January 1, 2025 • 22 mins
What if AI could revolutionize your sales strategy and open doors to opportunities you never knew existed? In our latest episode, we promise insights that will transform your approach to sales by leveraging the power of AI tools like ChatGPT. We uncover the secrets behind crafting precise prompts, a crucial skill that could make or break your AI-generated content. From maintaining the perfect tone to targeting specific audiences, our discussion emphasizes the importance of context in achieving optimal outcomes. Plus, we explore how visuals and SEO can amplify your AI outputs, ensuring you stay ahead of the competition in the sales world.
But that's not all—imagine having a custom GPT tailored just for your sales needs. We dive into the exciting world of building these personalized models, which can transform your sales training and management processes. Our conversation covers everything from uploading robust content to creating secure systems, ensuring that your business remains privacy-conscious. We also share guidance on effective sales strategy development, discussing the importance of understanding market needs and leveraging competitive insights. Whether you're a seasoned sales professional or just starting, this episode is packed with valuable takeaways and practical advice. Join us for an engaging session that not only offers strategies but also invites you to participate in upcoming workshops designed to enhance your mastery of AI in sales.
Join the AI-Powered Seller Newsletter: https://bit.ly/ai-powered-seller-newsletter
Sign up for Custom GPTs: https://link.meetjourney.ai/aff_ad?campaign_id=4859&aff_id=2751&source=Jake-KD-Podcast&hostNameId=24435
Connect with Jake: https://www.linkedin.com/in/jakedunlap/
Connect with KD: https://www.linkedin.com/in/kddorsey3/
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:05):
The reason we've been doing these sessions.
Ai and the generative AIprocess is one of the first new
things I have seen in sales intech really in a long time.
Almost all products that havebeen coming out over the last
six to seven years not many ofthem are new ideas.
They're new ways or better waysof doing something that was
(00:27):
already in place, whereas AI isa whole brand new field to play
in.
It's actually different.
It's creating opportunitiesthat did not exist before, and
so when this all startedhappening, I dove in head.
First.
I sat down and mapped out.
I was like what is it thatmakes me good as a leader?
(00:49):
What is it that I do that Ithink leads to some of the
results?
And I mapped it all out on apage, and then I wrote down what
I thought AI could do now, whatAI could do soon and what I
thought AI would be able toeventually do.
And when I wrote all of thatout, I was like, wow, it's going
to be able to do a lot of this.
And so I had this little phrasethat I used a lot of last year,
(01:11):
which is if anyone's going toreplace me, it's going to be me.
That's where I've beenoperating from is if anyone's
going to replace me, it's goingto be me, and so I want to learn
everything I can about AI, thedifferent use cases and
everything around it.
So today we're going to diveinto one just some of the key
learnings and, I'll say, maybe,some of the mistakes I've made
over the past year trying tolearn it, but then also some of
(01:34):
the use cases, some of theprompts and some of the ideas
that I see happening now.
The prompt is everything.
Prompt is everything.
The prompt is everything.
And when I see people trying AIfor the first time, trying
ChatGPT for the first time,where they'll do a prompt right
and it looks like this, whereit's like, hey, write me a blog
(01:59):
article for a sales leadershipcourse, and that's what they
prompt, right, and this is whatit comes back Transform your
sales game, join our premiersales leadership course.
In this fast-paced world ofsales, staying ahead of the game
is not just an advantage, it'sa necessity.
(02:20):
How's this feel so far?
How is this blog articlefeeling so far?
To everybody?
We say, hey, we want to do ablog article.
It's vague, it's boring.
Here's what you'll learnstrategic sales planning, team
building, motivation, highschool essay.
Right, because there we go.
(02:41):
The prompt is this write me ablog article for a sales
leadership course that's goingto give you results like this,
and early, early on.
Actually, I could even go backto some of my early, like you
know sessions in this one oflike that's what I would put in
there, like, hey, I need a coldcall script or help me handle
(03:03):
this objection.
The prompt is everything right,so things that you want to
include in the prompt.
Actually, I'll put you on thespot here real quick.
Okay, what would make thisprompt better?
Right, let's use this as anexample.
Again, I want a blog articlefor a sales leadership program.
(03:25):
But now here's the key.
When writing the prompt, firstthing we need to know is what
tone?
What tone do I want this in?
I want the tone to be casual nojargon, no fluff words, a human
(03:46):
tone.
Drew nailed it.
The ICP, for this is first timemanagers who maybe haven't had
much coaching in the past.
Okay, so we've got who we'rewriting it to.
What we want the tone to be inhere.
But then also, this is wherethe human element comes in is
(04:11):
you need to give it some ideas,right, reps, goals, connecting
(04:32):
that to their quota and mappingbackwards to make a plan.
Please ensure to have anactionable next step included.
Okay, now is this prompt alittle bit better, right, so we
can see what it'll come up withhere.
But it's this level ofgranularity you need to be going
into with your prompts.
Too often I see people havingbad prompts.
(04:55):
They go through it but theydon't build on it right.
So now the title's alreadybetter.
Let's make it simple andeffective.
Hey, there we're more casual.
Stepping in.
It starts to sound like a humanright, understanding your
team's goals.
It's doing a pretty good jobhere.
Linking goals to quotas isdoing a pretty good job here.
(05:16):
So the point I'm making on thisfirst one is the prompt is
everything right?
Like you want to use it more inmarketing use cases, you have
to feed it.
What you know about goodmarketing which is actually my
next biggest learning in AI allof last year is context is
everything.
Ai by itself can create content.
(05:38):
If you want it to be good, youhave to give it context, right.
So, actually, what would besomething else that makes a good
blog article?
What's something that wouldmake a good blog article right?
So it's already writing thisout.
But let's think about nowwhat's some further context?
What are things that we know gointo a great blog article that
(06:02):
we can give it some contextaround?
Ooh, okay.
So now diagrams, visuals Okay,this is great.
Could you make me aninfographic that covers these
main points?
Might have to flip over tomid-journey for it, but right.
(06:27):
So now we did the article, nowwe're asking for an infographic,
and then I could think ofanother example of like, and we
want to make it SEO optimizedfor blank.
That's what I mean in terms ofthe context you need to give it.
So, similarly, if you're tryingto go for a closing, I need
(06:47):
help prepping for a closing callyou have to give it the context
behind what good looks like.
The people that are going towin with you know, chatgpt and
everyone else are the ones thatcan give it the best context of
what they are looking for.
So now, look at this.
Now we just got an infographicand notice, I would like put my
(07:08):
own typo in there sales,leadership, rogals,
understanding your teals.
But look at this right, howlong did this take?
Now we have an infographic thatpeople could download, and now
we'd have to give itinstructions to fix it.
But now we went from nothing toa blog article, to an
infographic, and then we couldtell it to SEO, optimized for
(07:29):
these keywords, and it would dothat.
So this is how the prompt iseverything.
And then it's giving it thecontext right what does good
look like?
So it knows how to operate fromit.
So something I will use to allof my sales leaders, marketing
leaders, revops leaders startrecording everything.
(07:54):
When you are leading a training,record it.
When you are givinginstructions to somebody in
terms of how they can improvesomething, record it.
Revops, if you are reviewing adashboard, record yourself doing
that, because that's thecontext that you can feed into
ChatGPT to give it the insightsthat it needs.
(08:16):
So actually, I'll show you allthis real quick as an example.
This right here, this is overlike 15 hours of recorded
trainings that I'm leading atraining with my team.
I record it because I can feedit in.
(08:45):
It's that transfer.
That's the context, becausewhen you're teaching something,
you're explaining it.
Explaining it to chat GPT isthe key to making the results
better.
So the first big learning wasprompt is everything.
The second is context.
You need to tell it what goodlooks like.
You need to give the humanelement that you need to have in
(09:07):
there.
That's what makes it better.
Can you create trainingmaterials using that method?
Absolutely, so.
This is actually one Iliterally just did yesterday.
Okay, so I need you to help mecreate some lesson plans I have
done for other trainers todeliver.
Right, this is a training thatI want my own enablement team to
be able to deliver to mycurrent team, right?
(09:29):
So look at this y'all.
So I uploaded, right, thedocument.
This is the training, right?
This is the training that Idelivered.
I need your help to create thesesessions.
I need to outline these, thekey topics, with sub points
under them, with example,engagement questions to help ask
the students to apply.
I consult SAS companies.
I'll upload the transcriptionof the training.
So you have it.
I want them to be very thorough.
(09:51):
So this is the first thing andit spits out.
So it gives the sub points, theengagement questions of each
part of this training.
So far, so good.
I go further, right?
I'd like you to make the subpoints a bit more detailed.
Four to five points there foreach prompts, plus let's add in
an engagement question andexercise questions, meaning
(10:13):
questions that tasks.
So, danny, this is againfurther context.
Here's what I want you to doNow look what it comes back with
Sub points engagement,questions, exercise for each one
of the key portions of this andthen looking good.
Now the sub points.
I want the actual prompts thatit's talking about.
(10:36):
This took me 20 minutes to builda full training guide for one
of my trainings, completelymapped out, that then I can use
for my teams, for my managers,for my team leads, to be able to
actually apply it.
This is why, if you haverecordings, I actually did this
with one of my coaching clientsa few weeks ago with he had
(10:57):
actually what he did foronboarding Right, he uploaded a
persona.
Right, he uploaded a persona.
Right, he uploaded a persona.
And then we said, great, canyou write me a quiz to certify
my reps on?
Please include a mix ofmultiple choice questions, fill
in the blank questions and yesand no.
We're looking for an 80 to 90%pass rate, so don't make the
(11:20):
questions too hard, but weshould be able to pass this and
in five minutes, add a complete10 question quiz set up for one
of the enablement modules thatthen he can use with his team.
That's another big applicationthat we can use here, right?
So this is an example of atraining content that you can do
(11:40):
, and I don't know for any ofy'all, but, like writing,
writing this stuff is not great.
I'm a speaker, so I'll recordmyself giving the training.
I use a tool called Descript todo the transcriptions.
I know there's a lot of toolsout there to do transcriptions,
but I record my trainings, Iupload them into Descript and
then go from there.
(12:01):
Now I've got the transcriptionsand I can use it from there.
But my enablement team has usedthis for checklists, to map out
coaching plans, all of it, andit's great for creating quizzes
and things like that as well.
Powerpoints it seems tostruggle with a little bit more
than like PDFs and docs.
You still absolutely can, butit's a little bit better when
(12:24):
it's in a doc.
But, yes, you can upload it andhave it, do a quiz right then
and there for you, so it canwork in a big way.
So now, flipping kind of thetopic here, right?
So we prompt this king, give itthe context and I don't know if
y'all are catching this, but doyou notice whether this is
right or wrong how I talk to it?
(12:45):
How would you describe how Italk to it Right, so it did the
thing.
How would you describe myresponse back to it Right, it's
like, ok, it's starting to lookgood.
Now, in the sub points Iactually do talk to it like a
human, right.
I'm saying, hey, I liked this,so I'm training it like we're
(13:07):
close.
Keep doing more of that,because also learning lesson If
you don't this is so weird.
If you don't tell it you likedit, it changes the whole thing
sometimes, right, where if Ijust said, oh, I want more
bullet points, it changes thewhole thing sometimes, right,
where if I just said, oh, I wantmore bullet points, it changes
all of it versus hey, I likethis, we're looking good, I'm
(13:27):
coaching it along the way thesame way.
Now I want a short paragraphexplaining the bullet points,
how to speak to it, why thebullet point matters.
I want the instructor's guideto be almost like a scorecard
and script.
Talk to it like you would, ahuman.
This people.
It's weird Like they put ontheir prompt hat too and it gets
super, super complicated to doEngage with it, to get it to
(13:51):
where you want it to be.
Where I'm most excited is whatJake and I've been working on is
building of the custom GPTs.
Right, because each one ofthese is also right to go
through it.
Like it doesn't remember allthese conversations, whereas if
I go into the Bipsy bot that Ihave built here right, and come
(14:13):
into this all those you knowtrainings I showed you that's
already uploaded and built intothis bot, so it always knows it,
it always has the context to it, it always has the frameworks
already built in.
Is all of this content is inhere right?
The ability to build customGPTs.
(14:33):
You have to have the context todo it.
Look at all this context thatis built in here the sales
training, the coaching,everything else and then how I
want it to behave right.
Here's how I want you to do it.
I want you to act like a doctordoing a diagnosis, asking good
questions, ask clarifyingquestions.
Ideally, you can lead theperson to the answer through
(14:54):
questions.
Even better, give choices.
You map all of this out.
So now right, when you'reactually in here.
So again, it's like actualcontext, if I can say have a rep
that is losing deals at the endbecause whenever they ask for a
(15:17):
deposit, the prospect declinesand never comes back.
Create a training program withprompts and tips on how to
generate more deposits.
One thing we know works is byjustifying the deposit because
(15:47):
of how much upfront work andvalue we deliver.
Because of how much upfrontwork and value we deliver.
So now, because it has so muchof my sales, training, my
management, everything elseright, it can take all of that
and work with it.
Okay, so it can help make out atraining program right.
Enhance enough skill, secureand deposit duration.
(16:08):
This will be over two weeks,week one understand the value
and building confidence.
A training session focused onunderstanding the value activity
reps list out all the possiblevalues and benefits that justify
the deposit.
Right.
Crafting the justification.
Development like look at thisy'all, look at this.
Okay, this is big right Around.
(16:29):
All of this is and I don't evenhave a specific training on
deposits.
But can you see what this justdid?
Right, feedback sessionspersonalized feedback activity
reps create and respond to these.
Like this right, because it hasall the training and how to be
a good manager.
It can do this right.
(16:50):
So this is why I'm encouragingyou all to record everything.
Right Is because you can do itthis way.
So uploading it here does notgive everybody on ChatGPT access
to my documents.
That's not how it works.
The flip side, though, is alsotrue.
This is what Jake and I've beenbuilding is we've also created
a private instance for ourselvesand for companies to you.
(17:13):
That's completely contained.
So it's not on chat GPT, it'snot out there anywhere else.
It's a privately containedsystem that does the same things
with our content, the company'scontent and everything else.
So if you're afraid of privacy,you do need to build it in
internally.
If you want some ideas on how Ireach out to us, because we've
done it, we've actually built itand it's really starting to
(17:34):
work very well.
But, like, this is an exampleof what you can do, and I have
one for call scorecards too howto build a custom call scorecard
.
The custom GPTs are all of it,right Of all of it is like
taking the context, but if youdon't have context, it's going
to be very hard to build thecustom GPTs.
(17:55):
Even if you have the playbooks,you need to have the best
practices built out as well.
So the custom GPTs is one ofthe things I'm most excited
about.
We've built the manager one, webuilt a scorecard one, we built
a research one, a prep one, andmessaging the one that Jake
built out on creating snippetsto put into emails and LinkedIn
(18:16):
is just gold, right.
It's like chunking it down.
Which is my biggest tip I wantto give everyone is think big,
then think small.
The GPTs get better the morespecific the action you're
trying to get them to do is.
So in Bipsy, for example, Ihave my scorecard training in
(18:38):
there, but when I ask it tobuild a scorecard, it's not
nearly as good as the salesscorecard builder, because in
here I can give it even morespecific instructions.
I have the example scorecards.
I have the specificinstructions on how to do it.
So think big about what youwant it to do.
Then get small, build the GPTsfor it.
(19:00):
And I cannot express enoughy'all.
If you want help with this, werun workshops on this.
We can show you how to do this.
We have a tool that works wellfor this.
We have a whole prompt guide,right.
Like I don't know if manypeople have seen this.
We've talked about it a lot,but like, when we talk about the
prompt guide, like this is whatwe mean on the prompt guide,
(19:21):
and we have prompts for research, prompts for messaging, prompts
for closing deals, like it'sall in here already to get the
ball rolling for you.
That is the key on it, rightit's again, it's a great idea of
big to small when it comes tohow to build out these custom
(19:41):
GPTs, but you have to have thecontent and best practices to
make it work right.
So I was moving fast on thisone, but I hope this gave us
some ideas here.
One the prompt is everythingokay.
Again, if you want help, go getthe prompt guide.
It's like 50 bucks, y'all, it's$50.
(20:02):
Go buy it because it's going totrigger right.
I've literally I've spent wellover $500 on all other prompt
guides out there to learn itright.
Because you start to see thepatterns of good prompts and go,
ah, this is how it applies tomy world, those are the ways.
So scoop the prompt guide upy'all.
It'll just jumpstart yourprocess in learning this.
(20:25):
Okay, how do you know whenyou've done enough refinement
requests?
How many times Ooh, okay.
So I my my personal sweet spottends to be like four to five,
because if I'm giving it adifferent, I was like Ooh, I
liked this, but I didn't.
Similarly, I liked this, Ididn't like this part.
(20:46):
This part sounded too salesy.
Can you give me another examplewhere, instead of being salesy,
it's more human.
It doesn't include X, y, z.
If I give it four and five backand forth and I'm still not
getting what I want, I'll starta new channel and start with the
last prompt that I gave thatwas better and work from there,
(21:08):
because sometimes if you stay inthe same one, it falls.
It just doesn't change faster.
So I'd recommend four to five,and if it's still not getting
you to where you want it to be,then I would, I would say, start
a new chat to go through, andthat tends to get it back on
track.
One of the hardest parts ofearly go to market is not
(21:29):
knowing, maybe, what mattersright to the market, what
matters to what's going on outthere in the industry, what are
the trends, what are theinsights?
What's the language that yourpersonas like use?
Right, here's actually onestraight from the prompt guide.
Right, talk about likecompetitive insights, right of
like.
Hey, I came across thiscompetitor.
Why do you think people buythat?
(21:50):
How?
How would you compare it toours?
What are some keydifferentiators and how can we
leverage that information toclose more deals?
Right, so that's straight fromthe prompt-like guide.
So there's a lot of places wecan do that, but more than
anything, I'm just trying tostay on top of, like the prompt
engineering, the insights, andthen how can it connect the dots
(22:10):
?
That's one of my personalfavorite tactics within.
It is okay, told me this, nowconnect the dots to this, and
then it does such a great job ofdoing that.
Hopefully, you saw how fastthis works.
I appreciate y'all, as always.
If there's follow-up questions,hit me up, hit Jake up, but as
(22:34):
always, y'all deuces, we're out.
Appreciate you, thank you.
The reason we've been doing these sessions.
Ai and the generative AIprocess is one of the first new
things I have seen in sales intech really in a long time.
Almost all products that havebeen coming out over the last
six to seven years not many ofthem are new ideas.
They're new ways or better waysof doing something that was
(00:27):
already in place, whereas AI isa whole brand new field to play
in.
It's actually different.
It's creating opportunitiesthat did not exist before, and
so when this all startedhappening, I dove in head.
First.
I sat down and mapped out.
I was like what is it thatmakes me good as a leader?
(00:49):
What is it that I do that Ithink leads to some of the
results?
And I mapped it all out on apage, and then I wrote down what
I thought AI could do now, whatAI could do soon and what I
thought AI would be able toeventually do.
And when I wrote all of thatout, I was like, wow, it's going
to be able to do a lot of this.
And so I had this little phrasethat I used a lot of last year,
(01:11):
which is if anyone's going toreplace me, it's going to be me.
That's where I've beenoperating from is if anyone's
going to replace me, it's goingto be me, and so I want to learn
everything I can about AI, thedifferent use cases and
everything around it.
So today we're going to diveinto one just some of the key
learnings and, I'll say, maybe,some of the mistakes I've made
over the past year trying tolearn it, but then also some of
(01:34):
the use cases, some of theprompts and some of the ideas
that I see happening now.
The prompt is everything.
Prompt is everything.
The prompt is everything.
And when I see people trying AIfor the first time, trying
ChatGPT for the first time,where they'll do a prompt right
and it looks like this, whereit's like, hey, write me a blog
(01:59):
article for a sales leadershipcourse, and that's what they
prompt, right, and this is whatit comes back Transform your
sales game, join our premiersales leadership course.
In this fast-paced world ofsales, staying ahead of the game
is not just an advantage, it'sa necessity.
(02:20):
How's this feel so far?
How is this blog articlefeeling so far?
To everybody?
We say, hey, we want to do ablog article.
It's vague, it's boring.
Here's what you'll learnstrategic sales planning, team
building, motivation, highschool essay.
Right, because there we go.
(02:41):
The prompt is this write me ablog article for a sales
leadership course that's goingto give you results like this,
and early, early on.
Actually, I could even go backto some of my early, like you
know sessions in this one oflike that's what I would put in
there, like, hey, I need a coldcall script or help me handle
(03:03):
this objection.
The prompt is everything right,so things that you want to
include in the prompt.
Actually, I'll put you on thespot here real quick.
Okay, what would make thisprompt better?
Right, let's use this as anexample.
Again, I want a blog articlefor a sales leadership program.
(03:25):
But now here's the key.
When writing the prompt, firstthing we need to know is what
tone?
What tone do I want this in?
I want the tone to be casual nojargon, no fluff words, a human
(03:46):
tone.
Drew nailed it.
The ICP, for this is first timemanagers who maybe haven't had
much coaching in the past.
Okay, so we've got who we'rewriting it to.
What we want the tone to be inhere.
But then also, this is wherethe human element comes in is
(04:11):
you need to give it some ideas,right, reps, goals, connecting
(04:32):
that to their quota and mappingbackwards to make a plan.
Please ensure to have anactionable next step included.
Okay, now is this prompt alittle bit better, right, so we
can see what it'll come up withhere.
But it's this level ofgranularity you need to be going
into with your prompts.
Too often I see people havingbad prompts.
(04:55):
They go through it but theydon't build on it right.
So now the title's alreadybetter.
Let's make it simple andeffective.
Hey, there we're more casual.
Stepping in.
It starts to sound like a humanright, understanding your
team's goals.
It's doing a pretty good jobhere.
Linking goals to quotas isdoing a pretty good job here.
(05:16):
So the point I'm making on thisfirst one is the prompt is
everything right?
Like you want to use it more inmarketing use cases, you have
to feed it.
What you know about goodmarketing which is actually my
next biggest learning in AI allof last year is context is
everything.
Ai by itself can create content.
(05:38):
If you want it to be good, youhave to give it context, right.
So, actually, what would besomething else that makes a good
blog article?
What's something that wouldmake a good blog article right?
So it's already writing thisout.
But let's think about nowwhat's some further context?
What are things that we know gointo a great blog article that
(06:02):
we can give it some contextaround?
Ooh, okay.
So now diagrams, visuals Okay,this is great.
Could you make me aninfographic that covers these
main points?
Might have to flip over tomid-journey for it, but right.
(06:27):
So now we did the article, nowwe're asking for an infographic,
and then I could think ofanother example of like, and we
want to make it SEO optimizedfor blank.
That's what I mean in terms ofthe context you need to give it.
So, similarly, if you're tryingto go for a closing, I need
(06:47):
help prepping for a closing callyou have to give it the context
behind what good looks like.
The people that are going towin with you know, chatgpt and
everyone else are the ones thatcan give it the best context of
what they are looking for.
So now, look at this.
Now we just got an infographicand notice, I would like put my
(07:08):
own typo in there sales,leadership, rogals,
understanding your teals.
But look at this right, howlong did this take?
Now we have an infographic thatpeople could download, and now
we'd have to give itinstructions to fix it.
But now we went from nothing toa blog article, to an
infographic, and then we couldtell it to SEO, optimized for
(07:29):
these keywords, and it would dothat.
So this is how the prompt iseverything.
And then it's giving it thecontext right what does good
look like?
So it knows how to operate fromit.
So something I will use to allof my sales leaders, marketing
leaders, revops leaders startrecording everything.
(07:54):
When you are leading a training,record it.
When you are givinginstructions to somebody in
terms of how they can improvesomething, record it.
Revops, if you are reviewing adashboard, record yourself doing
that, because that's thecontext that you can feed into
ChatGPT to give it the insightsthat it needs.
(08:16):
So actually, I'll show you allthis real quick as an example.
This right here, this is overlike 15 hours of recorded
trainings that I'm leading atraining with my team.
I record it because I can feedit in.
(08:45):
It's that transfer.
That's the context, becausewhen you're teaching something,
you're explaining it.
Explaining it to chat GPT isthe key to making the results
better.
So the first big learning wasprompt is everything.
The second is context.
You need to tell it what goodlooks like.
You need to give the humanelement that you need to have in
(09:07):
there.
That's what makes it better.
Can you create trainingmaterials using that method?
Absolutely, so.
This is actually one Iliterally just did yesterday.
Okay, so I need you to help mecreate some lesson plans I have
done for other trainers todeliver.
Right, this is a training thatI want my own enablement team to
be able to deliver to mycurrent team, right?
(09:29):
So look at this y'all.
So I uploaded, right, thedocument.
This is the training, right?
This is the training that Idelivered.
I need your help to create thesesessions.
I need to outline these, thekey topics, with sub points
under them, with example,engagement questions to help ask
the students to apply.
I consult SAS companies.
I'll upload the transcriptionof the training.
So you have it.
I want them to be very thorough.
(09:51):
So this is the first thing andit spits out.
So it gives the sub points, theengagement questions of each
part of this training.
So far, so good.
I go further, right?
I'd like you to make the subpoints a bit more detailed.
Four to five points there foreach prompts, plus let's add in
an engagement question andexercise questions, meaning
(10:13):
questions that tasks.
So, danny, this is againfurther context.
Here's what I want you to doNow look what it comes back with
Sub points engagement,questions, exercise for each one
of the key portions of this andthen looking good.
Now the sub points.
I want the actual prompts thatit's talking about.
(10:36):
This took me 20 minutes to builda full training guide for one
of my trainings, completelymapped out, that then I can use
for my teams, for my managers,for my team leads, to be able to
actually apply it.
This is why, if you haverecordings, I actually did this
with one of my coaching clientsa few weeks ago with he had
(10:57):
actually what he did foronboarding Right, he uploaded a
persona.
Right, he uploaded a persona.
Right, he uploaded a persona.
And then we said, great, canyou write me a quiz to certify
my reps on?
Please include a mix ofmultiple choice questions, fill
in the blank questions and yesand no.
We're looking for an 80 to 90%pass rate, so don't make the
(11:20):
questions too hard, but weshould be able to pass this and
in five minutes, add a complete10 question quiz set up for one
of the enablement modules thatthen he can use with his team.
That's another big applicationthat we can use here, right?
So this is an example of atraining content that you can do
(11:40):
, and I don't know for any ofy'all, but, like writing,
writing this stuff is not great.
I'm a speaker, so I'll recordmyself giving the training.
I use a tool called Descript todo the transcriptions.
I know there's a lot of toolsout there to do transcriptions,
but I record my trainings, Iupload them into Descript and
then go from there.
(12:01):
Now I've got the transcriptionsand I can use it from there.
But my enablement team has usedthis for checklists, to map out
coaching plans, all of it, andit's great for creating quizzes
and things like that as well.
Powerpoints it seems tostruggle with a little bit more
than like PDFs and docs.
You still absolutely can, butit's a little bit better when
(12:24):
it's in a doc.
But, yes, you can upload it andhave it, do a quiz right then
and there for you, so it canwork in a big way.
So now, flipping kind of thetopic here, right?
So we prompt this king, give itthe context and I don't know if
y'all are catching this, but doyou notice whether this is
right or wrong how I talk to it?
(12:45):
How would you describe how Italk to it Right, so it did the
thing.
How would you describe myresponse back to it Right, it's
like, ok, it's starting to lookgood.
Now, in the sub points Iactually do talk to it like a
human, right.
I'm saying, hey, I liked this,so I'm training it like we're
(13:07):
close.
Keep doing more of that,because also learning lesson If
you don't this is so weird.
If you don't tell it you likedit, it changes the whole thing
sometimes, right, where if Ijust said, oh, I want more
bullet points, it changes thewhole thing sometimes, right,
where if I just said, oh, I wantmore bullet points, it changes
all of it versus hey, I likethis, we're looking good, I'm
(13:27):
coaching it along the way thesame way.
Now I want a short paragraphexplaining the bullet points,
how to speak to it, why thebullet point matters.
I want the instructor's guideto be almost like a scorecard
and script.
Talk to it like you would, ahuman.
This people.
It's weird Like they put ontheir prompt hat too and it gets
super, super complicated to doEngage with it, to get it to
(13:51):
where you want it to be.
Where I'm most excited is whatJake and I've been working on is
building of the custom GPTs.
Right, because each one ofthese is also right to go
through it.
Like it doesn't remember allthese conversations, whereas if
I go into the Bipsy bot that Ihave built here right, and come
(14:13):
into this all those you knowtrainings I showed you that's
already uploaded and built intothis bot, so it always knows it,
it always has the context to it, it always has the frameworks
already built in.
Is all of this content is inhere right?
The ability to build customGPTs.
(14:33):
You have to have the context todo it.
Look at all this context thatis built in here the sales
training, the coaching,everything else and then how I
want it to behave right.
Here's how I want you to do it.
I want you to act like a doctordoing a diagnosis, asking good
questions, ask clarifyingquestions.
Ideally, you can lead theperson to the answer through
(14:54):
questions.
Even better, give choices.
You map all of this out.
So now right, when you'reactually in here.
So again, it's like actualcontext, if I can say have a rep
that is losing deals at the endbecause whenever they ask for a
(15:17):
deposit, the prospect declinesand never comes back.
Create a training program withprompts and tips on how to
generate more deposits.
One thing we know works is byjustifying the deposit because
(15:47):
of how much upfront work andvalue we deliver.
Because of how much upfrontwork and value we deliver.
So now, because it has so muchof my sales, training, my
management, everything elseright, it can take all of that
and work with it.
Okay, so it can help make out atraining program right.
Enhance enough skill, secureand deposit duration.
(16:08):
This will be over two weeks,week one understand the value
and building confidence.
A training session focused onunderstanding the value activity
reps list out all the possiblevalues and benefits that justify
the deposit.
Right.
Crafting the justification.
Development like look at thisy'all, look at this.
Okay, this is big right Around.
(16:29):
All of this is and I don't evenhave a specific training on
deposits.
But can you see what this justdid?
Right, feedback sessionspersonalized feedback activity
reps create and respond to these.
Like this right, because it hasall the training and how to be
a good manager.
It can do this right.
(16:50):
So this is why I'm encouragingyou all to record everything.
Right Is because you can do itthis way.
So uploading it here does notgive everybody on ChatGPT access
to my documents.
That's not how it works.
The flip side, though, is alsotrue.
This is what Jake and I've beenbuilding is we've also created
a private instance for ourselvesand for companies to you.
(17:13):
That's completely contained.
So it's not on chat GPT, it'snot out there anywhere else.
It's a privately containedsystem that does the same things
with our content, the company'scontent and everything else.
So if you're afraid of privacy,you do need to build it in
internally.
If you want some ideas on how Ireach out to us, because we've
done it, we've actually built itand it's really starting to
(17:34):
work very well.
But, like, this is an exampleof what you can do, and I have
one for call scorecards too howto build a custom call scorecard
.
The custom GPTs are all of it,right Of all of it is like
taking the context, but if youdon't have context, it's going
to be very hard to build thecustom GPTs.
(17:55):
Even if you have the playbooks,you need to have the best
practices built out as well.
So the custom GPTs is one ofthe things I'm most excited
about.
We've built the manager one, webuilt a scorecard one, we built
a research one, a prep one, andmessaging the one that Jake
built out on creating snippetsto put into emails and LinkedIn
(18:16):
is just gold, right.
It's like chunking it down.
Which is my biggest tip I wantto give everyone is think big,
then think small.
The GPTs get better the morespecific the action you're
trying to get them to do is.
So in Bipsy, for example, Ihave my scorecard training in
(18:38):
there, but when I ask it tobuild a scorecard, it's not
nearly as good as the salesscorecard builder, because in
here I can give it even morespecific instructions.
I have the example scorecards.
I have the specificinstructions on how to do it.
So think big about what youwant it to do.
Then get small, build the GPTsfor it.
(19:00):
And I cannot express enoughy'all.
If you want help with this, werun workshops on this.
We can show you how to do this.
We have a tool that works wellfor this.
We have a whole prompt guide,right.
Like I don't know if manypeople have seen this.
We've talked about it a lot,but like, when we talk about the
prompt guide, like this is whatwe mean on the prompt guide,
(19:21):
and we have prompts for research, prompts for messaging, prompts
for closing deals, like it'sall in here already to get the
ball rolling for you.
That is the key on it, rightit's again, it's a great idea of
big to small when it comes tohow to build out these custom
(19:41):
GPTs, but you have to have thecontent and best practices to
make it work right.
So I was moving fast on thisone, but I hope this gave us
some ideas here.
One the prompt is everythingokay.
Again, if you want help, go getthe prompt guide.
It's like 50 bucks, y'all, it's$50.
(20:02):
Go buy it because it's going totrigger right.
I've literally I've spent wellover $500 on all other prompt
guides out there to learn itright.
Because you start to see thepatterns of good prompts and go,
ah, this is how it applies tomy world, those are the ways.
So scoop the prompt guide upy'all.
It'll just jumpstart yourprocess in learning this.
(20:25):
Okay, how do you know whenyou've done enough refinement
requests?
How many times Ooh, okay.
So I my my personal sweet spottends to be like four to five,
because if I'm giving it adifferent, I was like Ooh, I
liked this, but I didn't.
Similarly, I liked this, Ididn't like this part.
(20:46):
This part sounded too salesy.
Can you give me another examplewhere, instead of being salesy,
it's more human.
It doesn't include X, y, z.
If I give it four and five backand forth and I'm still not
getting what I want, I'll starta new channel and start with the
last prompt that I gave thatwas better and work from there,
(21:08):
because sometimes if you stay inthe same one, it falls.
It just doesn't change faster.
So I'd recommend four to five,and if it's still not getting
you to where you want it to be,then I would, I would say, start
a new chat to go through, andthat tends to get it back on
track.
One of the hardest parts ofearly go to market is not
(21:29):
knowing, maybe, what mattersright to the market, what
matters to what's going on outthere in the industry, what are
the trends, what are theinsights?
What's the language that yourpersonas like use?
Right, here's actually onestraight from the prompt guide.
Right, talk about likecompetitive insights, right of
like.
Hey, I came across thiscompetitor.
Why do you think people buythat?
(21:50):
How?
How would you compare it toours?
What are some keydifferentiators and how can we
leverage that information toclose more deals?
Right, so that's straight fromthe prompt-like guide.
So there's a lot of places wecan do that, but more than
anything, I'm just trying tostay on top of, like the prompt
engineering, the insights, andthen how can it connect the dots
(22:10):
?
That's one of my personalfavorite tactics within.
It is okay, told me this, nowconnect the dots to this, and
then it does such a great job ofdoing that.
Hopefully, you saw how fastthis works.
I appreciate y'all, as always.
If there's follow-up questions,hit me up, hit Jake up, but as
(22:34):
always, y'all deuces, we're out.
Appreciate you, thank you.
Popular Podcasts
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.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
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.