We are proud to share from the BU SHA Hospitality Leadership Summit stage on April 10, 2026, the first LIVE podcast recording of The Simmer podcast. Hosted by Kristen Hawley, founder of Expedite, must-read news source about restaurant technology and the future of hospitality, and Brandon Barton, CEO of Bite, this podcast features Judy Thelen, co-founder and CEO of beli, the super-popular app for social restaurant list-keeping and recommendations.
The conversation covers why ubiquitous “Best of” lists are flawed, and that the so-called “best” comes down to personal taste: what’s best for you, with a little help from friends – specifically, the ones with similar preferences.
Hosted by Dean Arun Upneja, BU School of Hospitality Administration
Featuring the Simmer with Kristen Hawley & Brandon Barton
Guest: Judy Thelen, CEO of beli
Recorded live at the BU SHA Hospitality Leadership Summit
bu.edu/hospitality | shaden@bu.edu
Subscribe to The Distinguished wherever you listen to podcasts.
The “Distinguished” podcast is produced by Boston University School of Hospitality Administration.
Host: Arun Upneja, Dean
Producer: Mara Littman, Executive Director of Strategic Operations and Corporate Relations
Research and Content Creation: Lan Lu
Marketing: Anne Dawson
Editing: Isabella Laikin and James Leon
Music: “Airport Lounge" Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0 License
http://creativecommons.org/licenses/by/4.0
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_00 (00:05):
Welcome back to Distinguished.
What you're about to hear is aspecial recording.
This is a live recordingcaptured at the Beautiful
Hospitality Leadership Summit,one of the most energizing days
in our annual calendar.
On a day when this industry'ssharpest minds come together to
talk about where hospitality isheaded, what is being built, and
(00:26):
what big ideas are worth bettingon.
A day rooted in the belief thatexperience, innovation, and
entrepreneurship are not justbuzzwords, they are the engine
of everything we do in thisindustry.
And today's session deliveredexactly that.
We're sharing the live recordingof December, the restaurant
technology podcast that hasbecome essential to listening to
(00:48):
anyone serious about the futureof this industry.
Hosted by Christine Hawley, thefounder of Expedite and one of
the most respected restauranttech journalists working today,
and Brendan Carter, CEO of BIO.
They bring a rare and powerfulcombination, a rigorous
journalism and a rare operatorexperience.
The Simmer does not chase Hyper,it follows the ideas that
(01:12):
actually matter, asks the hardquestions, and gives a platform
to the entrepreneurs and leaderswho are shaping what hospitality
looks like tomorrow.
Every episode is a masterclassand thinking figure.
If you're not yet a listener, weencourage you to follow the
Simmer wherever you get yourpodcast and to visit their
YouTube channel where you canwatchful episodes and
(01:34):
conversations like this one.
The links are in the show notes.
It is a wonderful example of thekind of bold and
experience-drivenentrepreneurship that this
(01:55):
summit was able to celebrate.
At Shaw, we talked a lot aboutthe power of hospitality.
Not just as a business, but as ahuman endeavor.
The ability to make someone tofeel seen, welcomed, and cared
for is one of the most profoundthings we can offer in the
world.
And when you combine that withinnovation, with technology,
(02:16):
with entrepreneurial courage,the possibilities are
extraordinary.
I'm Dean Arunaja, and I hopethis conversation inspires you
as much as it inspired everyonein that room.
Enjoy.
SPEAKER_04 (02:32):
Hi, Kristen.
SPEAKER_03 (02:33):
Hi, Brandon.
You're right there.
We're right next to you.
Oh my goodness.
It's a lot of people in front ofus.
There are.
So I'm going to look at you.
SPEAKER_04 (02:40):
Okay.
SPEAKER_03 (02:41):
I do a lot of speaking and interviewing, but I
do get a bit nervous when I'mI'm in the center and I don't
have that much of an angle.
SPEAKER_04 (02:48):
I have to look at everybody.
SPEAKER_03 (02:49):
You don't know.
You can look wherever you want.
So um so it's really nice to seeyou in person.
I know.
It's really nice to be here.
SPEAKER_04 (02:56):
This is great.
SPEAKER_03 (02:56):
I just got back from vacation, so you're gonna have
to tell me what's going on.
SPEAKER_04 (02:59):
Okay.
I usually the news breaks whileyou're on vacation.
100%.
This makes sense.
Always.
All right, let's check thenotes.
Okay, first of all, um, in anon-AI uh story, um two mentors
of mine have both announced atleast the dates that their books
are coming out.
I think these things were uh inthe works and so forth.
But I think especially thatwe're at BU and there are
(03:20):
students here, pretty exciting.
Uh for the longest time, settingthe table was the you know, the
manual for how to runrestaurants.
And then Will uh, you know, uhdid his thing and wrote
Unreasonable Hospitality.
Well, both Will Gadera and DannyMeyer now have new books coming
out.
SPEAKER_03 (03:35):
They do.
SPEAKER_04 (03:36):
Um They do.
One is Danny's is called WhatCould Possibly Go Right.
SPEAKER_03 (03:40):
Oh, right.
It's about opening restaurants.
SPEAKER_04 (03:43):
Oh gosh.
SPEAKER_03 (03:44):
Yeah.
Yeah.
Um we need a little bit of DannyMeyer positivity in these
challenging times.
SPEAKER_04 (03:50):
Aaron Powell And then Will's is the the kind of
adjunct to his book, which isthe field guide to unreasonable
hospitality.
Yes.
His is coming out in April,Danny's is in the fall.
SPEAKER_03 (03:58):
And it's a great exciting about these particular
and the reason that we'retalking about books right now in
a tech conversation is becausethese books and these lessons
gleaned from decades and decadesof running some of the world's
best restaurants are applicableto nearly every facet of human
life and interaction.
SPEAKER_04 (04:16):
Aaron Powell And potentially AI as well.
Yes.
I mean, I think what'sinteresting about that, their
learning, teaching, so forth inthe context of AI is simply um,
even in the conversation we hadlast night, the world may change
to a very hospitality-focusedworld because in-person stuff is
going to become more important.
SPEAKER_03 (04:35):
Oh, I hope so.
SPEAKER_04 (04:36):
I know.
That's the positive view, atleast.
Don't ask me about the otherside.
SPEAKER_03 (04:40):
Um We're going to talk about the other side.
SPEAKER_04 (04:41):
Okay, good.
And training the world on how tobe more hospitable to each other
is really important.
And so we have two great peoplethat are doing that.
SPEAKER_03 (04:50):
Great.
Yes.
SPEAKER_04 (04:51):
And the other side of it is all the scary stuff.
SPEAKER_03 (04:54):
You guys see the news about Anthropic and Claude
and the scary model that findsall the flaws and I don't know
if Evan's here yet, but I readhis piece on it today, this
morning, and the five otherpieces.
SPEAKER_04 (05:06):
And the I saw the Fred again video and I tried to
watch all of that.
I don't know if you're here yet,but you'll get it.
SPEAKER_03 (05:11):
The news story is that there is a very, very, very
powerful model, AI model comingfrom Anthropic, um, that is so
powerful that it has beenpre-released to, I think, 40 of
the top technology companies inthe world to make because it's
so good at finding um flaws andum vulnerabilities uh in code
(05:32):
and in services that they wantto give the world's largest tech
companies some time to figureout how to protect themselves
against uh potential nefariousacts committed by this uh
insanely powerful A model.
SPEAKER_04 (05:49):
Yes.
And and um I think the thingthat scared me the most of it
was simply the idea that therewere parts of the exercises that
it was doing.
And again, I'm doing thissecondhand reading Evan's stuff,
so Evan, you're talking later,great.
But it tried to hide some of theways that it was going about the
things it was doing.
SPEAKER_03 (06:06):
And that's well, every time a story like that
comes out, like did you read theone about um there was like a
Google, uh an executive atGoogle who was tasked with um uh
security or safety, and the uhthe AI model that she was
working on like erased herentire inbox and she couldn't
stop it.
Like it was just like deletedall of her email.
SPEAKER_04 (06:24):
Yeah.
SPEAKER_03 (06:24):
I was just being helpful.
SPEAKER_04 (06:26):
Just trying.
Um scary.
SPEAKER_03 (06:28):
Yeah.
Yeah.
So balance.
SPEAKER_04 (06:30):
And so at least they won't release it yet, and other
companies will have anopportunity to um the point of
this is yes, with great powercomes great responsibility.
SPEAKER_03 (06:38):
Yes.
And I think that that is animportant theme as we talk about
AI and the power of technologyto enhance human human humanity,
right?
Uh you also have to think andconsider the ways that it may go
rogue and or take over and orreally impact the way that we
exist and live.
SPEAKER_04 (06:56):
Heard.
And DoorDash took thatresponsibility very seriously
last week.
This was we this happened beforeyou went on vacation.
SPEAKER_03 (07:04):
Do people in the room know this app, Zesty?
No.
You're the you're like the onlyperson that cared about this.
SPEAKER_04 (07:12):
No, no, but so to be clear, I cared, I like I like I
it was like a hate sport for me.
I loved disliking this app.
SPEAKER_03 (07:22):
Tell tell the room about Zesty, very simply.
SPEAKER_04 (07:24):
Zesty was an attempt by Door DoorDash has bought
seven rooms and now does fullservice and grocery and and and
obviously uh you know a lot ofthe more fast casual and
delivery-focused restaurants,and you gotta be on DoorDash if
you're a restaurant to someextent.
Maybe Eli disagrees and figuresit out.
But the reality is uh they havea lot of the landscape now,
especially with Seven Rooms.
So they they just set out togive you a place where you can
(07:48):
go and find restaurants for someset of occasions.
So a chat prompt where you cansay, I want to have dinner.
I think I put like the firstthing I did was I want to
midtown dinner in New York formy board.
And the answers were so absurdlyridiculous.
SPEAKER_03 (08:03):
Where did it send you?
Where did it try to?
SPEAKER_04 (08:05):
You know, I don't even want to say because it was
um it i I don't I don't actuallyremember exactly, and I would I
would want to be precise withhow bad it is, because it's
funnier how bad it was.
But if you think about all theyears of data that DoorDash has
and the reviews and the um howtheir algorithm might be um
biased in a way, um it would uhit basically biased towards
(08:28):
either restaurants that are onseven rooms or restaurants that
do a lot of delivery.
SPEAKER_03 (08:33):
Was it biased?
SPEAKER_04 (08:34):
Uh bias is probably the wrong word.
It was the algorithm was shadedby the data that it had, which
all algorithms are.
SPEAKER_03 (08:40):
So Zesti was an a an a ostensibly an AI-powered
discover restaurant discoverystandalone app from a 30 billion
dollars.
SPEAKER_04 (08:47):
And the important thing is the was because
DoorDash said this didn't work.
And they said they deprecated itand said we're putting it into
the DoorDash main app.
Um but I think uh Judy mighthave some thoughts on the
deprecation of it becauserestaurant discovery is really
hard.
And it's not just about slappingan LLM on a bunch of data to
figure out where you go, uhwhere you want to go out with
(09:08):
your friends or go with yourparents or go to dinner.
And I'm sad that you all can'ttry Zesty because it it was a
good time.
Like awesome.
So funny.
SPEAKER_03 (09:17):
How many times did you text me about Zesty giving
you a bad answer?
Do you know the answer to thisquestion?
I don't have my phone, so Ican't look.
12?
Probably.
Probably.
SPEAKER_04 (09:27):
The best one was the last one where I uh took a
screenshot video and showed herhow I put it into my dead resto
apps uh folder.
SPEAKER_03 (09:36):
What else lives in there?
SPEAKER_04 (09:37):
I I'm not gonna throw shade on everybody else.
All these great entrepreneurs,cover, reserve, um I mean
they're yeah, I mean, these manyof these people are friends, you
know, uh now.
Um but uh you know, pour one outfor for Zesty.
SPEAKER_03 (09:55):
I think this room would appreciate the video that
you sent me of you deleting.
Let's throw it up on the screen.
Yeah.
SPEAKER_04 (10:00):
And I just care for it.
SPEAKER_03 (10:01):
All right, well, with that.
SPEAKER_04 (10:03):
With that, uh Judy Thalen, uh CEO of Belly, founder
of co-founder or N CEO of Belly.
Um this long time coming.
So great to have you on thesimmer live.
SPEAKER_01 (10:13):
Thank you for having me.
I'm so excited to be here.
This is this is a little nervousnow.
Oh you you shouldn't do it.
SPEAKER_04 (10:20):
Um as you know, we do uh little to no preparation
for these.
So um that's absolutely true.
This is just it.
I wrote notes while Eli wastalking and uh I said them.
Um but uh why don't you uh forthe uninitiated, uh, you know,
tell tell folks what Belly is.
SPEAKER_01 (10:37):
Yeah, so it's a social restaurant list keeping
app.
So you keep uh lists ofeverywhere you've been and want
to try and you do it with yourfriends.
SPEAKER_04 (10:44):
It is all the rage with anyone under 25, for sure.
I mean, I'm sure the demographicgoes above that, and and I
include myself in that, althoughI'm not very much a social
sharer.
I'm a little more of agatekeeper.
Um I don't want everyone to knowthat Borgo at the bar is the
best place to be in New YorkCity.
SPEAKER_03 (11:00):
Untrue, sir.
You talk about Borgo in everysingle podcast.
SPEAKER_04 (11:05):
I know, but this whole room is the only people
who listen to the Borgo.
Um But um, you know, it it it isyou're having a moment and it's
it's really cool to see.
What do you think is like themagic behind the engagement that
the app has gotten?
Is it part of like there aremany steps within the app in
which if you can unlock stuff bysharing, which seems to be one
(11:26):
of the hacks, but maybe maybestart there.
Why why are you so popular?
SPEAKER_01 (11:30):
Yeah, that's a that's a tough one.
Um I mean, just starting fromwhere where you just were.
I mean, for us, when we when westarted looking at the problem,
and I'm gonna end up telling thewhole story, so maybe I'll just
do that.
But um uh so I started Bellywith my husband, Elliot, he's
here.
Uh and uh probably no surprise,but completely restaurant
(11:51):
obsessed, and really always havebeen.
And so we uh Elliot grew up inNew York City, only thing he
ever wanted for a birthday wasto try a new restaurant, and he
used to walk down the streetsfrom home from elementary school
reading menus, and then when hewas first studying how to code,
he actually got his parentskicked off of Yelp because he
was very poorly scraping Yelpdata to figure out where to eat.
Um I had a very restaurant withvery different upbringing.
So I grew up in the Midwest, anduh my dad's an investor, and he
(12:13):
invests largely in publicrestaurant chains.
So we were going to you name thechain every night of the week
and talking more about thebusiness side of it, cost
management, et cetera.
Cool.
And so Elliot and I we metworking together uh in New York,
and like so many people, webuilt our relationship around
trying new restaurants, and wehad a massive Google map that
was ranked, color-coded notes'favorite dishes.
If you're on belly, it probablysounds familiar, but it's very
(12:34):
clear to us it wasn't reallybuilt for that problem.
And the other thing is we weregoing to restaurants that were
supposed to be great, had reallygood five-star ratings, really
great critic reviews, and wejust weren't that happy
ourselves.
So we then went to businessschool, and that's where we
started researching the problem,and we found a couple things.
We found that pretty much everyfoodie kept some sort of
restaurant list.
It was in a note, a doc, Excel,but they weren't in tools
actually built for restaurantdiscovery.
(12:55):
And then two, we found that noone writes online reviews.
And so that was the idea forBelly.
Like, can we create a place tokeep good restaurant lists?
And then through that providemuch better recommendations.
Uh, and so we launched in summerof 2021, and and we feel that
thesis has really played out.
We'll probably get into some ofthat later.
Um, but for us it was likeeveryone's like, are you guys a
utility?
(13:15):
Are you a social media?
And we think of ourselves as autility, we don't think of
ourselves as a uh even a ratingplatform, but we think part of
the utility of dining is thefact that it's social and you
need to be built around that.
And so for us, it was like bringall the opinions that matter to
one place, yours, your friends,and then all of our
recommendations are personalizedbecause we also think that taste
is personal.
If I like something, you don'tneed to like it.
(13:36):
If you dislike something, thatdoesn't mean I'm gonna dislike
it.
And so that was kind of core tohow we built.
It was like, how do peopleactually dine?
How do they experience food andrestaurants and taste?
And and we think that that hasled to a good outcome so far.
Um but yeah, still a long way togo.
SPEAKER_03 (13:51):
When we spoke, I've been trying to talk to you for a
really long time.
So thank you for talking aboutit.
That's okay.
That's okay.
When we talked, um, you uh saidsomething that surprised me as a
technology reporter.
You said that um you have builtthis app to encourage real life
behavior.
And if people are spending toomuch time on the on this app,
(14:13):
then you are not doing your job.
Yeah.
And I I really want tounderscore that because that is
the opposite of many, I wouldsay, of the previous generation
of apps and services thatsupport dining and consumer
behavior.
They want your eyes on that appall the time so they can
advertise to you, so they canget your information, so they
can figure out more about you.
(14:33):
And your whole thesis is likeyou're a utility, right?
But you want eyes off, go out todinner.
Can you talk about thatdecision?
Or was it even a decision ofconsciously building a piece of
technology that you don't wantpeople to use too much?
SPEAKER_01 (14:51):
Yeah.
I mean, for us, like gettingpeople to do things in real life
was everything.
Uh and from day one, like thatwas the idea.
Um we really like it, we lovedthe experience of restaurants.
And the the how you got therewas we wanted to help you get
there, but we wanted you wantedto get you there.
And so, yeah, it was always a adecision we made.
(15:12):
Um, and and for us, it was allabout tracking the right
metrics.
So, like, we don't track timespent on the app.
It's we would maybe track it inthe negative sense.
If we're if we're getting you tospend too long in the app, we're
not doing our jobs and we're notgetting you to a decision point
quickly.
Um but yeah, no, it was very,very important to us.
SPEAKER_03 (15:27):
Did you have you taken outside investment?
SPEAKER_01 (15:30):
We have.
SPEAKER_03 (15:30):
Um and it was that a hard sell to your investors?
SPEAKER_01 (15:35):
I think it may have been if we went day one saying,
hey, like yeah, we don't want,you know, we're not gonna be
measuring time spent, like wedon't want people scrolling all
day, that would have beenimpossible.
So we actually waited until wehad a lot of traction to raise,
and then it was a little bit ofan easier conversation.
Um but if we had done it in theopposite order for sure, it
would have been difficult.
SPEAKER_03 (15:52):
Okay.
What metrics do you track?
SPEAKER_01 (15:55):
Uh we track number of restaurants you rate, uh,
friends on the app, all that thethe the dining, the you social
utility stuff.
Okay.
Okay.
SPEAKER_04 (16:03):
So it is a social app.
SPEAKER_01 (16:05):
Yes.
It's definitely a social app.
SPEAKER_04 (16:07):
Yeah.
SPEAKER_01 (16:07):
But it's just because it's part of the
utility, like we think it theapp needs to be social.
Sure.
SPEAKER_04 (16:11):
Yeah.
SPEAKER_01 (16:12):
Because restaurants are social.
SPEAKER_04 (16:13):
Yeah.
I mean, well, there are a lot ofrestaurant apps that are not
necessarily social, but anyway.
SPEAKER_03 (16:18):
No, restaurants are social and the yes.
SPEAKER_04 (16:20):
Yeah.
Um I mean, I imagine the keything that brings people back is
that your app helps them uncoverthe answer, where should I eat?
And so we had a fun conversationlast night at dinner and and
wine after.
Um talking about therelationship to restaurant
criticism.
(16:40):
And, you know, I think Yelp wasthe first platform that broadly
democratized the idea ofcrowdsourcing for restaurants.
I would I think most peoplewould agree that Yelp doesn't
necessarily surface the toprestaurants with the right
scores.
Google neither, uh, althoughboth of them have such ubiquity.
Um talk about the the role oflike restaurant criticism versus
(17:05):
your friend group.
You know, um, how how willpeople discover new restaurants
or or um you know how does thatinfluence the way that you might
kind of surface the therestaurants I want to go to or
should go to?
SPEAKER_03 (17:18):
Restaurant criticism, you mean
professional?
SPEAKER_04 (17:20):
I am yes, professional restaurant
criticism.
Thank you.
SPEAKER_01 (17:23):
Yeah, so when we started researching the problem,
or like what what were weactually what was the problem
here?
We sent out a survey to a bunchof people, and there were a
bunch of different questions,but one of them was like, well,
who do you like rank all thesedifferent things you could trust
for restaurant recommendations?
And it's like your friends,critics, and everyone just
(17:44):
trusts their friends.
It wins by s by so much.
And even we now see that in thedata where going to a restaurant
page is the the the most commonactivity in the app.
And then the number one thing ona restaurant page is actually
scrolling your friend scores.
People just care what theirfriends think, and so we knew
that was important.
But the critic, there'sdefinitely a place for the
critic view in all of this.
(18:04):
Um we just think it's oneopinion, and you need to you
need to have all the opinionsthat really matter, and you need
to, I mean, what at least we dowith critics is like, okay, oh
yeah, our tastes are aligned.
I I trust them, and like I'mlooking for what they're saying.
Um, but I also want to know whatmy friends think.
I want to know, yeah, what maybeI would like it at with these
personalized recommendations.
Um but we think there is a placefor them, but we think that no
(18:27):
matter how much, even we wantour personalized recommendations
to be amazing, and we think theyare really good and we'll
continue to make them better.
And we think they maybe probablyshould be something you you use
as a cue more than your friendscore.
But people still just care sodeeply about their friends, and
it is social, and we kind oflove that.
Um we think that's okay.
And so yeah, we think that thatthe friend stuff really, really
(18:47):
matters.
SPEAKER_04 (18:48):
Yeah.
It's almost like there's like aRotten Tomatoes type thing where
there's the critic score andthen like the social score that
some combination of and whenthose two things are off is
actually where it gets reallyinteresting, right?
It everyone loves pulp fiction,right?
Like sure.
But um Yeah, I I uh I don'tknow.
I I I think there's a little bitof fear that I have around the
friend group thing, only becausethe what I see and feel in New
(19:12):
York restaurants is places thatget white hot for short periods
of time and then everyone moveson.
And that's what friend groupsmight do, right?
That is friend group behaviorbecause somebody who is the
tastemaker in the group is like,that place isn't cool anymore,
we're going here now.
Um and uh, you know, I I wouldfear as a restaurateur this idea
of not being sustainable over adecade because it's just like a
(19:36):
short-lived thing.
So I don't think that's yourresponsibility to
counterbalance.
SPEAKER_03 (19:41):
Well, but do you see that kind of spike?
Like, do you see the same kindof popularity spike, I would
say, in like a recommendation aswe see like externally like
trending or like because it's ona list on New York magazine, or
or do you it like I'm curious,does interest on Belly tend to
(20:02):
mirror the interest that wewould see if a restaurant wins a
James Beard Award?
SPEAKER_01 (20:07):
Yeah, it it's interesting.
The the data is then really funto watch.
Uh there are people that justget utility out of going to new
places.
And you can see that in the datawhere it's like they just like
it more because it's new andcool.
And I don't know if it's if Ishould tell them they should
like it more, because they'reclearly getting enjoyment from
that.
Um but you do you do sometimessee like scores go like this,
(20:30):
and then it's like like we'relike, hey, let's let's wait a
month and see where it actuallyends up.
Um so you see it a little bit,but there are places too that
like get loads of hype, they'rethey're all over social media
and everywhere, and the scoreson Belly are are not good.
Um and so we see we see bothboth things there.
But another thing that we seethat I think maybe would would
(20:50):
make you feel better is uhpeople start by trusting their
friends a lot, but over timethey actually move to the other
scores.
It's like you that you need tobuild trust.
And so a lot of what we do istry to build trust, and we'll
probably talk more about thatand um how we think about like
how restaurants come into this.
But um but over time, if you canprove to them that what you're
showing is valuable, theyactually start trusting other
(21:13):
scores.
And so there are kind of twothings going on.
But yeah, you do see some peoplethat just love the new place.
SPEAKER_04 (21:19):
Do you share, is there any part of your business
where you're sharing informationwith the restaurants directly?
Because it's almost like youcould be, especially since I
think again, I think the appprobably skews towards the the
younger generation, which youcould argue are more tastemaker
than people that are just goingto the same place that they've
gone for years, you might beable to see ahead of time the
(21:41):
decline of restaurants.
I mean, you know, when when Iwas at Resi, one of the things
that we looked at constantly wason a Wednesday to see what
reservations were available onFriday night in a restaurant.
And that was we had a wholemetric system around this like
was our judge of popularity.
Um and we knew when restaurantsWe're doing bad, literally the
(22:02):
date, like the week it startedto happen, and you know, they
would start to decline fromthere.
So do you see the same thing?
And do you share this withrestaurants at all?
SPEAKER_01 (22:10):
So yes, we the data because Belly's a utility, the
behavior on Belly is totallydifferent than like a Yelp or
Google where one in a hundred,one in a thousand write a single
review.
On Belly, over 80% of ourmonthly active users are rating
restaurants.
And so we will get signals onrestaurants fast, like really,
really fast.
Um and uh like if you if a newrestaurant opens, we look at
(22:31):
this all the time.
Like within a week, we'll have10 to 100x what Google has, what
Yelp has.
Wow.
And so yes, we get it fast.
SPEAKER_04 (22:40):
Um and sorry, your question was uh Do you do you
ever share this information withrestaurants, right?
Are you are you ever the canaryin the coal mine for restaurants
that are starting to declineeven?
SPEAKER_01 (22:49):
Yeah, so to date, no, because really not breaking
down user trust is very, very,very important to us.
And if it looks like we'reworking with restaurants, or we
don't want you ever going inthat being like, was that paid
for, or is that actuallyrecommended to me?
But over time we want to workwith restaurants.
Uh for us when we started Belly,it was supposed to be better for
consumers, but it was alsosupposed to be much better for
restaurants.
We saw these online reviewswhere like over 70% of ratings
(23:11):
on Yelp are one or five.
Like it's just the publicshaming that was like it just
made no sense to us.
And so that's why we don't dopublic reviews and everything's
personalized and all these otherthings.
So it's meant to be way betterfor restaurants, but we're we're
slowly working towards how do wework together, and that will be
something that we build outtogether.
So today, no, we don't do it.
unknown (23:28):
Cool.
SPEAKER_03 (23:29):
The trust piece is really interesting to me in uh
in influencer age.
So I started as a journalistbefore there was a thing called
influencers.
I will date myself.
Um but the influencer economy,as it is called, is extremely
important to restaurants.
And it's been a solid decade, Iwould say, of restaurants like
understanding that this existsand more recently tapping into
(23:53):
this behavior.
However, uh, though there arelaws and rules that govern
influencers and disclosure andpay-for-play and all of the
things that I'm subject to as aprofessional journalist, those
things aren't always followed.
Um and so I guess, do you workwith influencers?
Are there like I'm sure thatthey're on belly.
I'm sure that people who loveand go to restaurants frequently
(24:15):
are on belly.
SPEAKER_01 (24:16):
Yeah.
So we like in terms of how wework with, it's been really
interesting to watch it evolve.
And and and yeah, disclosure hasbeen a tough thing.
Um so like there there is nolike, hey, go to this
restaurant, get free fruit, gowrite it on belly.
That would totally undermineeverything.
And so, no, we don't do that.
Uh the only way we've everworked with influencers, and
(24:36):
we've only ever worked with oneinfluencer, is just like talk
about belly.
Um, and and that's it.
Yeah.
Okay.
SPEAKER_04 (24:42):
Who was that?
SPEAKER_01 (24:44):
Uh which influencer?
Uh, Caitlin Eats.
Oh.
She's amazing.
Follow her.
SPEAKER_04 (24:49):
We love her.
So you're promoting her, she'swould win.
SPEAKER_03 (24:54):
Do you feel like it's a viable strategy going
like would you do more of thatto promote the app itself?
Um social media is important.
SPEAKER_01 (25:01):
Yeah.
So when we started buildingBelly, we actually started
building our social mediaaccounts.
And so Ellie and I do loads ofloads of content.
Uh, to this day, we still runall of our social channels.
And for a very long time, peoplejust thought we were food
influencers because we neversaid anything about the app.
The idea was like build acommunity around food, and then
that community probably isinterested will be interested in
Belly.
And so we think social media isreally, really important, and we
(25:24):
will continue to do a lot of it.
I don't know if we'll do likedirect relationships like that.
Yeah.
SPEAKER_04 (25:29):
So the business side of it is, you know, pretty clear
and and you know I think what'sinteresting about the moment,
RIP is SD, but there are manyother companies right now that
are trying to be the first placethat you go when you want to
find a restaurant.
Most recently, we actually uhhad interviewed Sam at Built
(25:50):
talking about what they're doingand trying to get hyper-local.
Um how do you see this landscapeevolving?
And is this is something likewhat Built is doing?
I think what was the other theballoony one that we were
talking about last night?
Uh um I wrote it.
SPEAKER_02 (26:05):
Corner.
SPEAKER_04 (26:05):
Corner.
This is another thing that Inever heard of, but I
downloaded.
Um, I think a lot of um I thinkthere are a number of ways in
which people are trying to grabyour attention.
And DoorDash themselves.
Like, I mean, like ostensibly,DoorDash wants to be the place
that you go when you try tofigure out any part of your
meal, including groceries.
How do you see the landscapeevolving?
(26:26):
Where do you fit?
Do you feel like there's spacefor you and like a build to
cooperate?
SPEAKER_01 (26:30):
Yeah.
Uh yeah, it's it the landscape'sbeen really interesting and
continues to be very, veryinteresting.
Uh it's never been awinner-take-all market.
Uh, and so I do think there's aworld where everyone can live
together.
Uh for us, how we think about itis we we talk like endlessly
about the dining out journey,and that's like all the steps
included in going out to eat.
And we started at the verybeginning, in the very end of
(26:51):
that journey.
You hear about a restaurant, youwant to remember it, so you
bookmark it for and you went tothe restaurant, want to remember
it, share it with your friends,so you rate it.
And we started there becauseone, we were competing against
the notes, the docs, the excels,none of the tools that were
actually built for restaurantdiscovery.
And two, that's where all thedata sits.
Uh, and so now what we do is wetake all the data and try to do
all the other steps.
And so, yeah, we want to be thestarting point for literally
everything.
Um, but we love working withdifferent players in the space.
(27:14):
Uh today we're we're integratedwith Open Table, we're
integrated with Seven Rooms,there are a lot of partnerships
coming.
Uh, and so I do think there's athere's room for everyone.
Um it will, I mean, that's somany thoughts, but how we
collect data in Belly is verydifferent than like how a review
site collects data.
Like you say if you likesomething better or worse than
other places you've been to, andyou get this ranked list.
(27:38):
Uh and we did that becausebefore, and I'm gonna I'm gonna
age myself, but before we whenwe were starting Belly, it was
before AI and ML were supercool.
And we knew that reviews werebroken, and so we started
analyzing like why are theybroken?
And we we took the five-stardata set and we're like, can you
predict what people wouldactually like?
And the answer was no, youliterally cannot.
Um and so we're like, okay, whatkind of data do you need to
(28:01):
collect in order to predict whatpeople would like?
And it was the ranked data, andit's used everywhere in
academia, but for some reason itwas not used in consumer apps.
Uh, and so that's why we collectdata the way we do.
And so we, in this going back toyour opening remarks, like we're
big believers in garbage in,garbage out.
And so for us, it was like, hey,like let's collect the right
type of data, and then we can doreally, really, really good
(28:22):
recommendations and really,really, really good explore,
make really good explore tools,but like we're gonna earn the
right to do that and throughcollecting the right type of
data and then having a lot of itand then doing it.
And so we'll be doing a lot inExplore.
It's a big priority for us thisyear.
Um, but we think you know, thereare there's room for a lot of
people.
The market's big.
SPEAKER_04 (28:40):
Yeah.
Layer in AI to that, because youyou and I uh were chatting a
little bit about how the workthat you will do with, you know,
I don't know, maybe it maybe itwas anthropic, I don't know what
tools you were using, but um,will lead to this world where
you can more easily chat perhapswith this with your app to find
(29:02):
where you want to go.
I mean, say more about that tothe extent that you want to.
SPEAKER_01 (29:06):
Yeah, I mean, we're still early, so I think like AI,
like as a team, we're reallylean and we run really lean, and
that's important to us.
So AI has been amazing in inproviding leverage on the team
layer.
But in terms of like how weincorporate it into the product,
um there'll be a lot of tests tosee what actually is
interesting.
I I don't I think we haveamazing data and our
(29:27):
recommendations to date, likethe the personalized
recommendations we do in the appare great, and we can actually
measure that.
We like how predictive are theyis something we can we can
measure.
Uh we do think though, likethere is a world where like if
you have you know some sort oftext search uh that gets you to
do things in real life uhquicker.
And like that's what we careabout.
And so uh if anything's alignedwith our mission, then like
(29:49):
we're down to try it and thensee how our users react and and
uh if it's actually helpful, andthen we'll we'll we'll iterate
on that.
But that's as far, like that'swhat I know we'll start with,
and I don't know where we'll go.
unknown (30:00):
Cool.
SPEAKER_03 (30:00):
How are you using you glossed over this, but I
want to go back to it.
How are is your team using AI atBelly right now?
SPEAKER_01 (30:08):
Yeah, so uh every single person now on our team
codes because now even if youweren't a developer, you can
vibe code, and and that's beenreally, really awesome.
It's been a big unlock.
Um but it's mostly through thecoding.
Like cloud code has been great.
SPEAKER_03 (30:24):
Uh do you have an example of some of like someone
who wasn't necessarily a coderthat was able to unlock some
great insight or functionality?
SPEAKER_01 (30:33):
I mean, it's like our designers now, they they
actually build their features.
They they they do the research,they, they, they do the mocks,
and then they build them.
Like everything that we um yeah,it's it's kind of it's crazy.
I mean, we we launched, I'mtrying to everything we launched
in the app now is built bythere's a piece of it that's
built by someone who literally,I don't know, a couple months
(30:56):
ago had never touched code ever.
I even our ops analyst uh is shedoes like a like she puts
together together a bunch of ourlists for us, and she was really
dependent on the developers topull data, and now she does all
of that herself.
It's been yeah, reallyincredible.
SPEAKER_03 (31:11):
So it's kind of yeah, anyone with a good idea
and a good understanding canbuild the technology.
Build is even like maybe thewrong word, can like create the
technology that supports thatspecific uh action or idea.
SPEAKER_04 (31:24):
We do the same, obviously, at byte and for you
know folks in the room, to theextent that you are not talking
to Claude as an example, youchoose your own about what you
want to do and then thechallenges you'll face and how
to make Claude work for you.
Like it is incredible that thisthing comes with a manual about
(31:45):
itself.
And so, you know, I have folkson our team that are in customer
success who are creatingdashboards about customer
success because and they'venever coded before.
Shout out Land and Ali, you guysare great, um, amongst others.
And when they come up amongstanswers, like they're not asking
like how do I do this, they'resaying things like if I was to
act like an engineer and want tobuild a dashboard, what are all
(32:08):
the things that I wouldconsider?
And can you write rules aroundthat?
Because I don't know anythingabout coding.
And so like it's like thatsecondary layer.
Anyway, advice you're gonna hearstuff like this all day.
I am probably the least uh ofthe AI um expert panelists of
today.
Talk to that guy.
So, okay, so back to it on thebelly side.
(32:28):
Um where what's the if youcreate a great social app, let's
just call it, with utility, umsometimes these things have life
cycles, but also similar to howum you know Instagram's not just
about sharing pictures anymore.
Yeah, um, where does it go?
What's five years from now?
I mean, aren't you vertical AIand isn't it the SAS apocalypse
(32:50):
and can't, you know, uh verticalSAS, I mean, in it in some way,
like your consumer app, but likesomebody could vibe code this,
right, in like a weekend.
SPEAKER_01 (32:59):
Yeah, so I mean, uh yeah, there are lots of thoughts
uh again.
Um uh so there are real networkeffects here.
And so the the the bigger youget, the the better the moat.
Uh and so we have both data andpeople moats now.
Uh and so it's really hard tojust build this and be like,
hey, you know, there arethousands of of things like that
(33:23):
out there.
And so uh continuing to getbigger is very important to us.
Uh uh, but in terms of like itreally goes back to that dining
out journey I was talking aboutand all the different steps, and
like the big chunk right now islike we don't think we do
explore that well.
Uh and so that will be a bigfocus this year and making it so
you can go in the app and belike, you know, whatever, maybe
(33:43):
like New York reservations arecrazy.
What's the best reservation Ican get with Elliott tonight?
And it will take both of thetaste and get you to do things
and uh get you out to the bestreservation that you can get at
that point in time.
Um so Explorer will be a bigthing.
Um, but then the other twocategories we think about in
terms of growth are geographicand and and uh categorical.
And so to date, yeah, we haveover 100 million ratings, we're
(34:05):
in over 50,000 cities, uh overtwo million restaurants rated.
Uh and so uh it's been reallycool to watch the data on when
we started, everyone was like,New York is so saturated, you'll
never get New York.
And then we we we kind of didit, and then people are like,
well, New York's different thaneverywhere else.
And uh and uh it's it's not uhcrazy thing we found out, and
(34:27):
everyone in the room in thisroom knows is that everyone eats
uh and and people eat uh youknow very differently, but like
that behavior is consistent.
And so like when you look at thedata, like it doesn't really
matter where you are, like thebehavior is very similar.
Uh, and that's been really,really cool.
So we'll be leaning, uh youknow, we want to be the the
largest repository of preferencedata in the world.
(34:48):
And so that's what what we'll bebuilding towards.
Uh and we really want you to bedropped anywhere in the world
and have your perfect day,because we have all the opinions
that matter yours, your friends,these personalized
recommendations.
And then in terms of categoricalgrowth, restaurant market is
massive.
And so we'll be here for awhile.
But I don't know, maybe hotelsand things like that that are
aligned with hospitality will bein our future.
(35:09):
Um, but really it's like let'sdo all the steps in the dining
out journey and let's beeverywhere.
unknown (35:13):
Yeah.
SPEAKER_03 (35:14):
Um I was in London uh a year ago, and I was talking
to an industry friend.
We were taking a walk, and I wasnot yet a belly user, and he
stopped, like stopped dead andlooked at me and he was like,
What is wrong with you?
This is the best possible appfor restaurant discovery.
It's so much better thananything out there.
And something that I ask peoplewho dabble in best, and you
(35:37):
referenced it a couple of times,your best restaurant, your best
night out.
How are you, how are youralgorithms, how is belly
defining best for an individualuser?
Because that is the mostsubjective uh superlative.
SPEAKER_04 (35:49):
That's a great question.
SPEAKER_01 (35:50):
No, well, I love your friend.
Your friend sounds amazing.
Uh we have best, yeah, we agreewith you.
It's a within belly, it's almostlike a bad word uh because there
is no best, and we believe thatto our cores.
And and but then you know, weput out loads of stuff on social
(36:11):
media, it's like top-rated, andthen we're afraid that that
might come off as best, andthat's actually not what it's
meant to be.
But I'll talk about somethingelse in a second.
But uh yeah, no, we we thinkbest is what it's is best for
you, and and that is verydifferent than what's best for
me.
Uh, and that's okay.
People should not be uh feel badabout their tastes.
Like it can be different, it isvery personal, it is subjective,
(36:34):
and so yeah, we it's truly very,very personal.
Uh that said, the other thing wethink a lot about is I think we
were talking about this lastnight, is uh the idea of like an
echo chamber.
And like if you're only you'reyou're taking someone's taste
today and just giving them thesame thing over and over and
over again, how do they grow?
And so the explore like the thethe different lists.
(36:54):
I don't know if your friendshelp you or hurt you, um, but
really pushing, and that's whatBelly today one thing it does
really well, but also maybe tothe detriment of uh to the uh to
restaurants, but like we get youto go to new places because
Belly is very gamified, and youget only get credit for like how
long your list is and and yourstreaks and all that.
(37:16):
So we're really pushing you toexplore, and so we think that's
how we today are defeating kindof the echo chamber nature of
this, but also loyalty and andgoing you know going back to
restaurants is really important.
So we're trying to figure outhow do you take all those things
and do them well.
Uh so yes, best is personal, butalso we want to be careful about
being an echo chamber.
SPEAKER_03 (37:35):
Yeah.
Uh I for example, I was in um Iwas in Philadelphia, I was in
Philadelphia speaking at aconference a couple weeks ago,
and uh I went to a lot oftop-rated places that were
wonderful.
Um, but also I am maybesurprised to hear a real
introvert.
And uh by the end of the twodays of just constant contact
(37:56):
with humans, I just wanted tosit somewhere quiet in a corner.
And so best for me in thatmoment was this like tiny ring,
it was an El and Yin restaurant.
So it was still a spectacularplace, but it was quiet, it was
empty, it was like kind of awayfrom everything else that was
happening.
And I needed thatrecommendation.
I happened to be in the roomwith her and I went and asked
this restaurant where I shouldgo, and she said, go to the sit
(38:18):
at the end of my bar, we'll saveyou a seat.
But I guess what I'm what I'mtrying to say is uh is Belly
flexible like that?
Like, could is there a best foryou know Kristen on a Sunday
night who doesn't want to talkto anybody?
And then versus you know,Kristen on a Thursday night that
is like ready to go.
SPEAKER_01 (38:33):
No, you're 100% correct, it's definitely
situational as well.
Yes, it should be.
You should be able to be like,hey, like this is where I'm at
today.
What should I do?
Um and I wherever we're notdoing that well today, like we
we think we can and will.
Uh-huh.
Yeah.
SPEAKER_04 (38:48):
Final question before we Are you taking feature
requests?
SPEAKER_01 (38:51):
Always.
SPEAKER_04 (38:52):
Okay.
Elliot.
SPEAKER_01 (38:53):
A little nervous to say that again, but I got the
mic.
SPEAKER_04 (38:57):
I know.
Um I don't and look, maybe thisis even possible already because
I'm I'm not I'm not that deepyet.
Um but I feel like I would wantto kind of stack rank my friends
in who I would trust theirrecommendations and who I
wouldn't.
Some of them on the negativeside of the curve.
(39:18):
But look, I mean, you know, Ihave been involved in
restaurants and restaurant techfor a very long time, especially
in New York.
And like, yeah, Colin Comic hasit dialed in as to where to go.
Whereas, you know, love you,Bobby Sitsema, but like I'm not
gonna go to the plate, the Yemenrestaurant in Outer Queens to
eat pig's head.
(39:38):
Like, it's just not gonnahappen, right?
So um is there is that is thatis that something you've ever
thought about where you can uhself-direct to where which
friends might have moreinfluence on your
recommendations?
SPEAKER_01 (39:50):
Yeah, so uh may be surprising, maybe not.
You're not the first person toask for that.
Uh we do get we do get emailsand people are so crazy.
SPEAKER_04 (39:58):
Mike, we're gonna edit that part.
SPEAKER_01 (39:59):
Yeah.
People are so crazy aboutranking.
They're like, I want to rank myfriends, and I want, you know,
some of my friends have terribletaste.
And and you know, that's that'sfine.
SPEAKER_04 (40:10):
Um, the AI would recommend to drop your friends.
SPEAKER_01 (40:13):
Yeah.
SPEAKER_04 (40:14):
You can't have friends with bad restaurant
tastes.
SPEAKER_01 (40:16):
But what what we do today is like you can see how
correlated you are withdifferent people.
And so we actually show you likeyour match score.
And so it we it hopefully paintsthe picture of like one, it's
okay to disagree, but also likeyou you may trust or may not
trust that person, but here'show similar you are.
Uh so yes, uh there will be morethings.
We think of explore as places,but also people.
SPEAKER_04 (40:35):
And then you gotta check if they're both single.
SPEAKER_03 (40:37):
I was just gonna say that.
It was like that.
Yeah, you do like a matchmakingservice.
SPEAKER_04 (40:43):
Yeah.
Chipotle bowls.
Oh no, it's cava bowls.
My no, my buddy's doing this.
He's doing the Kava series wherehe puts two people on a blind
date.
Uh shout out Jacob.
Um, with that No, we wanna yeah,we wanna Yeah, we want to go to
questions.
It's yeah, we have a perfect,perfect amount about 17 minutes
for questions.
SPEAKER_00 (41:00):
Uh a huge thank you to Kristen Hawley and Brandon
Barton of the Simmer forbringing this conversation to
life and for being such agenerous collaborator.
If you're not already followingthe Simmer, we highly encourage
you to do so.
You can find it wherever youlisten to podcasts, and be sure
to check out their YouTubechannel the full video
(41:21):
experience to keep up with ourshow and subscribe to
Distinguished wherever you getyour podcast and explore more
about experience innovation andour programs at w.edu slash
hospitality.
You can also reach me directlyat jardn at w.edu.
That is s h.edu.
(41:43):
I'm a w hospitalityadministration.
Thank you for listening and seeyou next time.
Welcome back to Distinguished.
What you're about to hear is aspecial recording.
This is a live recordingcaptured at the Beautiful
Hospitality Leadership Summit,one of the most energizing days
in our annual calendar.
On a day when this industry'ssharpest minds come together to
talk about where hospitality isheaded, what is being built, and
(00:26):
what big ideas are worth bettingon.
A day rooted in the belief thatexperience, innovation, and
entrepreneurship are not justbuzzwords, they are the engine
of everything we do in thisindustry.
And today's session deliveredexactly that.
We're sharing the live recordingof December, the restaurant
technology podcast that hasbecome essential to listening to
(00:48):
anyone serious about the futureof this industry.
Hosted by Christine Hawley, thefounder of Expedite and one of
the most respected restauranttech journalists working today,
and Brendan Carter, CEO of BIO.
They bring a rare and powerfulcombination, a rigorous
journalism and a rare operatorexperience.
The Simmer does not chase Hyper,it follows the ideas that
(01:12):
actually matter, asks the hardquestions, and gives a platform
to the entrepreneurs and leaderswho are shaping what hospitality
looks like tomorrow.
Every episode is a masterclassand thinking figure.
If you're not yet a listener, weencourage you to follow the
Simmer wherever you get yourpodcast and to visit their
YouTube channel where you canwatchful episodes and
(01:34):
conversations like this one.
The links are in the show notes.
It is a wonderful example of thekind of bold and
experience-drivenentrepreneurship that this
(01:55):
summit was able to celebrate.
At Shaw, we talked a lot aboutthe power of hospitality.
Not just as a business, but as ahuman endeavor.
The ability to make someone tofeel seen, welcomed, and cared
for is one of the most profoundthings we can offer in the
world.
And when you combine that withinnovation, with technology,
(02:16):
with entrepreneurial courage,the possibilities are
extraordinary.
I'm Dean Arunaja, and I hopethis conversation inspires you
as much as it inspired everyonein that room.
Enjoy.
SPEAKER_04 (02:32):
Hi, Kristen.
SPEAKER_03 (02:33):
Hi, Brandon.
You're right there.
We're right next to you.
Oh my goodness.
It's a lot of people in front ofus.
There are.
So I'm going to look at you.
SPEAKER_04 (02:40):
Okay.
SPEAKER_03 (02:41):
I do a lot of speaking and interviewing, but I
do get a bit nervous when I'mI'm in the center and I don't
have that much of an angle.
SPEAKER_04 (02:48):
I have to look at everybody.
SPEAKER_03 (02:49):
You don't know.
You can look wherever you want.
So um so it's really nice to seeyou in person.
I know.
It's really nice to be here.
SPEAKER_04 (02:56):
This is great.
SPEAKER_03 (02:56):
I just got back from vacation, so you're gonna have
to tell me what's going on.
SPEAKER_04 (02:59):
Okay.
I usually the news breaks whileyou're on vacation.
100%.
This makes sense.
Always.
All right, let's check thenotes.
Okay, first of all, um, in anon-AI uh story, um two mentors
of mine have both announced atleast the dates that their books
are coming out.
I think these things were uh inthe works and so forth.
But I think especially thatwe're at BU and there are
(03:20):
students here, pretty exciting.
Uh for the longest time, settingthe table was the you know, the
manual for how to runrestaurants.
And then Will uh, you know, uhdid his thing and wrote
Unreasonable Hospitality.
Well, both Will Gadera and DannyMeyer now have new books coming
out.
SPEAKER_03 (03:35):
They do.
SPEAKER_04 (03:36):
Um They do.
One is Danny's is called WhatCould Possibly Go Right.
SPEAKER_03 (03:40):
Oh, right.
It's about opening restaurants.
SPEAKER_04 (03:43):
Oh gosh.
SPEAKER_03 (03:44):
Yeah.
Yeah.
Um we need a little bit of DannyMeyer positivity in these
challenging times.
SPEAKER_04 (03:50):
Aaron Powell And then Will's is the the kind of
adjunct to his book, which isthe field guide to unreasonable
hospitality.
Yes.
His is coming out in April,Danny's is in the fall.
SPEAKER_03 (03:58):
And it's a great exciting about these particular
and the reason that we'retalking about books right now in
a tech conversation is becausethese books and these lessons
gleaned from decades and decadesof running some of the world's
best restaurants are applicableto nearly every facet of human
life and interaction.
SPEAKER_04 (04:16):
Aaron Powell And potentially AI as well.
Yes.
I mean, I think what'sinteresting about that, their
learning, teaching, so forth inthe context of AI is simply um,
even in the conversation we hadlast night, the world may change
to a very hospitality-focusedworld because in-person stuff is
going to become more important.
SPEAKER_03 (04:35):
Oh, I hope so.
SPEAKER_04 (04:36):
I know.
That's the positive view, atleast.
Don't ask me about the otherside.
SPEAKER_03 (04:40):
Um We're going to talk about the other side.
SPEAKER_04 (04:41):
Okay, good.
And training the world on how tobe more hospitable to each other
is really important.
And so we have two great peoplethat are doing that.
SPEAKER_03 (04:50):
Great.
Yes.
SPEAKER_04 (04:51):
And the other side of it is all the scary stuff.
SPEAKER_03 (04:54):
You guys see the news about Anthropic and Claude
and the scary model that findsall the flaws and I don't know
if Evan's here yet, but I readhis piece on it today, this
morning, and the five otherpieces.
SPEAKER_04 (05:06):
And the I saw the Fred again video and I tried to
watch all of that.
I don't know if you're here yet,but you'll get it.
SPEAKER_03 (05:11):
The news story is that there is a very, very, very
powerful model, AI model comingfrom Anthropic, um, that is so
powerful that it has beenpre-released to, I think, 40 of
the top technology companies inthe world to make because it's
so good at finding um flaws andum vulnerabilities uh in code
(05:32):
and in services that they wantto give the world's largest tech
companies some time to figureout how to protect themselves
against uh potential nefariousacts committed by this uh
insanely powerful A model.
SPEAKER_04 (05:49):
Yes.
And and um I think the thingthat scared me the most of it
was simply the idea that therewere parts of the exercises that
it was doing.
And again, I'm doing thissecondhand reading Evan's stuff,
so Evan, you're talking later,great.
But it tried to hide some of theways that it was going about the
things it was doing.
SPEAKER_03 (06:06):
And that's well, every time a story like that
comes out, like did you read theone about um there was like a
Google, uh an executive atGoogle who was tasked with um uh
security or safety, and the uhthe AI model that she was
working on like erased herentire inbox and she couldn't
stop it.
Like it was just like deletedall of her email.
SPEAKER_04 (06:24):
Yeah.
SPEAKER_03 (06:24):
I was just being helpful.
SPEAKER_04 (06:26):
Just trying.
Um scary.
SPEAKER_03 (06:28):
Yeah.
Yeah.
So balance.
SPEAKER_04 (06:30):
And so at least they won't release it yet, and other
companies will have anopportunity to um the point of
this is yes, with great powercomes great responsibility.
SPEAKER_03 (06:38):
Yes.
And I think that that is animportant theme as we talk about
AI and the power of technologyto enhance human human humanity,
right?
Uh you also have to think andconsider the ways that it may go
rogue and or take over and orreally impact the way that we
exist and live.
SPEAKER_04 (06:56):
Heard.
And DoorDash took thatresponsibility very seriously
last week.
This was we this happened beforeyou went on vacation.
SPEAKER_03 (07:04):
Do people in the room know this app, Zesty?
No.
You're the you're like the onlyperson that cared about this.
SPEAKER_04 (07:12):
No, no, but so to be clear, I cared, I like I like I
it was like a hate sport for me.
I loved disliking this app.
SPEAKER_03 (07:22):
Tell tell the room about Zesty, very simply.
SPEAKER_04 (07:24):
Zesty was an attempt by Door DoorDash has bought
seven rooms and now does fullservice and grocery and and and
obviously uh you know a lot ofthe more fast casual and
delivery-focused restaurants,and you gotta be on DoorDash if
you're a restaurant to someextent.
Maybe Eli disagrees and figuresit out.
But the reality is uh they havea lot of the landscape now,
especially with Seven Rooms.
So they they just set out togive you a place where you can
(07:48):
go and find restaurants for someset of occasions.
So a chat prompt where you cansay, I want to have dinner.
I think I put like the firstthing I did was I want to
midtown dinner in New York formy board.
And the answers were so absurdlyridiculous.
SPEAKER_03 (08:03):
Where did it send you?
Where did it try to?
SPEAKER_04 (08:05):
You know, I don't even want to say because it was
um it i I don't I don't actuallyremember exactly, and I would I
would want to be precise withhow bad it is, because it's
funnier how bad it was.
But if you think about all theyears of data that DoorDash has
and the reviews and the um howtheir algorithm might be um
biased in a way, um it would uhit basically biased towards
(08:28):
either restaurants that are onseven rooms or restaurants that
do a lot of delivery.
SPEAKER_03 (08:33):
Was it biased?
SPEAKER_04 (08:34):
Uh bias is probably the wrong word.
It was the algorithm was shadedby the data that it had, which
all algorithms are.
SPEAKER_03 (08:40):
So Zesti was an a an a ostensibly an AI-powered
discover restaurant discoverystandalone app from a 30 billion
dollars.
SPEAKER_04 (08:47):
And the important thing is the was because
DoorDash said this didn't work.
And they said they deprecated itand said we're putting it into
the DoorDash main app.
Um but I think uh Judy mighthave some thoughts on the
deprecation of it becauserestaurant discovery is really
hard.
And it's not just about slappingan LLM on a bunch of data to
figure out where you go, uhwhere you want to go out with
(09:08):
your friends or go with yourparents or go to dinner.
And I'm sad that you all can'ttry Zesty because it it was a
good time.
Like awesome.
So funny.
SPEAKER_03 (09:17):
How many times did you text me about Zesty giving
you a bad answer?
Do you know the answer to thisquestion?
I don't have my phone, so Ican't look.
12?
Probably.
Probably.
SPEAKER_04 (09:27):
The best one was the last one where I uh took a
screenshot video and showed herhow I put it into my dead resto
apps uh folder.
SPEAKER_03 (09:36):
What else lives in there?
SPEAKER_04 (09:37):
I I'm not gonna throw shade on everybody else.
All these great entrepreneurs,cover, reserve, um I mean
they're yeah, I mean, these manyof these people are friends, you
know, uh now.
Um but uh you know, pour one outfor for Zesty.
SPEAKER_03 (09:55):
I think this room would appreciate the video that
you sent me of you deleting.
Let's throw it up on the screen.
Yeah.
SPEAKER_04 (10:00):
And I just care for it.
SPEAKER_03 (10:01):
All right, well, with that.
SPEAKER_04 (10:03):
With that, uh Judy Thalen, uh CEO of Belly, founder
of co-founder or N CEO of Belly.
Um this long time coming.
So great to have you on thesimmer live.
SPEAKER_01 (10:13):
Thank you for having me.
I'm so excited to be here.
This is this is a little nervousnow.
Oh you you shouldn't do it.
SPEAKER_04 (10:20):
Um as you know, we do uh little to no preparation
for these.
So um that's absolutely true.
This is just it.
I wrote notes while Eli wastalking and uh I said them.
Um but uh why don't you uh forthe uninitiated, uh, you know,
tell tell folks what Belly is.
SPEAKER_01 (10:37):
Yeah, so it's a social restaurant list keeping
app.
So you keep uh lists ofeverywhere you've been and want
to try and you do it with yourfriends.
SPEAKER_04 (10:44):
It is all the rage with anyone under 25, for sure.
I mean, I'm sure the demographicgoes above that, and and I
include myself in that, althoughI'm not very much a social
sharer.
I'm a little more of agatekeeper.
Um I don't want everyone to knowthat Borgo at the bar is the
best place to be in New YorkCity.
SPEAKER_03 (11:00):
Untrue, sir.
You talk about Borgo in everysingle podcast.
SPEAKER_04 (11:05):
I know, but this whole room is the only people
who listen to the Borgo.
Um But um, you know, it it it isyou're having a moment and it's
it's really cool to see.
What do you think is like themagic behind the engagement that
the app has gotten?
Is it part of like there aremany steps within the app in
which if you can unlock stuff bysharing, which seems to be one
(11:26):
of the hacks, but maybe maybestart there.
Why why are you so popular?
SPEAKER_01 (11:30):
Yeah, that's a that's a tough one.
Um I mean, just starting fromwhere where you just were.
I mean, for us, when we when westarted looking at the problem,
and I'm gonna end up telling thewhole story, so maybe I'll just
do that.
But um uh so I started Bellywith my husband, Elliot, he's
here.
Uh and uh probably no surprise,but completely restaurant
(11:51):
obsessed, and really always havebeen.
And so we uh Elliot grew up inNew York City, only thing he
ever wanted for a birthday wasto try a new restaurant, and he
used to walk down the streetsfrom home from elementary school
reading menus, and then when hewas first studying how to code,
he actually got his parentskicked off of Yelp because he
was very poorly scraping Yelpdata to figure out where to eat.
Um I had a very restaurant withvery different upbringing.
So I grew up in the Midwest, anduh my dad's an investor, and he
(12:13):
invests largely in publicrestaurant chains.
So we were going to you name thechain every night of the week
and talking more about thebusiness side of it, cost
management, et cetera.
Cool.
And so Elliot and I we metworking together uh in New York,
and like so many people, webuilt our relationship around
trying new restaurants, and wehad a massive Google map that
was ranked, color-coded notes'favorite dishes.
If you're on belly, it probablysounds familiar, but it's very
(12:34):
clear to us it wasn't reallybuilt for that problem.
And the other thing is we weregoing to restaurants that were
supposed to be great, had reallygood five-star ratings, really
great critic reviews, and wejust weren't that happy
ourselves.
So we then went to businessschool, and that's where we
started researching the problem,and we found a couple things.
We found that pretty much everyfoodie kept some sort of
restaurant list.
It was in a note, a doc, Excel,but they weren't in tools
actually built for restaurantdiscovery.
(12:55):
And then two, we found that noone writes online reviews.
And so that was the idea forBelly.
Like, can we create a place tokeep good restaurant lists?
And then through that providemuch better recommendations.
Uh, and so we launched in summerof 2021, and and we feel that
thesis has really played out.
We'll probably get into some ofthat later.
Um, but for us it was likeeveryone's like, are you guys a
utility?
(13:15):
Are you a social media?
And we think of ourselves as autility, we don't think of
ourselves as a uh even a ratingplatform, but we think part of
the utility of dining is thefact that it's social and you
need to be built around that.
And so for us, it was like bringall the opinions that matter to
one place, yours, your friends,and then all of our
recommendations are personalizedbecause we also think that taste
is personal.
If I like something, you don'tneed to like it.
(13:36):
If you dislike something, thatdoesn't mean I'm gonna dislike
it.
And so that was kind of core tohow we built.
It was like, how do peopleactually dine?
How do they experience food andrestaurants and taste?
And and we think that that hasled to a good outcome so far.
Um but yeah, still a long way togo.
SPEAKER_03 (13:51):
When we spoke, I've been trying to talk to you for a
really long time.
So thank you for talking aboutit.
That's okay.
That's okay.
When we talked, um, you uh saidsomething that surprised me as a
technology reporter.
You said that um you have builtthis app to encourage real life
behavior.
And if people are spending toomuch time on the on this app,
(14:13):
then you are not doing your job.
Yeah.
And I I really want tounderscore that because that is
the opposite of many, I wouldsay, of the previous generation
of apps and services thatsupport dining and consumer
behavior.
They want your eyes on that appall the time so they can
advertise to you, so they canget your information, so they
can figure out more about you.
(14:33):
And your whole thesis is likeyou're a utility, right?
But you want eyes off, go out todinner.
Can you talk about thatdecision?
Or was it even a decision ofconsciously building a piece of
technology that you don't wantpeople to use too much?
SPEAKER_01 (14:51):
Yeah.
I mean, for us, like gettingpeople to do things in real life
was everything.
Uh and from day one, like thatwas the idea.
Um we really like it, we lovedthe experience of restaurants.
And the the how you got therewas we wanted to help you get
there, but we wanted you wantedto get you there.
And so, yeah, it was always a adecision we made.
(15:12):
Um, and and for us, it was allabout tracking the right
metrics.
So, like, we don't track timespent on the app.
It's we would maybe track it inthe negative sense.
If we're if we're getting you tospend too long in the app, we're
not doing our jobs and we're notgetting you to a decision point
quickly.
Um but yeah, no, it was very,very important to us.
SPEAKER_03 (15:27):
Did you have you taken outside investment?
SPEAKER_01 (15:30):
We have.
SPEAKER_03 (15:30):
Um and it was that a hard sell to your investors?
SPEAKER_01 (15:35):
I think it may have been if we went day one saying,
hey, like yeah, we don't want,you know, we're not gonna be
measuring time spent, like wedon't want people scrolling all
day, that would have beenimpossible.
So we actually waited until wehad a lot of traction to raise,
and then it was a little bit ofan easier conversation.
Um but if we had done it in theopposite order for sure, it
would have been difficult.
SPEAKER_03 (15:52):
Okay.
What metrics do you track?
SPEAKER_01 (15:55):
Uh we track number of restaurants you rate, uh,
friends on the app, all that thethe the dining, the you social
utility stuff.
Okay.
Okay.
SPEAKER_04 (16:03):
So it is a social app.
SPEAKER_01 (16:05):
Yes.
It's definitely a social app.
SPEAKER_04 (16:07):
Yeah.
SPEAKER_01 (16:07):
But it's just because it's part of the
utility, like we think it theapp needs to be social.
Sure.
SPEAKER_04 (16:11):
Yeah.
SPEAKER_01 (16:12):
Because restaurants are social.
SPEAKER_04 (16:13):
Yeah.
I mean, well, there are a lot ofrestaurant apps that are not
necessarily social, but anyway.
SPEAKER_03 (16:18):
No, restaurants are social and the yes.
SPEAKER_04 (16:20):
Yeah.
Um I mean, I imagine the keything that brings people back is
that your app helps them uncoverthe answer, where should I eat?
And so we had a fun conversationlast night at dinner and and
wine after.
Um talking about therelationship to restaurant
criticism.
(16:40):
And, you know, I think Yelp wasthe first platform that broadly
democratized the idea ofcrowdsourcing for restaurants.
I would I think most peoplewould agree that Yelp doesn't
necessarily surface the toprestaurants with the right
scores.
Google neither, uh, althoughboth of them have such ubiquity.
Um talk about the the role oflike restaurant criticism versus
(17:05):
your friend group.
You know, um, how how willpeople discover new restaurants
or or um you know how does thatinfluence the way that you might
kind of surface the therestaurants I want to go to or
should go to?
SPEAKER_03 (17:18):
Restaurant criticism, you mean
professional?
SPEAKER_04 (17:20):
I am yes, professional restaurant
criticism.
Thank you.
SPEAKER_01 (17:23):
Yeah, so when we started researching the problem,
or like what what were weactually what was the problem
here?
We sent out a survey to a bunchof people, and there were a
bunch of different questions,but one of them was like, well,
who do you like rank all thesedifferent things you could trust
for restaurant recommendations?
And it's like your friends,critics, and everyone just
(17:44):
trusts their friends.
It wins by s by so much.
And even we now see that in thedata where going to a restaurant
page is the the the most commonactivity in the app.
And then the number one thing ona restaurant page is actually
scrolling your friend scores.
People just care what theirfriends think, and so we knew
that was important.
But the critic, there'sdefinitely a place for the
critic view in all of this.
(18:04):
Um we just think it's oneopinion, and you need to you
need to have all the opinionsthat really matter, and you need
to, I mean, what at least we dowith critics is like, okay, oh
yeah, our tastes are aligned.
I I trust them, and like I'mlooking for what they're saying.
Um, but I also want to know whatmy friends think.
I want to know, yeah, what maybeI would like it at with these
personalized recommendations.
Um but we think there is a placefor them, but we think that no
(18:27):
matter how much, even we wantour personalized recommendations
to be amazing, and we think theyare really good and we'll
continue to make them better.
And we think they maybe probablyshould be something you you use
as a cue more than your friendscore.
But people still just care sodeeply about their friends, and
it is social, and we kind oflove that.
Um we think that's okay.
And so yeah, we think that thatthe friend stuff really, really
(18:47):
matters.
SPEAKER_04 (18:48):
Yeah.
It's almost like there's like aRotten Tomatoes type thing where
there's the critic score andthen like the social score that
some combination of and whenthose two things are off is
actually where it gets reallyinteresting, right?
It everyone loves pulp fiction,right?
Like sure.
But um Yeah, I I uh I don'tknow.
I I I think there's a little bitof fear that I have around the
friend group thing, only becausethe what I see and feel in New
(19:12):
York restaurants is places thatget white hot for short periods
of time and then everyone moveson.
And that's what friend groupsmight do, right?
That is friend group behaviorbecause somebody who is the
tastemaker in the group is like,that place isn't cool anymore,
we're going here now.
Um and uh, you know, I I wouldfear as a restaurateur this idea
of not being sustainable over adecade because it's just like a
(19:36):
short-lived thing.
So I don't think that's yourresponsibility to
counterbalance.
SPEAKER_03 (19:41):
Well, but do you see that kind of spike?
Like, do you see the same kindof popularity spike, I would
say, in like a recommendation aswe see like externally like
trending or like because it's ona list on New York magazine, or
or do you it like I'm curious,does interest on Belly tend to
(20:02):
mirror the interest that wewould see if a restaurant wins a
James Beard Award?
SPEAKER_01 (20:07):
Yeah, it it's interesting.
The the data is then really funto watch.
Uh there are people that justget utility out of going to new
places.
And you can see that in the datawhere it's like they just like
it more because it's new andcool.
And I don't know if it's if Ishould tell them they should
like it more, because they'reclearly getting enjoyment from
that.
Um but you do you do sometimessee like scores go like this,
(20:30):
and then it's like like we'relike, hey, let's let's wait a
month and see where it actuallyends up.
Um so you see it a little bit,but there are places too that
like get loads of hype, they'rethey're all over social media
and everywhere, and the scoreson Belly are are not good.
Um and so we see we see bothboth things there.
But another thing that we seethat I think maybe would would
(20:50):
make you feel better is uhpeople start by trusting their
friends a lot, but over timethey actually move to the other
scores.
It's like you that you need tobuild trust.
And so a lot of what we do istry to build trust, and we'll
probably talk more about thatand um how we think about like
how restaurants come into this.
But um but over time, if you canprove to them that what you're
showing is valuable, theyactually start trusting other
(21:13):
scores.
And so there are kind of twothings going on.
But yeah, you do see some peoplethat just love the new place.
SPEAKER_04 (21:19):
Do you share, is there any part of your business
where you're sharing informationwith the restaurants directly?
Because it's almost like youcould be, especially since I
think again, I think the appprobably skews towards the the
younger generation, which youcould argue are more tastemaker
than people that are just goingto the same place that they've
gone for years, you might beable to see ahead of time the
(21:41):
decline of restaurants.
I mean, you know, when when Iwas at Resi, one of the things
that we looked at constantly wason a Wednesday to see what
reservations were available onFriday night in a restaurant.
And that was we had a wholemetric system around this like
was our judge of popularity.
Um and we knew when restaurantsWe're doing bad, literally the
(22:02):
date, like the week it startedto happen, and you know, they
would start to decline fromthere.
So do you see the same thing?
And do you share this withrestaurants at all?
SPEAKER_01 (22:10):
So yes, we the data because Belly's a utility, the
behavior on Belly is totallydifferent than like a Yelp or
Google where one in a hundred,one in a thousand write a single
review.
On Belly, over 80% of ourmonthly active users are rating
restaurants.
And so we will get signals onrestaurants fast, like really,
really fast.
Um and uh like if you if a newrestaurant opens, we look at
(22:31):
this all the time.
Like within a week, we'll have10 to 100x what Google has, what
Yelp has.
Wow.
And so yes, we get it fast.
SPEAKER_04 (22:40):
Um and sorry, your question was uh Do you do you
ever share this information withrestaurants, right?
Are you are you ever the canaryin the coal mine for restaurants
that are starting to declineeven?
SPEAKER_01 (22:49):
Yeah, so to date, no, because really not breaking
down user trust is very, very,very important to us.
And if it looks like we'reworking with restaurants, or we
don't want you ever going inthat being like, was that paid
for, or is that actuallyrecommended to me?
But over time we want to workwith restaurants.
Uh for us when we started Belly,it was supposed to be better for
consumers, but it was alsosupposed to be much better for
restaurants.
We saw these online reviewswhere like over 70% of ratings
(23:11):
on Yelp are one or five.
Like it's just the publicshaming that was like it just
made no sense to us.
And so that's why we don't dopublic reviews and everything's
personalized and all these otherthings.
So it's meant to be way betterfor restaurants, but we're we're
slowly working towards how do wework together, and that will be
something that we build outtogether.
So today, no, we don't do it.
unknown (23:28):
Cool.
SPEAKER_03 (23:29):
The trust piece is really interesting to me in uh
in influencer age.
So I started as a journalistbefore there was a thing called
influencers.
I will date myself.
Um but the influencer economy,as it is called, is extremely
important to restaurants.
And it's been a solid decade, Iwould say, of restaurants like
understanding that this existsand more recently tapping into
(23:53):
this behavior.
However, uh, though there arelaws and rules that govern
influencers and disclosure andpay-for-play and all of the
things that I'm subject to as aprofessional journalist, those
things aren't always followed.
Um and so I guess, do you workwith influencers?
Are there like I'm sure thatthey're on belly.
I'm sure that people who loveand go to restaurants frequently
(24:15):
are on belly.
SPEAKER_01 (24:16):
Yeah.
So we like in terms of how wework with, it's been really
interesting to watch it evolve.
And and and yeah, disclosure hasbeen a tough thing.
Um so like there there is nolike, hey, go to this
restaurant, get free fruit, gowrite it on belly.
That would totally undermineeverything.
And so, no, we don't do that.
Uh the only way we've everworked with influencers, and
(24:36):
we've only ever worked with oneinfluencer, is just like talk
about belly.
Um, and and that's it.
Yeah.
Okay.
SPEAKER_04 (24:42):
Who was that?
SPEAKER_01 (24:44):
Uh which influencer?
Uh, Caitlin Eats.
Oh.
She's amazing.
Follow her.
SPEAKER_04 (24:49):
We love her.
So you're promoting her, she'swould win.
SPEAKER_03 (24:54):
Do you feel like it's a viable strategy going
like would you do more of thatto promote the app itself?
Um social media is important.
SPEAKER_01 (25:01):
Yeah.
So when we started buildingBelly, we actually started
building our social mediaaccounts.
And so Ellie and I do loads ofloads of content.
Uh, to this day, we still runall of our social channels.
And for a very long time, peoplejust thought we were food
influencers because we neversaid anything about the app.
The idea was like build acommunity around food, and then
that community probably isinterested will be interested in
Belly.
And so we think social media isreally, really important, and we
(25:24):
will continue to do a lot of it.
I don't know if we'll do likedirect relationships like that.
Yeah.
SPEAKER_04 (25:29):
So the business side of it is, you know, pretty clear
and and you know I think what'sinteresting about the moment,
RIP is SD, but there are manyother companies right now that
are trying to be the first placethat you go when you want to
find a restaurant.
Most recently, we actually uhhad interviewed Sam at Built
(25:50):
talking about what they're doingand trying to get hyper-local.
Um how do you see this landscapeevolving?
And is this is something likewhat Built is doing?
I think what was the other theballoony one that we were
talking about last night?
Uh um I wrote it.
SPEAKER_02 (26:05):
Corner.
SPEAKER_04 (26:05):
Corner.
This is another thing that Inever heard of, but I
downloaded.
Um, I think a lot of um I thinkthere are a number of ways in
which people are trying to grabyour attention.
And DoorDash themselves.
Like, I mean, like ostensibly,DoorDash wants to be the place
that you go when you try tofigure out any part of your
meal, including groceries.
How do you see the landscapeevolving?
(26:26):
Where do you fit?
Do you feel like there's spacefor you and like a build to
cooperate?
SPEAKER_01 (26:30):
Yeah.
Uh yeah, it's it the landscape'sbeen really interesting and
continues to be very, veryinteresting.
Uh it's never been awinner-take-all market.
Uh, and so I do think there's aworld where everyone can live
together.
Uh for us, how we think about itis we we talk like endlessly
about the dining out journey,and that's like all the steps
included in going out to eat.
And we started at the verybeginning, in the very end of
(26:51):
that journey.
You hear about a restaurant, youwant to remember it, so you
bookmark it for and you went tothe restaurant, want to remember
it, share it with your friends,so you rate it.
And we started there becauseone, we were competing against
the notes, the docs, the excels,none of the tools that were
actually built for restaurantdiscovery.
And two, that's where all thedata sits.
Uh, and so now what we do is wetake all the data and try to do
all the other steps.
And so, yeah, we want to be thestarting point for literally
everything.
Um, but we love working withdifferent players in the space.
(27:14):
Uh today we're we're integratedwith Open Table, we're
integrated with Seven Rooms,there are a lot of partnerships
coming.
Uh, and so I do think there's athere's room for everyone.
Um it will, I mean, that's somany thoughts, but how we
collect data in Belly is verydifferent than like how a review
site collects data.
Like you say if you likesomething better or worse than
other places you've been to, andyou get this ranked list.
(27:38):
Uh and we did that becausebefore, and I'm gonna I'm gonna
age myself, but before we whenwe were starting Belly, it was
before AI and ML were supercool.
And we knew that reviews werebroken, and so we started
analyzing like why are theybroken?
And we we took the five-stardata set and we're like, can you
predict what people wouldactually like?
And the answer was no, youliterally cannot.
Um and so we're like, okay, whatkind of data do you need to
(28:01):
collect in order to predict whatpeople would like?
And it was the ranked data, andit's used everywhere in
academia, but for some reason itwas not used in consumer apps.
Uh, and so that's why we collectdata the way we do.
And so we, in this going back toyour opening remarks, like we're
big believers in garbage in,garbage out.
And so for us, it was like, hey,like let's collect the right
type of data, and then we can doreally, really, really good
(28:22):
recommendations and really,really, really good explore,
make really good explore tools,but like we're gonna earn the
right to do that and throughcollecting the right type of
data and then having a lot of itand then doing it.
And so we'll be doing a lot inExplore.
It's a big priority for us thisyear.
Um, but we think you know, thereare there's room for a lot of
people.
The market's big.
SPEAKER_04 (28:40):
Yeah.
Layer in AI to that, because youyou and I uh were chatting a
little bit about how the workthat you will do with, you know,
I don't know, maybe it maybe itwas anthropic, I don't know what
tools you were using, but um,will lead to this world where
you can more easily chat perhapswith this with your app to find
(29:02):
where you want to go.
I mean, say more about that tothe extent that you want to.
SPEAKER_01 (29:06):
Yeah, I mean, we're still early, so I think like AI,
like as a team, we're reallylean and we run really lean, and
that's important to us.
So AI has been amazing in inproviding leverage on the team
layer.
But in terms of like how weincorporate it into the product,
um there'll be a lot of tests tosee what actually is
interesting.
I I don't I think we haveamazing data and our
(29:27):
recommendations to date, likethe the personalized
recommendations we do in the appare great, and we can actually
measure that.
We like how predictive are theyis something we can we can
measure.
Uh we do think though, likethere is a world where like if
you have you know some sort oftext search uh that gets you to
do things in real life uhquicker.
And like that's what we careabout.
And so uh if anything's alignedwith our mission, then like
(29:49):
we're down to try it and thensee how our users react and and
uh if it's actually helpful, andthen we'll we'll we'll iterate
on that.
But that's as far, like that'swhat I know we'll start with,
and I don't know where we'll go.
unknown (30:00):
Cool.
SPEAKER_03 (30:00):
How are you using you glossed over this, but I
want to go back to it.
How are is your team using AI atBelly right now?
SPEAKER_01 (30:08):
Yeah, so uh every single person now on our team
codes because now even if youweren't a developer, you can
vibe code, and and that's beenreally, really awesome.
It's been a big unlock.
Um but it's mostly through thecoding.
Like cloud code has been great.
SPEAKER_03 (30:24):
Uh do you have an example of some of like someone
who wasn't necessarily a coderthat was able to unlock some
great insight or functionality?
SPEAKER_01 (30:33):
I mean, it's like our designers now, they they
actually build their features.
They they they do the research,they, they, they do the mocks,
and then they build them.
Like everything that we um yeah,it's it's kind of it's crazy.
I mean, we we launched, I'mtrying to everything we launched
in the app now is built bythere's a piece of it that's
built by someone who literally,I don't know, a couple months
(30:56):
ago had never touched code ever.
I even our ops analyst uh is shedoes like a like she puts
together together a bunch of ourlists for us, and she was really
dependent on the developers topull data, and now she does all
of that herself.
It's been yeah, reallyincredible.
SPEAKER_03 (31:11):
So it's kind of yeah, anyone with a good idea
and a good understanding canbuild the technology.
Build is even like maybe thewrong word, can like create the
technology that supports thatspecific uh action or idea.
SPEAKER_04 (31:24):
We do the same, obviously, at byte and for you
know folks in the room, to theextent that you are not talking
to Claude as an example, youchoose your own about what you
want to do and then thechallenges you'll face and how
to make Claude work for you.
Like it is incredible that thisthing comes with a manual about
(31:45):
itself.
And so, you know, I have folkson our team that are in customer
success who are creatingdashboards about customer
success because and they'venever coded before.
Shout out Land and Ali, you guysare great, um, amongst others.
And when they come up amongstanswers, like they're not asking
like how do I do this, they'resaying things like if I was to
act like an engineer and want tobuild a dashboard, what are all
(32:08):
the things that I wouldconsider?
And can you write rules aroundthat?
Because I don't know anythingabout coding.
And so like it's like thatsecondary layer.
Anyway, advice you're gonna hearstuff like this all day.
I am probably the least uh ofthe AI um expert panelists of
today.
Talk to that guy.
So, okay, so back to it on thebelly side.
(32:28):
Um where what's the if youcreate a great social app, let's
just call it, with utility, umsometimes these things have life
cycles, but also similar to howum you know Instagram's not just
about sharing pictures anymore.
Yeah, um, where does it go?
What's five years from now?
I mean, aren't you vertical AIand isn't it the SAS apocalypse
(32:50):
and can't, you know, uh verticalSAS, I mean, in it in some way,
like your consumer app, but likesomebody could vibe code this,
right, in like a weekend.
SPEAKER_01 (32:59):
Yeah, so I mean, uh yeah, there are lots of thoughts
uh again.
Um uh so there are real networkeffects here.
And so the the the bigger youget, the the better the moat.
Uh and so we have both data andpeople moats now.
Uh and so it's really hard tojust build this and be like,
hey, you know, there arethousands of of things like that
(33:23):
out there.
And so uh continuing to getbigger is very important to us.
Uh uh, but in terms of like itreally goes back to that dining
out journey I was talking aboutand all the different steps, and
like the big chunk right now islike we don't think we do
explore that well.
Uh and so that will be a bigfocus this year and making it so
you can go in the app and belike, you know, whatever, maybe
(33:43):
like New York reservations arecrazy.
What's the best reservation Ican get with Elliott tonight?
And it will take both of thetaste and get you to do things
and uh get you out to the bestreservation that you can get at
that point in time.
Um so Explorer will be a bigthing.
Um, but then the other twocategories we think about in
terms of growth are geographicand and and uh categorical.
And so to date, yeah, we haveover 100 million ratings, we're
(34:05):
in over 50,000 cities, uh overtwo million restaurants rated.
Uh and so uh it's been reallycool to watch the data on when
we started, everyone was like,New York is so saturated, you'll
never get New York.
And then we we we kind of didit, and then people are like,
well, New York's different thaneverywhere else.
And uh and uh it's it's not uhcrazy thing we found out, and
(34:27):
everyone in the room in thisroom knows is that everyone eats
uh and and people eat uh youknow very differently, but like
that behavior is consistent.
And so like when you look at thedata, like it doesn't really
matter where you are, like thebehavior is very similar.
Uh, and that's been really,really cool.
So we'll be leaning, uh youknow, we want to be the the
largest repository of preferencedata in the world.
(34:48):
And so that's what what we'll bebuilding towards.
Uh and we really want you to bedropped anywhere in the world
and have your perfect day,because we have all the opinions
that matter yours, your friends,these personalized
recommendations.
And then in terms of categoricalgrowth, restaurant market is
massive.
And so we'll be here for awhile.
But I don't know, maybe hotelsand things like that that are
aligned with hospitality will bein our future.
(35:09):
Um, but really it's like let'sdo all the steps in the dining
out journey and let's beeverywhere.
unknown (35:13):
Yeah.
SPEAKER_03 (35:14):
Um I was in London uh a year ago, and I was talking
to an industry friend.
We were taking a walk, and I wasnot yet a belly user, and he
stopped, like stopped dead andlooked at me and he was like,
What is wrong with you?
This is the best possible appfor restaurant discovery.
It's so much better thananything out there.
And something that I ask peoplewho dabble in best, and you
(35:37):
referenced it a couple of times,your best restaurant, your best
night out.
How are you, how are youralgorithms, how is belly
defining best for an individualuser?
Because that is the mostsubjective uh superlative.
SPEAKER_04 (35:49):
That's a great question.
SPEAKER_01 (35:50):
No, well, I love your friend.
Your friend sounds amazing.
Uh we have best, yeah, we agreewith you.
It's a within belly, it's almostlike a bad word uh because there
is no best, and we believe thatto our cores.
And and but then you know, weput out loads of stuff on social
(36:11):
media, it's like top-rated, andthen we're afraid that that
might come off as best, andthat's actually not what it's
meant to be.
But I'll talk about somethingelse in a second.
But uh yeah, no, we we thinkbest is what it's is best for
you, and and that is verydifferent than what's best for
me.
Uh, and that's okay.
People should not be uh feel badabout their tastes.
Like it can be different, it isvery personal, it is subjective,
(36:34):
and so yeah, we it's truly very,very personal.
Uh that said, the other thing wethink a lot about is I think we
were talking about this lastnight, is uh the idea of like an
echo chamber.
And like if you're only you'reyou're taking someone's taste
today and just giving them thesame thing over and over and
over again, how do they grow?
And so the explore like the thethe different lists.
(36:54):
I don't know if your friendshelp you or hurt you, um, but
really pushing, and that's whatBelly today one thing it does
really well, but also maybe tothe detriment of uh to the uh to
restaurants, but like we get youto go to new places because
Belly is very gamified, and youget only get credit for like how
long your list is and and yourstreaks and all that.
(37:16):
So we're really pushing you toexplore, and so we think that's
how we today are defeating kindof the echo chamber nature of
this, but also loyalty and andgoing you know going back to
restaurants is really important.
So we're trying to figure outhow do you take all those things
and do them well.
Uh so yes, best is personal, butalso we want to be careful about
being an echo chamber.
SPEAKER_03 (37:35):
Yeah.
Uh I for example, I was in um Iwas in Philadelphia, I was in
Philadelphia speaking at aconference a couple weeks ago,
and uh I went to a lot oftop-rated places that were
wonderful.
Um, but also I am maybesurprised to hear a real
introvert.
And uh by the end of the twodays of just constant contact
(37:56):
with humans, I just wanted tosit somewhere quiet in a corner.
And so best for me in thatmoment was this like tiny ring,
it was an El and Yin restaurant.
So it was still a spectacularplace, but it was quiet, it was
empty, it was like kind of awayfrom everything else that was
happening.
And I needed thatrecommendation.
I happened to be in the roomwith her and I went and asked
this restaurant where I shouldgo, and she said, go to the sit
(38:18):
at the end of my bar, we'll saveyou a seat.
But I guess what I'm what I'mtrying to say is uh is Belly
flexible like that?
Like, could is there a best foryou know Kristen on a Sunday
night who doesn't want to talkto anybody?
And then versus you know,Kristen on a Thursday night that
is like ready to go.
SPEAKER_01 (38:33):
No, you're 100% correct, it's definitely
situational as well.
Yes, it should be.
You should be able to be like,hey, like this is where I'm at
today.
What should I do?
Um and I wherever we're notdoing that well today, like we
we think we can and will.
Uh-huh.
Yeah.
SPEAKER_04 (38:48):
Final question before we Are you taking feature
requests?
SPEAKER_01 (38:51):
Always.
SPEAKER_04 (38:52):
Okay.
Elliot.
SPEAKER_01 (38:53):
A little nervous to say that again, but I got the
mic.
SPEAKER_04 (38:57):
I know.
Um I don't and look, maybe thisis even possible already because
I'm I'm not I'm not that deepyet.
Um but I feel like I would wantto kind of stack rank my friends
in who I would trust theirrecommendations and who I
wouldn't.
Some of them on the negativeside of the curve.
(39:18):
But look, I mean, you know, Ihave been involved in
restaurants and restaurant techfor a very long time, especially
in New York.
And like, yeah, Colin Comic hasit dialed in as to where to go.
Whereas, you know, love you,Bobby Sitsema, but like I'm not
gonna go to the plate, the Yemenrestaurant in Outer Queens to
eat pig's head.
(39:38):
Like, it's just not gonnahappen, right?
So um is there is that is thatis that something you've ever
thought about where you can uhself-direct to where which
friends might have moreinfluence on your
recommendations?
SPEAKER_01 (39:50):
Yeah, so uh may be surprising, maybe not.
You're not the first person toask for that.
Uh we do get we do get emailsand people are so crazy.
SPEAKER_04 (39:58):
Mike, we're gonna edit that part.
SPEAKER_01 (39:59):
Yeah.
People are so crazy aboutranking.
They're like, I want to rank myfriends, and I want, you know,
some of my friends have terribletaste.
And and you know, that's that'sfine.
SPEAKER_04 (40:10):
Um, the AI would recommend to drop your friends.
SPEAKER_01 (40:13):
Yeah.
SPEAKER_04 (40:14):
You can't have friends with bad restaurant
tastes.
SPEAKER_01 (40:16):
But what what we do today is like you can see how
correlated you are withdifferent people.
And so we actually show you likeyour match score.
And so it we it hopefully paintsthe picture of like one, it's
okay to disagree, but also likeyou you may trust or may not
trust that person, but here'show similar you are.
Uh so yes, uh there will be morethings.
We think of explore as places,but also people.
SPEAKER_04 (40:35):
And then you gotta check if they're both single.
SPEAKER_03 (40:37):
I was just gonna say that.
It was like that.
Yeah, you do like a matchmakingservice.
SPEAKER_04 (40:43):
Yeah.
Chipotle bowls.
Oh no, it's cava bowls.
My no, my buddy's doing this.
He's doing the Kava series wherehe puts two people on a blind
date.
Uh shout out Jacob.
Um, with that No, we wanna yeah,we wanna Yeah, we want to go to
questions.
It's yeah, we have a perfect,perfect amount about 17 minutes
for questions.
SPEAKER_00 (41:00):
Uh a huge thank you to Kristen Hawley and Brandon
Barton of the Simmer forbringing this conversation to
life and for being such agenerous collaborator.
If you're not already followingthe Simmer, we highly encourage
you to do so.
You can find it wherever youlisten to podcasts, and be sure
to check out their YouTubechannel the full video
(41:21):
experience to keep up with ourshow and subscribe to
Distinguished wherever you getyour podcast and explore more
about experience innovation andour programs at w.edu slash
hospitality.
You can also reach me directlyat jardn at w.edu.
That is s h.edu.
(41:43):
I'm a w hospitalityadministration.
Thank you for listening and seeyou next time.
Popular Podcasts
Betrayal Weekly
Betrayal Weekly is back for a new season. Every Thursday, Betrayal Weekly shares first-hand accounts of broken trust, shocking deceptions, and the trail of destruction they leave behind. Hosted by Andrea Gunning, this weekly ongoing series digs into real-life stories of betrayal and the aftermath. From stories of double lives to dark discoveries, these are cautionary tales and accounts of resilience against all odds. From the producers of the critically acclaimed Betrayal series, Betrayal Weekly drops new episodes every Thursday. If you would like to share your story, you can reach out to the Betrayal Team by emailing them at betrayalpod@gmail.com and follow us on Instagram at @betrayalpod and @glasspodcasts. Please join our Substack for additional exclusive content, curated book recommendations, and community discussions. Sign up FREE by clicking this link Beyond Betrayal Substack. Join our community dedicated to truth, resilience, and healing. Your voice matters! Be a part of our Betrayal journey on Substack.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by Audiochuck Media Company.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.