like, meaningful interaction are critical.
Both ways.
It's interesting.
So to be poetic about it, I like to spout poetry from time to time.
It takes care of the mundane to give you time for the meaningful.
Yeah.
Oh, I feel that should be, like, on a wall somewhere.
(35:12):
A t-shirt.
Do you know, I have to share a little, you know, CX experience.
My partner had, just before Christmas, had bought a coat for her father, let's say two weeks before Christmas.
It was the present for the family, for their father.
And the coat never turned up and it was, let's say, Monday and Christmas Day was, I don't know, was the day that they were presenting it.
(35:36):
It was Wednesday, Christmas Eve and it still hadn't arrived and she phoned up the brand and I obviously, working in ecommerce and actually know the brand, said, you know, kind of trained her a little bit and said, look, they may or may not be able to do a lot for you.
It's almost certainly stuck in a depot somewhere.
You know, take it easy on them.
(35:56):
Don't get, you know, don't start hammering into them and maybe they'll help you out.
So she phoned them up anyway, and the company is from the north of the country, and they're very lucky in that they have a lovely accent in English, when they speak English, but also they use the word wee a lot as in, you know, small or little.
(36:17):
And Maureen came off the phone anyway and I said, how did you get on?
And she said, oh, the girl said she'll have a wee look.
And it was just, like, with that kind of, I don't know, very, I don't know how to put it but it was such a charming way of saying I'm going to look into it for you - and got it sorted actually and basically an hour before we had to leave for her parents, the coat arrived in the door.
(36:42):
So it was a great story all round, but I thought I must use this term wee if anybody ever wants to complain about something, I'll have a wee look at it for them.
I love it.
I mean, you know, earlier, I guess in Q4 of 2024, I, like, similarly had an experience where I've gotten, we kind of covered earlier, I'm incredibly cheap.
(37:07):
I would say that's, like, my dominant personality trait.
Cheap? I thought you were the queen of sustainability.
Yeah, sure, sure, a lot of that.
Was that not it?
No, I misunderstood.
But I, we have, like, a lot of sheets.
Again, I used to work at a bedding company.
So there's a lot of sheets in this house that are looking, you know, pretty old and dingy.
And so I've recently gotten into this idea of dye baths, D-Y-E baths, where, like, you can send, you know, clothing, bedding there and then they can dye it a different colour and make it look kind of exciting and new to you again.
(37:39):
And I sent something to, some sheets to this dye bath, and they sent me back a dress.
And I was, like, this is not what I sent you, and not my dress.
And despite, you know, working in CX, I really got myself into a lather.
I was, like, oh, like, I want this...
like, now I have this person's dress in my house, and I called them and it was, like, the second I spoke to, like, a reasonable, cool human on the other side of the phone, I was, like, oh, it's fine.
(38:11):
Like, this will get worked out.
Because, is it a nice dress?
No, it's a horrible dress.
That's why it wasn't...
It was a very bright pink, which is really, I live in New York, you know.
But yeah, I think, like, we sometimes, like, forget about kind of the humanity in all of this.
(38:32):
And, like, you just need a little reminder.
Yeah.
Well, I think what's interesting about both of those stories is as soon as we connected with somebody that was able to diffuse the situation, either by being cool or through their use of wee,
and then, like, it would have been fine if the coat hadn't been delivered.
(38:54):
And immediately you feel better about the problem that's happened with this dress.
And I don't think we're in a position where AI can actually do that right now.
But importantly, it gives time for people, like, for the CX agents to be able to also not to be so knackered by answering, you know, where's my order all the time that they get the opportunity to actually deal with it humanely.
(39:21):
I actually think that everybody should be forced to work either in a contact centre...
100 per cent.
Or as a CX agent, or in retail or in a restaurant at some point.
Dealing with people.
Like, they should learn how to do that.
And then that will inject some sanity into the strange process exceptions that they're making.
(39:44):
Yeah, and that's a cool thing about Glossier.
When I joined, I think my first week, they were, like, okay, and on Friday, you're going to go work at the store for a day.
And, you know, are very good about being, like, great, like, you need to go to the store and, like, see how this works or have this experience, because it's - I think it's about, I don't think you really, like, understand, like, the full ecosystem of how the different parts of the business work together or, like, where there's overlap and where there's differences until you're, like, really in those spaces.
(40:12):
And then you understand when someone's, like, oh, can we just make an exception?
Just call the warehouse.
And that's the other place somebody should go and actually work.
Everybody should actually be on the receiving end of one of those calls to say,, like, can you just cancel this order that's already shipped, and deal with an exception, to understand what is actually happening.
(40:35):
When that's triggered.
You can't.
I think, you know, if it's going to the wrong address, send another package.
Like, it's really cruel.
Cruel to the team member in the warehouse.
Who's on their hands and knees, like, digging through, like, a pallet of, you know...
(40:58):
On the 20th of December...
Hasn't seen their family in a week.
Yeah.
That's interesting.
So you guys get to work in various departments and a little bit of cross-pollination of knowledge and ideas within the business, right?
Yeah, and that's, like, another, I think, special thing about Glossier is, you know, my title, right, is, like, omnichannel customer experience.
(41:25):
I have sat, you know, in the retail function of the business where it was, like, everything customer-facing is going to be in retail.
I now sit in, like, kind of more of the ecommerce function of the business.
So, like, that is a lot of kind of my day-over-day touchstones.
But my responsibility is really kind of, like, that cohesive omnichannel brand experience.
And I, you know, I hope that other brands, like, start to think about things in those ways too.
(41:48):
It's very interesting.
So Glossier in general, you know, where do you see the business going or, you know, what kinds of things are you interested in in the future just as an overall business?
I presume there's only so much, so many categories of make-up.
Maybe there isn't.
You can never have enough moisturiser.
(42:08):
Never?
Never.
Ever.
OK, so I'm wrong about that.
More moisturiser.
So after you've made more moisturiser, what's next for the business?
There are so many exciting things coming down the pipeline and I'm at my heart a real follower and I would hate to accidentally share something that I'm not supposed to share.
Oh, go on.
Go on.
More moisturiser, it's right.
(42:29):
Share a wee secret.
Share a wee secret with us.
But I, you know, I think we are having, you know, for so long the customer was really kind of focused on that ecommerce experience and, you know, and obviously then, like, ecommerce plus some retail and now we're also in wholesale.
So I think, you know, we're really kind of enjoying and figuring out the evolution of, like, what does it mean to be a brand that exists in all these different places and all these different ways and it's a very exciting problem to kind of, not even a problem, opportunity to come to.
(43:02):
And then I would say other than that, like, follow us on Instagram.
Smash the like button, subscribe.
I think if you can solve that problem of taking a very controllable customer experience of one channel of, okay, it might not be perfect, it might not be easy control, easy to control, but then you add in retail and then you add in wholesale and you still want to maintain that brand experience around CX.
(43:32):
That's an enormous challenge.
Like, that's going to keep you busy for the next, I don't know how many years.
Yeah, I will say this is an interesting, Glossier really does in many ways feel like a community.
I think the brand kind of emerged during such a, like, exciting time of during so many people's lives.
(43:54):
Like, there really is kind of, like, this deep affinity and I do feel like we're really still connected and tapped into that community.
You know, you mentioned earlier, like, no one ever, like, emails to say that they're having a good time.
I feel like I'm an incredibly privileged person to say, like, actually, people do email us quite often to let us know that they're having an incredibly good time or, you know, they're getting ready for their wedding and they're like, here I am, like, putting on my Glossier Ultralip.
(44:21):
And it's like, thank you so much for, like, including us to be, like, part of this amazing day in your life or we get, like, people write letters to our office, like, telling us about, you know, ways that we've been a part of their journey.
And we read every single one of those letters and, like, really try to respond to as many as we can.
So, yeah, I do feel like, in many ways, like, there's something really exceptional and special about Glossier.
(44:47):
Fantastic.
That is a way to wrap up.
Amy, thank you so much for joining us this morning.
I can't drop my mic because it's attached to something.
But that was, that was a...
The sound engineer is nearly going to jump out of her chair at me.
Stop grabbing your mic.
Sorry, Elaine.
Yeah, stop it. Amy, thank you so much for joining us this afternoon, this morning.
(45:12):
And we hope you...
I'm going to have to do that bit again.
We're going to have to do an edit.
Are we going to edit out my mic drop that I didn't drop?
We're going to have to do a little edit there and rewind a little bit.
Amy, thank you so much for joining us and telling us about some of the things that Glossier have been up to.
(45:33):
He's laughing at me now.
I can't do it.
You're going to have to do it now.
You're going to have to do it,
I can't do it.
Amy, it's been a pleasure talking with you today.
Thanks so much for joining us.
Before I say goodbye, because you liked my pirate joke so much, can I tell you one more joke?
First, we need to tell the first pirate joke, because that was pre-recorded.
(45:56):
That was not on record.
You're going to have to double-joke.
That was PR.
Okay, so.
Okay, what's a pirate's favourite letter?
R.
You'd think it's the R, but it's the C they love.
Hardy-har-har-harr. So what's your second pirate joke?
How many tickles does it take to make an octopus laugh?
(46:17):
I don't know.
How many tickles does it take to make an octopus laugh?
10 tickles.
Oh, God Almighty, you'd make a great dad.
Is this your father-in-law who's arrived recently giving you all these good jokes?
No, I heard that on a podcast.
(46:37):
That was, like, the one of the first things I said to my partner of nearly 10 years.
That's a brilliant joke.
Thank you so much for joining us, Amy.
Thanks, Amy.
So, customer CX.
Obviously, we can't call this episode CX is Sexy because we did that last year, right?
(46:57):
We've already done it.
That would just be like going over old ground.
But I think this is, like, CXCA.
CXCA, yeah, yeah.
You pronounce it..
CXAI.
Give it a French pronunciation.
It sounds ultra good.
What, le CX?
But it's interesting.
I quite, I like the the CXS retention remark and I think it makes an awful lot of sense to look at how customer service and customer experience has influenced people to come back or not, you know, as the case may be.
(47:28):
I'd say by the time somebody gets in touch with customer service, you're probably almost halfway to losing them as a customer anyway, right?
Well, unless they're writing you a letter to say they've had a great time.
But yes, they're not reaching out to you because they're happy.
Normally.
And being able to tie in that customer outreach resolution and future recurring revenue, like, is an analytics quest that if you can go on really would prove the investment in some of the changes that people could be making.
(48:06):
But it must be totally, I mean, I presume you can tag easily enough, tag customers that got in touch with you through your various channels.
I'm sure, I'm sure.
Theoretically, that's very easy.
Train an AI to do it.
And who is going to pick up that piece of analysis on a day-to-day basis while everything else is on fire?
(48:27):
There is a couple of interesting ones, though.
I really like the comment on training the AI specifically for the tasks, because I know, you know, you get an AI and you just kind of want to dump everything into it and make one big washing machine of data and then hope that everything comes out.
And what is not a strategy?
(48:47):
Hope is not a strategy.
We hope that the AI is going to give the right answer.
Ah, there's a lot of hope and strategy in the world these days.
But we've managed, we've managed actually to go through two episodes without mentioning any geopolitical stuff.
We'll keep that going, right?
Yeah, this is a non-geopolitical zone.
(49:08):
This is a bubble, a bubble of its own ecommerce greatness.
But I am interested in that that point of having the different AIs for different jobs.
And identifying the jobs that are really good for AI to do and the actions that AI can take.
Like, I love the cancellation window.
I think that's brilliant because it's, like, that is a job that AI can do.
It's done.
(49:30):
So you haven't got to worry about it.
Two things.
One, your agent isn't having to do really mundane, boring stuff...
Yeah.
But number two, there's a better chance that the customer is going to get the service that they need in that situation.
So that's a brilliant, brilliant use case for AI.
I love the ticket example where, okay, so you've got tracking information and the AI then does nothing with it.
(49:56):
It's just like, oh, yeah, being able to then do something with that and make that bit meaningful - because there's nothing more frustrating than, like, I'm not contacting you because I don't know how to use a tracking number.
Of course, I know how to use a tracking number.
I get packages delivered every day.
So finding those areas to focus on and fix and have meaningful impact into the business, I think, is a spectacular use case for it.
(50:23):
I'd love to know who their vendors are.
Yeah.
Well, we don't.
No, we don't.
I think we did ask the question.
We did ask the question.
Yeah, it's super-confidential.
But those vendors sound super-good.
Super-good.
So if they want to respond to the LinkedIn post about it...
The, training the bots, that is definitely one that is of interest.
(50:46):
Parsing the tracking code, or the tracking information, is interesting as well.
So, you know, because the furthest most systems go is to say your order is processing, your order is processed, your order is part-shipped or whatever.
But that extra piece where you go to the, you follow, the AI follows the tracking link and then parses the information on the tracking page and then uses that somehow?
(51:09):
And infers whether or not this is an escalation requirement.
This is something that is working as it should.
Or this is something that is working as it shouldn't.
And then taking an action on it would be really interesting.
So to give you an example, package is stuck in a sort centre.
It's been there for five days.
Package is in Athlone for five days.
(51:29):
Yeah, your package is in Athlone.
I wonder what could be in Athlone that delays packages?
Most of the couriers in the country.
Yeah, everybody.
But your package has been in Athlone for five days and the AI is, like, right, okay, we're going to have to recall that package.
And it's going to take X number of days.
So instead, we're just going to send you a separate order right now.
We're going to send you a fresh order.
(51:50):
And then we're going to raise the request to recover that package from the carrier where it's stuck.
That's where it gets super-interesting because not only have you got the saving with the customer, but you also have the operational saving of making that contact, reaching out and resolving with the carrier.
Or the AI agent just files the carrier claim straight away.
(52:10):
And you're taking all of that operational, all of those operational issues away and moving.
And these are the kind of applications for AI within business and certainly within ecommerce operations that I find really exciting.
Not, like, oh, yeah, we can serve you the right level of article.
It's, we can get you this information, and then we can interpret the information, and then we can take action.
(52:37):
Yeah, that I love.
On very focused, frequently occurring issues where, you know, you can iteratively improve continuously,
it makes an enormous amount of sense.
And that's a really interesting approach to using AI.
What else did we learn as we were talking, apart from the possibility of dyeing sheets?
We learned some great jokes.
(52:58):
Are you slightly obsessed with these sheets now?
No, no, I quite like the company that is involved, though.
The jokes, obviously excellent.
What else did we learn about?
We learned about how important it is to manage the experience of employees.
Yeah.
And the impact that, like, working in CX can have on you as an individual.
(53:21):
Yeah.
Well, you know what I thought was interesting is it is the opportunity for people to work in different parts of the business.
So, like, Amy was mentioning she gets, she was, early doors,
she got the opportunity to work in the shop floor.
So she found out what it was like to work in the stores, which means you get to meet the customers face-to-face and know what they're like, and you know why they're excited about the products.
(53:41):
You get to meet the store assistants, who can teach you about the products.
Sometimes you might not know about the products, even though you work in the business.
You get the opportunity to do the pick-pack.
Sometimes you're answering the customer service queries.
I think it would be quite humbling and quite useful for business when staff are working in the different parts of the business where they also touch, like, how else would you know that you should never really ask for the warehouse to make an exception for you if you haven't tried to implement an exception yourself?
(54:08):
But do you know the thing that, like, managers dread?
Dread?
They dread it.
It's when, like, the CEO goes on, like, the front line for an hour and they experience a thing and then suddenly that thing becomes the most important thing to fix in the entire company.
(54:29):
And it could be an exception that happens, like, once every, like, 10,000 times.
And I know managers absolutely dread that.
But I do agree with the principle that everybody should spend time on the front line.
I think the other one is the CEO's friends.
Because the CEO's friend tried to do something in your business, buy something or whatever and didn't manage and then does.
(54:52):
Yeah, and it will be, and again it will be, like, one in those one in 10,000 exceptions.
Or the MD's grandmother's BlackBerry phone.
Yeah, any number of the above.
And then that becomes issue number one to fix whilst you've got lots of fires going on.
But wrapping it up, AI is incredibly useful and CX is a point of differentiation at a time when technology in the general, basic ecommerce technology the playing field is relatively level.
(55:26):
So you've got to find ways for your brand to differentiate and service is one of those areas where you can really excel.
Yeah, absolutely.
Great, Gordon, it's been a pleasure as ever.
The pleasure was all yours.
This is Functional & Fabulous, the omnichannel podcast where we unbox tales of online retail and digital transformation.
In this episode, celebrities lose their temper...
There was an incident where a celeb accused me of calling the paparazzi on them.
You'll hear horrifying stories that will curdle your blood...
(00:25):
That makes it sound very dramatic.
Ger has another bloody poem...
I like to spout poetry from time to time.
It takes care of the mundane to give you time for the meaningful.
Gordon explains what it's like to listen to Ger's poems...
Oh, you hurt me.
And oh, you hurt me again.
And he suggests a punishment even worse than listening to Ger's poems...
(00:50):
I actually think that everybody should be forced to work either in a contact centre.
100 per cent.
Or as a CX agent.
Ger gets his revenge...
He's laughing at me now.
I can't do it.
You're going to have to do it now.
You're going to have to do it,
I can't do it.
And there's another sodding pirate joke...
(01:13):
Hardy-har-har-harr. So what's your second pirate joke?
Two.
Two sodding pirate jokes.
This episode of Functional & Fabulous is brought to you with pride by StudioForty9, retail ecommerce experts, omnichannel growth consultants and cut-through performance marketing specialists.
StudioForty9, where your digital retail success is built.
(01:37):
Hello and you're very welcome to another episode of Functional & Fabulous.
Today we're joined by Amy Kemp, Director of Omnichannel Customer Experience at Glossier, a powerhouse beauty brand based in New York City.
Amy's career has been a little funky, to use her own words.
She's gone from NYC restaurants to med school startups, to leading CX at Glossier, and has tackled everything from panicked students to supply chain chaos on the journey.
(02:03):
Now she's making AI work for customers, using gizmos to handle Wizmos, and discovering that pick-pack satisfaction is very real.
We're talking good AI, bad AI, and what the hell AI, retail and digital, and why CX is the best job nobody notices until something goes wrong.
Amy Kemp, welcome to Functional & Fabulous.
Thank you for having me.
(02:25):
So Amy, what would you say is the important aspect of working in New York City restaurant scene as an ideal place to start to learn about handling customer experience issues?
You know, somebody who is hungry is really a particular customer.
And I think if you can learn the ins and outs of, you know, somebody who's maybe, like, on the edge of a nervous breakdown because they haven't eaten, making them feel settled and excited, you can probably tackle anything.
(02:52):
I feel like it's way more dramatic as well.
Like, the drama must be real in a New York City...
You know, my hunger is so immense.
I'm going to have a nervous breakdown if you don't bring me my spaghetti immediately.
Immediately.
I will also share, again, setting the scene.
I'm, you know, I'm 23.
Indie sleaze is still popular.
(03:12):
I was living in a basement, you know, really trying to figure out my life.
Working in the, you know, West Village in New York, which is a pretty chi-chi part of town, if you will.
Lots of celebs were coming in, you know, and I will share,
there was an incident where a celeb accused me of calling the paparazzi on them while they were eating at the restaurant.
(03:33):
And there I am, you know, fighting back tears, being, like, I don't even know how to do that.
Honest to God.
I think Robert De Niro is really reaching, you know?
I mean, seriously, what does he think?
Is he that important?
If you get to the other side of that, you can kind of take on anything.
So I can well imagine.
I mean, like, there's a number of different crucibles that we talked about.
(03:54):
You have the New York restaurant scene, panicking medical students.
The wire cut pick that effectively turned a small mattress company you worked for into a viral sensation.
So you've had a good few kind of interesting peaks of having to handle irritated people, I would say, over your career path.
(04:15):
I would agree with that.
And also, you know, managing my own feelings in this crazy world.
I'm probably the toughest customer out there.
I don't think that gets talked about enough.
That's true.
Yeah, 100 per cent.
It's totally real.
So tell us a little bit about Glossier.
How did it get started as a business?
(04:36):
And where is it today?
So Glossier was founded by Emily Weiss about 11 years ago at this point.
She was working predominantly in fashion.
She was going to the backstage of runway shows, working with models, working with Vogue, and started kind of getting very curious about all of these people who work in beauty, create all these beautiful looks.
(04:59):
What's in their kind of top shelf?
What's in their medicine cabinet at home that allows them to make people look so fabulous all the time?
And she started a blog called Into the Gloss that was primarily focused on kind of exploring what's in people's medicines cabinets that have these really interesting beauty backgrounds.
And after doing this for a number of years, you know, she understood, like, hey, there's kind of a gap in the market of some really amazing products, right?
(05:25):
Like, all of these people on the runway are using foundation, but they're first sheering it out with moisturiser.
Like, we have the opportunity to really develop something, like, in that space that fills that hole.
And she launched Glossier, which is, you know, the French dossier, but with gloss, Glossier.
And the brand really kind of took off and here we are still going strong.
(05:48):
I mean, it's an absolutely gorgeous website.
It looks like an absolutely beautiful product.
What's the clientele like?
I think there isn't any, like, one particular Glossier customer.
We, a couple, when I joined, we relaunched our Soho flagship store.
And sometimes when we open new stores, people will wait in, like, line for an insane number of hours to get in for the first time.
(06:14):
And we were walking down the line, kind of talking to people and hearing their stories.
And there was a woman who I think was probably in her seventies who told us, like, I am going to wear Glossier products until I'm in ashes.
Which I thought was so funny.
But because, like, our motto really is skincare first, make-up second, I think we offer a large catalogue of products where, like, anyone can really find their journey.
(06:40):
And within that journey for the customer, you're the director of Omnichannel Experience.
So what is your role in helping with that customer's journey?
Yeah, I think, so I have, I would think that I have kind of two big functions within my job.
One is, like, basically everything customer service kind of flows into me.
(07:02):
So really understanding, like, okay, you know, we're always going to have those questions, like, where is my package?
What's going on, et cetera, et cetera?
Like, how can we make sure we really have the right team in place or responding to people?
We have the right process and policy to support our customer.
So kind of that whole customer service piece.
And then the other function of my job is really understanding kind of our customer more globally and, like, what we need to do to continue iterating on our experiences holistically to keep our customer engaged, to keep that retention, that loyalty, et cetera.
(07:34):
And to me, that's really kind of one of my favourite parts of the job is, like, spending all this time in customer data, getting these clear insights and then basically going to other parts of the company and, like, trying to influence them to be, like, you know what would be really cool? Is if we did this instead of that.
And this is why.
(07:54):
So yeah, I would say, like, I'm an influencer.
I love that.
So it's kind of split, isn't it?
It's make customers happy and then make changes to keep them happy and keep them.
Yeah.
And, you know, I think in our, like, modern ecommerce world, pretty much every brand has competition, right?
And I think more and more brands are really competing on this idea of, like, loyalty and retention or, like, how are you distinct and how do you have a relationship with your customer where they're going to continue to, like, pick you versus somebody else?
(08:27):
Recording in progress.
"Recording in progress" - versus kind of somebody else in the space.
And I think so much of customer experience is about that, is about, like, really listening to your customer, like, hearing where we can still deliver and, like, making those changes to really build those meaningful relationships.
Yeah.
Do you think, like, the role of technology then plays a big part in this?
(08:50):
Because on one hand, you've got the commerce experience now, which is fairly leveled across most retailers.
And then you have the experience component where, okay, this is the battleground where we're going to compete.
That makes it sound very dramatic.
But if you're entering in that battleground, is there a role that technology needs to play and what role do you think it plays?
(09:14):
First of all, I went to performing arts high school, so I love the drama.
You can keep it coming through the entire conversation.
But absolutely.
I was thinking about kind of this morning, you know, as, like, a customer experience team, like, most teams in ecommerce, right?
We have, like, a couple, like, governing KPIs that we're, you know, responsible for delivering to the brand.
(09:37):
And one of those big KPIs is FRT first response time, right?
Like, how quickly do we get back to that customer when they reach out?
Like, I'm sure we've all had that experience of, you know, somebody, I don't know, charged us incorrectly.
And, like, we need to talk immediately.
Like, I need to get refunded.
And so I think first response time is really a critical part of that customer journey.
(09:58):
And I think holistically, we've kind of come at that problem really from, like, a staffing POV of looking at, like, okay, we know that we get more inbound tickets -
you know, we call people who reach out to us tickets - on Mondays.
So, like, we're going to staff for, like, this on Mondays and then this on Tuesday, and we'll put this team here and this team here.
(10:18):
And I think where the industry is going, which makes a lot of sense to me, is instead of looking at this from, like, a big holistic staffing POV is really looking at kind of the different reasons why people reach out to your brand.
And then kind of making the right processes and protocols based on contact reason.
And so, for example, you know, to give, like, an obvious one that I always give, you know, what we call, like, order changes.
(10:43):
So cancel requests and, like, address changes.
If somebody realises, okay, I've already said I'm moving, right?
It is, I bet you three to five times in this process, I'm going to deliver something and send it to my old apartment.
And I'm going to realise the second I look at that confirmation email, and I'm going to panic and cold sweat and email that brand immediately.
(11:07):
And, you know, at Glossier, we have a one-hour cancellation window.
If I can connect with someone in that window, great.
They can, you know, update my address.
It's fine.
If I can't, it's going to be a really bad experience.
So that to me feels like a great use case for AI, right?
Like, AI responds almost immediately.
So it's going to guarantee that we can connect with that person, perform that action, cancel that order, whatever it is, that again, is going to, like, build that customer loyalty.
(11:36):
And then I think I say this, like, kind of nested in the overall conversation of first response time, because we can still hit that first response time kind of on average by prioritising the tickets where speed really, really matters.
And then where it's less important, you know, we can follow different processes while still, like, holistically really ensuring that we're hitting those KPIs.
(12:01):
That's really interesting that you mentioned AI, because I think we're probably going to talk about that quite a bit.
One of the things that I think is probably important for us to flag earlier on is the difference between ticket deflection and actually providing meaningful resolution for customers using AI.
(12:24):
Because the temptation is, I'm going to deflect as many tickets as I can from agents.
I'm going to reduce the amount of headcount and the amount of costs that I've got.
And we're going to tick this big box to say, yes, we've done this and we're using AI.
But unless the resolution is meaningful, then you're just causing the brand a longer term problem, right?
(12:46):
I could not agree more.
I don't think I'm being hyperbolic when I say probably once or twice a day, I get an email from some type of AI sales company saying, I can promise you 100 per cent deflection.
And for my POV, I'm, like, I'm sure I could deliver 100 per cent deflection to myself and then I wouldn't have a business anymore, right?
(13:06):
Because why would someone bother reaching out to us when they're never going to connect with a human agent?
So I think what we have to really look at things, I try to make it all decisions kind of on a triangle.
So really thinking about the customer experience, but also the business experience, like, what's the right move for the experience?
So the business, excuse me, and then also kind of through an employee experience.
(13:30):
And when we can use that triangle to really look at kind of decision-making, generally we're doing something right.
Brilliant.
So when you think about an AI, I'm just going to obsess over the cancellation example for one second.
The AI has to not only be able to respond to the customer, but it has to be able to complete the action of cancellation, or the action of an address change.
(13:55):
And I think for people who haven't experienced this on the business side, this will appear quite revolutionary.
I'm assuming that's what happens with you guys.
The AI just, it can handle the process as well as the conversation.
Yeah, definitely.
And I think AI, depending on how well you train your AI, to use a kind of goofy term, it can also escalate things, right?
(14:20):
It can be that first line of defence.
And if something truly goofy is going on, it can be, like, this is outside of my realm of expertise.
Let's make sure we're connecting you with the right person.
Yeah, and again, that is fundamentally important.
But it deals with the customer because you've got, okay, here's the urgent thing that's very simple.
And that's now resolved.
Customer very, very happy.
(14:42):
Customer's got a bit more complexity, needs to talk to a real person.
And that person then needs to be empowered to, okay, we can fix this.
From anyone thinking about this, because I'm sure there'll be retailers listening, going, oh, this sounds, this is fantastic.
How do you even start on that journey of implementation?
Yeah, so for us, I think Glossier was a pretty early adopter of AI.
(15:06):
It predates me joining the brand.
So sadly, I can't report, like, what year they brought that on as a resource.
But in 2023, for kind of a series of reasons, we took a step back and looked at our AI strategy holistically.
I mean, obviously there've been a lot of advancements in AI over recent years.
And we really started by kind of obsessing over the reasons, like, the contact reasons, the reasons why customers, like, reach out to us in the first place.
(15:33):
And I think that, and not only why people reach out to us, like, when do we see X versus Y, right?
Like, we know if we have a promotion, we have an exciting product launch.
You know, probably, like, seven to 12 days later, that's when we're going to get a huge influx of, like, where's my package tickets, right?
And so really understanding kind of what our customer's journey is, and then working backwards again, looking at the use cases in which having again, that incredibly fast first response time is most important.
(16:07):
Got it.
So you have that.
And then how do you then deal with the more complex things?
Is that a case of, okay, we're going to route, and then we're going to train our, we're going to route the query over towards our agents, and we're going to train our agents, or do you then double down on your, how can we train the AI to do this better?
I think obviously with AI, like, you're constantly giving it feedback.
(16:31):
So in some ways, like, when we do see maybe something that wasn't perfectly handled, or we see something being escalated that we think that the AI can handle, like, we're going to give feedback and we're going to iterate and improve, but predominantly, like, pass it to a human agent, right?
I think for me, if the AI is no longer providing the value that we think the AI should be providing, which again is that incredibly fast response time or the ability to do that small action that has a downstream impact, why then try to fit, what's that expression - like, a square peg into a round hole?
(17:03):
Like, that's when I think we should definitely pass it to a human agent.
And when you're looking at the different types of tickets that you've got coming in, I'm assuming that 80 or 90 per cent of the tickets can be classified in very few ways.
Is that right?
Do you categorise that way?
Definitely.
I think we definitely see patterns and trends.
(17:25):
And then also there are only so many reasons, like, why someone's going to reach out to an ecommerce brand.
You know, it's so...
So Wizmo tickets, where's my package?
Or we sometimes call them package status at Glossier, so that's what I mean.
Obviously, that's a huge reason why people reach out to us.
And I don't think, I think sometimes the ways that, like, AI technology companies or technology companies in general, like, talk about package status tickets is, like, a little bit maybe unkind to customers.
(17:58):
You know, there's this idea, like, oh, you know, most brands use Shopify.
We use Shopify.
You know, Shopify has statuses that don't, you know, denote, like, oh, this order has been shipped.
This hasn't been, et cetera.
And so the AI promises, like, oh, well, we can go into Shopify.
We can, like, understand the order status and, like, we can pass it back to the customer.
Done.
(18:19):
Pretty much everyone buys everything online, right?
Like, everybody on a computer, like, probably has some working knowledge of how to understand what's going on on a tracking link.
I have a friend who works for the Home Shopping Network.
I don't know if you have, like, the equivalent?
(18:41):
Mhmm.
Okay, great. So could not be further away from ecommerce, you know, in terms of, like, her day-to-day job.
But recently made a joke.
She was, like, oh, you know, you're in trouble when you land on that, like, Narvar tracking page.
And I was, like, how do you know what Narvar is?
Right?
(19:01):
Like, I think most people have, like, a pretty good fundamental idea of, like, how this part of ecommerce works.
So if a customer is taking the time to reach out to you and say, where's my package?
What's going on?
It's probably not as simple as, like, it shipped.
And so for us finding kind of an AI partner who could help us with some of these queries, which again, we think if someone's scared about where their package is, a fast response really matters here.
(19:29):
It was having an AI partner who could actually open a tracking link and parse that information, like, a human and understand, okay, yep, this got lost somewhere in Norway.
And therefore, like, I understand that with in mind, like, let's take the next action step versus, like, well, it shipped. AI out.
(19:50):
Yeah, we all know how to check the status of a package.
But how do you have that contextual awareness of, this package is stuck?
Because, hey, that status hasn't moved for four days and it should have done.
What do you do next?
And that's pretty impressive because the standard, oh, yeah, we've got an AI response is,
(20:13):
here's your status.
Here's your status, order processing.
It sounds, like, as well, it sounds, like, it's important that you ring fence what the AI does and knows about, we'll say, and what it doesn't.
And you have to be pretty clear about that, you know.
So, you know, do you typically train the AIs with a lot of general knowledge or is it really, okay, here's what you know about and then after that, you just hand it over?
(20:39):
Yeah, so when we were looking at AI partners and we met with a lot of different people in the field, we wanted to go with a partner that makes, like, a specific AI bot for each contact reason.
And the reason why that was important to us is because we could basically give each bot, like, the amount of information it needs to do one very specific function and nothing else, right?
(21:04):
Reducing that chance for the AI to potentially hallucinate and, like, give the customer information that they don't really need or is not helpful.
I think, so that was important to us from, like, a decision lens.
Yeah, that's quite clever because I mean, there must be a huge, well, there must be a large temptation to just say, okay, we'll dump the entire knowledge base on the AI and then figure it all out, but no.
(21:25):
I have worked at Glossier for over two years and I would say, like, once a quarter, we're, like, this is the quarter we're going to go through our knowledge,
make sure everything is up to date, and we have never done that task.
And we will eventually, because I'm an optimist, but it just, to me, there was a lot of inherent danger, I think, in just giving the AI a ton of information that I can't even confidently say is up to date.
(21:52):
Yeah, keeping your knowledge base up to date... and these things just grow and grow and grow and you change process because, like, one of the fundamental parts of your job is changing the process.
So then is every single article in the knowledge base that references that process getting updated and how do you manage that?
If anyone can tell me, I would love to know.
(22:16):
Excellent, so we'll put that call out if anybody knows how to do that without having a small army of content editors, then we would all like to know, I think.
Answers on a postcard.
What coverage would you say would you, of your tickets, are, like, entirely handled, we'll say, to the satisfaction of customers by AI?
(22:36):
Would you know?
A lady never tells her secret.
A lady never tells.
A lady never tells - 82 per cent.
Okay.
100 per cent.
100 per cent ticket deflection.
Do you know, actually, on a slightly separate
but also a wondering note,
because I've often wondered, in Ireland
it's unusual for ecommerce merchants
(22:57):
to have support,
so chat support after, let's say, 5pm, 5.30pm,
that kind of a way,
but most people are actually buying in the evenings,
and I've always thought it would be
a really useful experiment to test
to see how having that additional support
in the evenings,
when people are on the stores and buying,
(23:17):
how useful that is?
From a customer experience point of view, have you experimented in that?
Would your staff be on after hours or in antisocial hours and things like that?
So I will say I think we're lucky
in that we span,
our team spans three time zones,
so we have some members in the UK,
(23:38):
some EST and then PST,
so we do have wide coverage on that front,
but I don't think we have full force
kind of explored it in the way that you're imagining,
but it's definitely something that we think about
in terms of you can't do it all,
and you can only really do
(23:58):
one or two things well at a time,
and so I think eventually we'll get there
on the priority matrix.
So, okay, so you've done quite a few things.
It sounds like you've done a lot more than a lot of businesses have in the area of trying to approach customer satisfaction via AI.
What else, what's next for Glossier?
(24:19):
What other things are you experimenting with?
That's a good question.
I think, right now,
we have really spent a lot of time
delving into kind of customer data
from a number of different lenses,
you know, NPS, CSAT,
like, ticket data, et cetera, et cetera, et cetera.
(24:40):
And have spent a lot of time
kind of pulling out themes
and also kind of mapping,
you know, obviously people give us feedback
and I, like, to kind of run that against
data regarding if and when
people return to shop with us again.
Because talking about prioritisation again,
there are 900 things
(25:00):
that I want to do in any given day,
but in order to make the right decisions,
what is the thing that is really going to
hurt our relationship with the customer?
And not to bring this full circle,
but I think one of the things
that the data shows very clearly,
we don't get a lot of,
like, in terms of NPS, let's say,
(25:21):
we don't get a lot of detractors
who say, like, customer experience
is the reason that I'm not coming,
or is the reason I'm unhappy
with this experience,
but they have one of the highest instances
of being kind of, like, spurned
and then not wanting to come back
and kind of shop with us again.
And I think at first, when I saw that, I was kind of scratching my head, being, like, have we done something that's really that horrible?
(25:45):
But it didn't take me long to really think, well, you're only interacting with customer service, right?
It's, like, something else has already gone wrong.
And then to have that experience and then to also not have the experience you want kind of with your CX, you know, with the customer service piece, like, yeah, you're probably not going to shop with this brand again.
(26:07):
And so I think a big focus is really on, like, CX as, like, retention.
That's an interesting one.
So, yeah, so basically you're looking at what the experience is and rather than, as you say, rather than measuring just NPS and CSAT, you're measuring what interactions they had and whether they came back and bought again or not.
(26:28):
Makes an awful lot of sense.
I think that goes beyond that and the point of, like, it being double bad.
It's, like, you're not, nobody ever calls you to say, hey, just calling to say everything's fantastic.
Just calling to say it's brilliant.
Love it.
They're calling, well, not calling or they're interacting with you because something awful is, well, not awful, but it's probably a bit dramatic.
(26:57):
Their package might be late and they're, like, why is my package stuck?
Or, like, why have you double-billed my credit card?
That could be pretty awful for somebody.
Not that double-billing of credit cards happens that often, but then to not be able to either connect with somebody and not be able to, or not be able to get a resolution to your satisfaction, then why would you go back to that brand?
(27:21):
Because you're, like, oh, you hurt me. And, oh, you hurt me again.
Something went wrong and also something could not be fixed.
Yeah, exactly.
So that's a super-valid point of, like, what is the role of CX?
It's not a cost centre.
We are fixing the problems that are caused by other things within our business.
(27:45):
And I don't even know it's necessarily problems with the business.
It's just, there's only so much we can really control.
I think so much of, like, when you interact with a customer who clearly is feeling quite frazzled, it's because something scary that they can't control is happening.
(28:08):
And so, yeah, it's really, I think, our value to make people feel seen and heard.
That's also a really valid point.
It reminds me, I was once asked a question, I'll pose the same question to you.
And it comes back to carriers.
I seem to be obsessed with carriers at the moment.
But if your carrier loses a package or if your carrier is not hitting their delivery promise, whose responsibility is that?
(28:36):
Is that the carrier's responsibility or is it the brand's responsibility?
Because very often it's outside of the brand's control.
Well, I might just adjust that a tiny bit.
I know that you're, I know that's a totally valid question, but I think in the mind of the customer, who are they holding responsible?
And yeah, who the customer holds responsible, I think is, I think that's obvious.
(29:01):
I think that's quite an obvious answer to that question.
So does the other answer matter then?
But what is the reality?
Tell us, Amy Kemp.
No, I totally agree.
Like, you know, I think no one's going to kind of look at the experience cohesively and say, like, well, at this point it really became the carrier's responsibility.
(29:23):
It's, like, no, it's their experience shopping with your brand, right?
And that really ends even after the package is delivered, right?
And they open it up and the order is as expected.
And even sometimes it goes further, right?
In which, like, I don't know, an item is missing and then they have that experience with, like, CX, et cetera, et cetera.
So yeah, you have to, like, look at it as a holistic experience.
(29:46):
I think you need to look at it holistically.
It can be very challenging in terms of level of influence.
And one of the things I started to see, which is quite interesting, again, on the carrier front and feedback in various tickets is why do you use X carrier?
It's a question that then gets asked.
(30:11):
So assuming that customers are not carrier-aware... I'm not quite sure where I'm going with this, by the way, just carrier bashing, probably.
Yeah, we definitely see that a great deal.
And there's no pattern, right?
It's, like, we send something in the US via USPS and it's, like, oh, I hate USPS.
And then we send something to someone else via DHL and they're, like, oh, I hate DHL.
(30:34):
Like, you know, like, it's just people have different experiences, different carriers and we can't really control for that.
It kind of comes down to if you've had one bad experience with one carrier that stays with you.
And the point that you've just raised of if you have a bad experience with the brand, you ain't coming back or you'd certainly go back under duress.
(30:56):
And tell us, Amy, do you get on the pick line much?
Do you get to the coalface?
Are you answering the tickets occasionally?
Or are you on a pedestal?
I pontificate over tickets.
Or do you actually participate in the tickets?
I try to do about an hour of tickets per week.
(31:19):
Sometimes it's more, sometimes it's less.
And to be honest with you, I'm actually, like, not a very good CX agent.
There are people who have a lot more talent at it than I do.
I will share somewhat recently, we had a launch that sold out very quickly, much more quickly than we anticipated.
(31:40):
And one of the tickets I worked at was a customer kind of expressing their frustration of, like, I really wanted to buy this.
And I couldn't because it was sold out.
And I think I probably had been having a long day.
And I was, like, I'm going to really take a moment and connect with this customer and really give them a peek behind the curtain.
No one's more disappointed than we are that we sell out too quickly.
(32:05):
And I'm so sorry about this.
I think I even sent an email, maybe, like, if I had a time machine... I really was on this journey.
And then they responded back to me very quickly with something along the lines of, get your stuff together.
And it made me realise, I am most effective when I am really looking at this kind of puzzle holistically and understanding what are the levers that we need to change as a business to again, really drive that loyalty and retention.
(32:38):
And then again, how can I influence those priorities across the business, in a way that makes sense with, again, like, our customer experience, our business needs, and also our employee experience.
I am never going to make every single person happy.
Mm-hmm.
Even though you're empathising and you're offering the sneak peek and all the rest of it.
(32:59):
In reality, they just didn't get what they wanted.
Everybody else got it.
They have enormous FOMO and they don't care.
And you need to get your shit together, is what I'd imagine was actually in that email, right?
It's interesting you touch on employee experience there.
I always think it's very much kind of a left behind thought or concern in businesses.
(33:21):
But, you know, amongst all other people who can most affect what's going on in your business, the employee satisfaction and keeping employees happy and the fact that employees are enjoying, let's say, to the most that they can, their work environment and so on is important and is going to have a huge benefit to your business.
(33:42):
How does Glossier think about that?
What are you guys doing in that regard?
Yeah, I mean, I kind of, to go back to AI for a second, I think one thing that's been very exciting about AI is, you know, in some ways, it helps with those kind of, like, very quick, kind of, like, immediate tickets that are probably not, you know, not the most, maybe, like, memorable customer interaction someone's going to have in a day.
(34:09):
And it frees up some of that, like, time and capacity for those moments in which you really need to spend time connecting with someone, right?
Or, like, somebody had, you know, a bad experience.
We need to do something kind of unique again to, like, rebuild that trust.
That's when the work is meaningful.
I think we say this all the time, but I'll just say it again here.
(34:29):
Like, being a CX agent is an incredibly hard job.
It's really hard.
People are, you know, the anger is coming towards you and we're not really in a position where we can say, well, it's not my fault.
Like, it's, you should talk to logistics, you know, because again, for the customer, right, it's, like, one cohesive brand experience.
(34:50):
And so I think really taking those moments to carve out times for, like, meaningful interaction are critical.
Both ways.
It's interesting.
So to be poetic about it, I like to spout poetry from time to time.
It takes care of the mundane to give you time for the meaningful.
Yeah.
Oh, I feel that should be, like, on a wall somewhere.
(35:12):
A t-shirt.
Do you know, I have to share a little, you know, CX experience.
My partner had, just before Christmas, had bought a coat for her father, let's say two weeks before Christmas.
It was the present for the family, for their father.
And the coat never turned up and it was, let's say, Monday and Christmas Day was, I don't know, was the day that they were presenting it.
(35:36):
It was Wednesday, Christmas Eve and it still hadn't arrived and she phoned up the brand and I obviously, working in ecommerce and actually know the brand, said, you know, kind of trained her a little bit and said, look, they may or may not be able to do a lot for you.
It's almost certainly stuck in a depot somewhere.
You know, take it easy on them.
(35:56):
Don't get, you know, don't start hammering into them and maybe they'll help you out.
So she phoned them up anyway, and the company is from the north of the country, and they're very lucky in that they have a lovely accent in English, when they speak English, but also they use the word wee a lot as in, you know, small or little.
(36:17):
And Maureen came off the phone anyway and I said, how did you get on?
And she said, oh, the girl said she'll have a wee look.
And it was just, like, with that kind of, I don't know, very, I don't know how to put it but it was such a charming way of saying I'm going to look into it for you - and got it sorted actually and basically an hour before we had to leave for her parents, the coat arrived in the door.
(36:42):
So it was a great story all round, but I thought I must use this term wee if anybody ever wants to complain about something, I'll have a wee look at it for them.
I love it.
I mean, you know, earlier, I guess in Q4 of 2024, I, like, similarly had an experience where I've gotten, we kind of covered earlier, I'm incredibly cheap.
(37:07):
I would say that's, like, my dominant personality trait.
Cheap? I thought you were the queen of sustainability.
Yeah, sure, sure, a lot of that.
Was that not it?
No, I misunderstood.
But I, we have, like, a lot of sheets.
Again, I used to work at a bedding company.
So there's a lot of sheets in this house that are looking, you know, pretty old and dingy.
And so I've recently gotten into this idea of dye baths, D-Y-E baths, where, like, you can send, you know, clothing, bedding there and then they can dye it a different colour and make it look kind of exciting and new to you again.
(37:39):
And I sent something to, some sheets to this dye bath, and they sent me back a dress.
And I was, like, this is not what I sent you, and not my dress.
And despite, you know, working in CX, I really got myself into a lather.
I was, like, oh, like, I want this...
like, now I have this person's dress in my house, and I called them and it was, like, the second I spoke to, like, a reasonable, cool human on the other side of the phone, I was, like, oh, it's fine.
(38:11):
Like, this will get worked out.
Because, is it a nice dress?
No, it's a horrible dress.
That's why it wasn't...
It was a very bright pink, which is really, I live in New York, you know.
But yeah, I think, like, we sometimes, like, forget about kind of the humanity in all of this.
(38:32):
And, like, you just need a little reminder.
Yeah.
Well, I think what's interesting about both of those stories is as soon as we connected with somebody that was able to diffuse the situation, either by being cool or through their use of wee,
and then, like, it would have been fine if the coat hadn't been delivered.
(38:54):
And immediately you feel better about the problem that's happened with this dress.
And I don't think we're in a position where AI can actually do that right now.
But importantly, it gives time for people, like, for the CX agents to be able to also not to be so knackered by answering, you know, where's my order all the time that they get the opportunity to actually deal with it humanely.
(39:21):
I actually think that everybody should be forced to work either in a contact centre...
100 per cent.
Or as a CX agent, or in retail or in a restaurant at some point.
Dealing with people.
Like, they should learn how to do that.
And then that will inject some sanity into the strange process exceptions that they're making.
(39:44):
Yeah, and that's a cool thing about Glossier.
When I joined, I think my first week, they were, like, okay, and on Friday, you're going to go work at the store for a day.
And, you know, are very good about being, like, great, like, you need to go to the store and, like, see how this works or have this experience, because it's - I think it's about, I don't think you really, like, understand, like, the full ecosystem of how the different parts of the business work together or, like, where there's overlap and where there's differences until you're, like, really in those spaces.
(40:12):
And then you understand when someone's, like, oh, can we just make an exception?
Just call the warehouse.
And that's the other place somebody should go and actually work.
Everybody should actually be on the receiving end of one of those calls to say,, like, can you just cancel this order that's already shipped, and deal with an exception, to understand what is actually happening.
(40:35):
When that's triggered.
You can't.
I think, you know, if it's going to the wrong address, send another package.
Like, it's really cruel.
Cruel to the team member in the warehouse.
Who's on their hands and knees, like, digging through, like, a pallet of, you know...
(40:58):
On the 20th of December...
Hasn't seen their family in a week.
Yeah.
That's interesting.
So you guys get to work in various departments and a little bit of cross-pollination of knowledge and ideas within the business, right?
Yeah, and that's, like, another, I think, special thing about Glossier is, you know, my title, right, is, like, omnichannel customer experience.
(41:25):
I have sat, you know, in the retail function of the business where it was, like, everything customer-facing is going to be in retail.
I now sit in, like, kind of more of the ecommerce function of the business.
So, like, that is a lot of kind of my day-over-day touchstones.
But my responsibility is really kind of, like, that cohesive omnichannel brand experience.
And I, you know, I hope that other brands, like, start to think about things in those ways too.
(41:48):
It's very interesting.
So Glossier in general, you know, where do you see the business going or, you know, what kinds of things are you interested in in the future just as an overall business?
I presume there's only so much, so many categories of make-up.
Maybe there isn't.
You can never have enough moisturiser.
(42:08):
Never?
Never.
Ever.
OK, so I'm wrong about that.
More moisturiser.
So after you've made more moisturiser, what's next for the business?
There are so many exciting things coming down the pipeline and I'm at my heart a real follower and I would hate to accidentally share something that I'm not supposed to share.
Oh, go on.
Go on.
More moisturiser, it's right.
(42:29):
Share a wee secret.
Share a wee secret with us.
But I, you know, I think we are having, you know, for so long the customer was really kind of focused on that ecommerce experience and, you know, and obviously then, like, ecommerce plus some retail and now we're also in wholesale.
So I think, you know, we're really kind of enjoying and figuring out the evolution of, like, what does it mean to be a brand that exists in all these different places and all these different ways and it's a very exciting problem to kind of, not even a problem, opportunity to come to.
(43:02):
And then I would say other than that, like, follow us on Instagram.
Smash the like button, subscribe.
I think if you can solve that problem of taking a very controllable customer experience of one channel of, okay, it might not be perfect, it might not be easy control, easy to control, but then you add in retail and then you add in wholesale and you still want to maintain that brand experience around CX.
(43:32):
That's an enormous challenge.
Like, that's going to keep you busy for the next, I don't know how many years.
Yeah, I will say this is an interesting, Glossier really does in many ways feel like a community.
I think the brand kind of emerged during such a, like, exciting time of during so many people's lives.
(43:54):
Like, there really is kind of, like, this deep affinity and I do feel like we're really still connected and tapped into that community.
You know, you mentioned earlier, like, no one ever, like, emails to say that they're having a good time.
I feel like I'm an incredibly privileged person to say, like, actually, people do email us quite often to let us know that they're having an incredibly good time or, you know, they're getting ready for their wedding and they're like, here I am, like, putting on my Glossier Ultralip.
(44:21):
And it's like, thank you so much for, like, including us to be, like, part of this amazing day in your life or we get, like, people write letters to our office, like, telling us about, you know, ways that we've been a part of their journey.
And we read every single one of those letters and, like, really try to respond to as many as we can.
So, yeah, I do feel like, in many ways, like, there's something really exceptional and special about Glossier.
(44:47):
Fantastic.
That is a way to wrap up.
Amy, thank you so much for joining us this morning.
I can't drop my mic because it's attached to something.
But that was, that was a...
The sound engineer is nearly going to jump out of her chair at me.
Stop grabbing your mic.
Sorry, Elaine.
Yeah, stop it. Amy, thank you so much for joining us this afternoon, this morning.
(45:12):
And we hope you...
I'm going to have to do that bit again.
We're going to have to do an edit.
Are we going to edit out my mic drop that I didn't drop?
We're going to have to do a little edit there and rewind a little bit.
Amy, thank you so much for joining us and telling us about some of the things that Glossier have been up to.
(45:33):
He's laughing at me now.
I can't do it.
You're going to have to do it now.
You're going to have to do it,
I can't do it.
Amy, it's been a pleasure talking with you today.
Thanks so much for joining us.
Before I say goodbye, because you liked my pirate joke so much, can I tell you one more joke?
First, we need to tell the first pirate joke, because that was pre-recorded.
(45:56):
That was not on record.
You're going to have to double-joke.
That was PR.
Okay, so.
Okay, what's a pirate's favourite letter?
R.
You'd think it's the R, but it's the C they love.
Hardy-har-har-harr. So what's your second pirate joke?
How many tickles does it take to make an octopus laugh?
(46:17):
I don't know.
How many tickles does it take to make an octopus laugh?
10 tickles.
Oh, God Almighty, you'd make a great dad.
Is this your father-in-law who's arrived recently giving you all these good jokes?
No, I heard that on a podcast.
(46:37):
That was, like, the one of the first things I said to my partner of nearly 10 years.
That's a brilliant joke.
Thank you so much for joining us, Amy.
Thanks, Amy.
So, customer CX.
Obviously, we can't call this episode CX is Sexy because we did that last year, right?
(46:57):
We've already done it.
That would just be like going over old ground.
But I think this is, like, CXCA.
CXCA, yeah, yeah.
You pronounce it..
CXAI.
Give it a French pronunciation.
It sounds ultra good.
What, le CX?
But it's interesting.
I quite, I like the the CXS retention remark and I think it makes an awful lot of sense to look at how customer service and customer experience has influenced people to come back or not, you know, as the case may be.
(47:28):
I'd say by the time somebody gets in touch with customer service, you're probably almost halfway to losing them as a customer anyway, right?
Well, unless they're writing you a letter to say they've had a great time.
But yes, they're not reaching out to you because they're happy.
Normally.
And being able to tie in that customer outreach resolution and future recurring revenue, like, is an analytics quest that if you can go on really would prove the investment in some of the changes that people could be making.
(48:06):
But it must be totally, I mean, I presume you can tag easily enough, tag customers that got in touch with you through your various channels.
I'm sure, I'm sure.
Theoretically, that's very easy.
Train an AI to do it.
And who is going to pick up that piece of analysis on a day-to-day basis while everything else is on fire?
(48:27):
There is a couple of interesting ones, though.
I really like the comment on training the AI specifically for the tasks, because I know, you know, you get an AI and you just kind of want to dump everything into it and make one big washing machine of data and then hope that everything comes out.
And what is not a strategy?
(48:47):
Hope is not a strategy.
We hope that the AI is going to give the right answer.
Ah, there's a lot of hope and strategy in the world these days.
But we've managed, we've managed actually to go through two episodes without mentioning any geopolitical stuff.
We'll keep that going, right?
Yeah, this is a non-geopolitical zone.
(49:08):
This is a bubble, a bubble of its own ecommerce greatness.
But I am interested in that that point of having the different AIs for different jobs.
And identifying the jobs that are really good for AI to do and the actions that AI can take.
Like, I love the cancellation window.
I think that's brilliant because it's, like, that is a job that AI can do.
It's done.
(49:30):
So you haven't got to worry about it.
Two things.
One, your agent isn't having to do really mundane, boring stuff...
Yeah.
But number two, there's a better chance that the customer is going to get the service that they need in that situation.
So that's a brilliant, brilliant use case for AI.
I love the ticket example where, okay, so you've got tracking information and the AI then does nothing with it.
(49:56):
It's just like, oh, yeah, being able to then do something with that and make that bit meaningful - because there's nothing more frustrating than, like, I'm not contacting you because I don't know how to use a tracking number.
Of course, I know how to use a tracking number.
I get packages delivered every day.
So finding those areas to focus on and fix and have meaningful impact into the business, I think, is a spectacular use case for it.
(50:23):
I'd love to know who their vendors are.
Yeah.
Well, we don't.
No, we don't.
I think we did ask the question.
We did ask the question.
Yeah, it's super-confidential.
But those vendors sound super-good.
Super-good.
So if they want to respond to the LinkedIn post about it...
The, training the bots, that is definitely one that is of interest.
(50:46):
Parsing the tracking code, or the tracking information, is interesting as well.
So, you know, because the furthest most systems go is to say your order is processing, your order is processed, your order is part-shipped or whatever.
But that extra piece where you go to the, you follow, the AI follows the tracking link and then parses the information on the tracking page and then uses that somehow?
(51:09):
And infers whether or not this is an escalation requirement.
This is something that is working as it should.
Or this is something that is working as it shouldn't.
And then taking an action on it would be really interesting.
So to give you an example, package is stuck in a sort centre.
It's been there for five days.
Package is in Athlone for five days.
(51:29):
Yeah, your package is in Athlone.
I wonder what could be in Athlone that delays packages?
Most of the couriers in the country.
Yeah, everybody.
But your package has been in Athlone for five days and the AI is, like, right, okay, we're going to have to recall that package.
And it's going to take X number of days.
So instead, we're just going to send you a separate order right now.
We're going to send you a fresh order.
(51:50):
And then we're going to raise the request to recover that package from the carrier where it's stuck.
That's where it gets super-interesting because not only have you got the saving with the customer, but you also have the operational saving of making that contact, reaching out and resolving with the carrier.
Or the AI agent just files the carrier claim straight away.
(52:10):
And you're taking all of that operational, all of those operational issues away and moving.
And these are the kind of applications for AI within business and certainly within ecommerce operations that I find really exciting.
Not, like, oh, yeah, we can serve you the right level of article.
It's, we can get you this information, and then we can interpret the information, and then we can take action.
(52:37):
Yeah, that I love.
On very focused, frequently occurring issues where, you know, you can iteratively improve continuously,
it makes an enormous amount of sense.
And that's a really interesting approach to using AI.
What else did we learn as we were talking, apart from the possibility of dyeing sheets?
We learned some great jokes.
(52:58):
Are you slightly obsessed with these sheets now?
No, no, I quite like the company that is involved, though.
The jokes, obviously excellent.
What else did we learn about?
We learned about how important it is to manage the experience of employees.
Yeah.
And the impact that, like, working in CX can have on you as an individual.
(53:21):
Yeah.
Well, you know what I thought was interesting is it is the opportunity for people to work in different parts of the business.
So, like, Amy was mentioning she gets, she was, early doors,
she got the opportunity to work in the shop floor.
So she found out what it was like to work in the stores, which means you get to meet the customers face-to-face and know what they're like, and you know why they're excited about the products.
(53:41):
You get to meet the store assistants, who can teach you about the products.
Sometimes you might not know about the products, even though you work in the business.
You get the opportunity to do the pick-pack.
Sometimes you're answering the customer service queries.
I think it would be quite humbling and quite useful for business when staff are working in the different parts of the business where they also touch, like, how else would you know that you should never really ask for the warehouse to make an exception for you if you haven't tried to implement an exception yourself?
(54:08):
But do you know the thing that, like, managers dread?
Dread?
They dread it.
It's when, like, the CEO goes on, like, the front line for an hour and they experience a thing and then suddenly that thing becomes the most important thing to fix in the entire company.
(54:29):
And it could be an exception that happens, like, once every, like, 10,000 times.
And I know managers absolutely dread that.
But I do agree with the principle that everybody should spend time on the front line.
I think the other one is the CEO's friends.
Because the CEO's friend tried to do something in your business, buy something or whatever and didn't manage and then does.
(54:52):
Yeah, and it will be, and again it will be, like, one in those one in 10,000 exceptions.
Or the MD's grandmother's BlackBerry phone.
Yeah, any number of the above.
And then that becomes issue number one to fix whilst you've got lots of fires going on.
But wrapping it up, AI is incredibly useful and CX is a point of differentiation at a time when technology in the general, basic ecommerce technology the playing field is relatively level.
(55:26):
So you've got to find ways for your brand to differentiate and service is one of those areas where you can really excel.
Yeah, absolutely.
Great, Gordon, it's been a pleasure as ever.
The pleasure was all yours.
Popular Podcasts
Bookmarked by Reese's Book Club
Welcome to Bookmarked by Reese’s Book Club — the podcast where great stories, bold women, and irresistible conversations collide! Hosted by award-winning journalist Danielle Robay, each week new episodes balance thoughtful literary insight with the fervor of buzzy book trends, pop culture and more. Bookmarked brings together celebrities, tastemakers, influencers and authors from Reese's Book Club and beyond to share stories that transcend the page. Pull up a chair. You’re not just listening — you’re part of the conversation.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!