Robb Clarke and AI Support
Adam Robinson is joined by RB2B's Head of Tech Ops - Robb Clarke. Robb takes us through how he built RB2B's support system, which is 98% automated.
Somebody a few weeks ago, when we got asked, basically, something along lines
of like,
how do you support so many visitors with so few people?
I was like, well, we have this guy Rob, and like he uses this tool, Finn, that
Intercom,
you know, creates and between workflows and AI, we basically only have humans
responding
to like 2% of the total support interactions that we have.
And whoever that person was, I forgot.
They said, can you do a webinar on that?
So here we go.
Rob Clark is the presenter today.
I'll be the host, San Tosha will fire some questions at him too.
But Rob, can you take over Rob again to be clear?
Rob works with us at RPDB.
He, I'll let him tell you about himself.
Cool.
Hey everyone.
So like I've said, I'm Rob Clark on the head of technical operations for RPDB.
One of the many hats that I wear is building and managing our supported
infrastructure.
So making sure that when our users need support, they have a good time is one
of the main
things that I'm doing.
In that is, you know, building out our knowledge base.
So all the articles that people can read, building out our self-serve workflows
and our AI
support agent training it, making sure it is being effective and it's sort of
everything
correctly.
Those three things together seem to work really well.
And that's kind of how we're making our support experience as frictionless as
possible
for our users because I think I was talking to somebody earlier today and there
's 19,000
individual users across just about 13,000 accounts right now.
And that's a lot of people.
And trying to keep everybody going and happy and all that stuff is kind of one
of my main
priorities.
Describe this to Adam a while ago is I'm trying to do as much work as possible
by doing
as little work as possible.
And that's where all this comes in.
And the main benefit there obviously is that our users can get support 24/7,
get help when
they need it, solve their own problems and then should they need to escalate it
to a human
member of our team, somebody from the team can jump in and help.
So sweet.
So what do you have for us today?
Is it a presentation or are we just going to do a Q&A?
Like what is the format of this?
So I've got a little bit of info mostly because we get a lot of questions about
what we're
doing and how it's set up and why I get a lot of questions from people that see
what we're
doing for stats and then go how do we do this too or they don't know where to
start and
so foreign to them.
So I've got slides that have some data that back of you that's in all of the
teams.
So by the way, just to go over some top line stats, it's around 19,000 users.
We get around 700 paying customers, almost 2.5 million ARR, 2,800 support
interactions
per week.
That's like anyone touching any part of the support work, 2.0% involvement with
a human being,
93% CSAT.
And it's currently he's seeking another someone to help who has an engineering
background,
but it's a one man show, which is unbelievable.
So she just accepted today.
So is it a Julius sister?
It's Olivia Constantino.
So start on Monday.
Okay, cool.
Never mind.
We will withdraw that job offer.
So okay, we will officially now turn it over to Rob who can explain what the
heck he built
and why it works so well.
Yeah.
Give me a split second.
Well, as always, we'll be watching the chat, drop questions and then we can
answer in the
end and you know, you can drop in the Q and A also.
Yeah, let me find the appropriate way to share this one second share desktop.
Cool.
Can you see that?
Yep, I can see it.
Can we get a lot of people can see it in the chat?
Why?
Yeah, we got some yeses.
Awesome.
So, like I said earlier, we've kind of got a three pronged approach to self
serve.
We have our AI support agent.
The self serve workflows that people can go through and then our knowledge
based articles
and then the fallback is human support should they need it.
So just to back up a couple of stats that Adam and I were talking about, there
's our.
I did these two days ago on the 11th.
I did it on the weekend.
From August 1st, August 11th, we had 40 600 support instances and 97 of them
were handled
by a human.
So 2.1%, which is kind of awesome.
And then just kind of showing that it's improving and getting better from July
1st, August 11th,
we had 17,000 with 2.8% or 472 being handled by a human.
So that's pretty neat.
Our spin involvement has been getting really good.
Rob, what's fin?
So yeah, fin is our, it's intercom AI support agent.
We've named ours R2B2 because my dyslexia kicked in one day when I was typing
out our company
name.
And I was like, it's perfect name for a bot.
Thank you, Star Wars.
But please, Disney don't sue us.
So yeah, it's R2B2 and is our fin.
But yeah, fin is our AI support agent.
Right now, fin is involved in about 74% of support conversations and resolve 60
% of those
on its own.
So when people chat with fin, they're coming out with an answer about 60% of
the time.
Which is pretty cool.
And then from July 1st, August 11th, our resolution rate again, our involvement
rate again 75% with
the resolution rate of 56.
So it again is getting better.
It's learning and we're kind of evolving that over time.
Every time somebody has an interaction with the AI, we look at the conversation
and see
what can be improved, especially any time that they kind of rate it with the
bad rating.
I'll go through and look at every single one of those to see how the
conversations may
be improved and what were the AI went wrong.
And then especially any time somebody asks to speak to a human member of our
team, that's
the other thing I look at is why did they need to speak to us?
But did the AI not answer how can we prevent it from doing that in the future?
We're not always able to prevent it because there's a lot of stuff that the AI
simply doesn't
have access to at the moment.
I'm looking to change that in the future.
We can feed in an API into it and give it real time data on people's accounts.
They're up to the minute usage or what plan they're on, all that stuff.
We're currently not feeding that in, but I think when we do, we're going to see
a lot
of those, like, can I speak to human things go away?
Because those are a lot of the things the AI can't answer at the moment because
it simply
doesn't have that knowledge to be able to answer somebody.
So the other thing that I kind of talked about was our workflows.
I think I might have liked too far.
So when a user opens up our support widget, I do have a screenshot of that here
in a bit,
opens up the support widget, whether they're on our support site or within our
dashboard.
The initial thing that they see is how can we help?
And they've got a few ways to choose their own adventure.
They can ask a member of our team which engages the AI.
They can use a self-serve support, which is a bunch of workflows that are kind
of a choose
your own adventure.
Like, are you looking for help with installation or verification or privacy
policy, stuff like
that?
Which way they go, they can get to the answers that they want, or they can
search our knowledge
base, at which point we kind of prompt them to go to the right spot.
Now, if they ask to speak to that member of our team, it brings up the initial
interface
where they can speak with the AI.
So we used to introduce it specifically as an AI and say, I'm an AI support
agent.
But we had like an overwhelming number of people that immediately just went, I
want to talk
to the human because there's still a bit of a distrust of AI.
Adam Santosh and I are all old enough to remember terrible, terrible chatbot
days where there
was no intelligence behind it.
It was just garbage and wasted tons of your time.
So Rob, just to be clear here, we are saying connect to an agent.
In the agent there being connected to, they think is a person, but it is our
two B2.
And they're not pissed off about that.
I'm just not saying because I genuinely don't know.
No, we give them the option to ask our team, we're considering our AI part of
our team.
We don't explicitly say that it's a human, but we don't explicitly say that it
's not.
So can I ask you a question?
What makes this good enough to not piss people off?
Why is it that, how can it be that good?
I think there is a lot of financing along the way with how these initial three
messages have evolved over time.
So you think that how we're framing people, going into that, it's how we're
setting them
up.
It's sort of diffusing potential anger when they sort of figure out that this
is not an
actual person.
Exactly.
We're setting them up for success.
We definitely need to send these slides to people so that they can watch the
setup and
attempt to emulate it.
If you genuinely believe that that's the trick.
It seems to be working because as a team, yeah, be like previous messaging and
stuff that
didn't say like, kind of it didn't set the person up.
If you have multiple questions, let's focus on one at a time because people
used to come
in and just bump guard us and the AI with like a dozen questions.
And it's, I'm going to struggle with answering 12 minutes on.
Please just give me a minute.
I'm at all at that point.
And it's the same way.
So like if we set them up for success, people are having way more success with
it.
I had somebody today that actually ignored this and threw out a dozen questions
at it.
It did really well, but it didn't answer all of them successfully because it
was trying
to, it was struggling to handle all of it contextually.
Yeah.
Okay.
Yeah.
And then we kind of throw out there that, you know, if at any time they want to
do their
own self-surfing, they can say that and they'll take them back to that self-s
urf workflow.
And then if they ask to speak to a human at any time, it connects us.
Like they don't have any hoops to jump through.
It simply connects to the human team.
It gives a little note about timing.
No, I don't have that one.
So that's kind of the initial prompt for the AI, which seems to be pretty
effective with
users.
It sets them up for success with the interaction.
If they do do the self-surf workflow, they get into like what can we help with
today and
then they can choose their broad topics.
So like, like I said, script install, script verification, privacy policy,
understanding
data, technical support, whatever.
We do have what are called reusable workflows set up.
So that way, it's not one massive workflow.
They're chunked into different topics.
So that way we can reference those different topics and pass people to them
quickly and easily.
So we have one for privacy policy.
You can see there.
So if somebody picks privacy policy, it sends them to the privacy policy,
repeat it like
reusable workflow because at that, any other point I need to send them there, I
can do that.
I can just call that coding term would be like an include.
Same idea.
We can pull that little segment.
So the other last thing is just our our our user knowledge base.
This is set up twofold.
It's for our users.
It's it's organized in a way that it's useful to our users.
We've got like an overview.
We have a getting started section that has everything you need on day one.
A million one install guys debugging your script verification script install
your had update
your privacy policy, how that your cookie banner, it gives you a checklist of
everything
you need to do on day one.
And then it's broken into other things like here's how you use your dashboard,
the individual
integrations, whether it's HubSpot, Apollo, Clay, what have you different
features, some
billing stuff, etc.
So it's all organized for our users, but all the content is also organized and
formatted
in such a way that our AI gets the most out of it because this knowledge base
is the foundation
of our AI's brain.
So everything written in here is how the AI knows how to answer everything
amongst some
other sources, but that's the main part of it.
So anytime the AI kind of stumbles on a question or maybe it confuses some
context, I'll always
look at the original sources and figure out how it got two plus two equals five
And most of the time when it gets two plus two equals five and gives kind of
the wrong answer
contextually, it's because of praising I had won the other day that I used a
comma instead
of a period and it took two thoughts as one and using that it tied another
article there
and it answered wrong based on that.
And I was like, oh, that comma, cause that problem changed it to a period made
two unique
thoughts and like now it literally, I've tested that same question and it doesn
't have
the same problem, which is cool.
So really it's like setting the expectations in the front end and then having a
super dialed
in knowledge base on the back end are like the two most important ways to get
people
good support in this regard.
Yeah, exactly.
Really quick, I do want to just go back to one thing, one question.
I just saw a pop up in chat.
So they said, if the AI is so smart, why do we need to define these connections
I'm assuming you're talking about these workflows.
So what you see on the screen here has nothing to do with the AI.
This is just the users, sorry, the users self-serve support.
So if they're doing the choose your own adventure thing, this part is
completely independent
of the AI altogether.
This is just the user guiding themselves to finding the right answer.
So I just wanted to answer that contextually.
Somebody also said, what are you using for your software?
And we are using intercom.
It's kick ass and they've got all kinds of stuff in there under the hood to
make this super
valuable.
Okay, so that intercom knowledge base right there, that's like the boot page
and then this
is just an example of an article.
It's all, this is how to add our script to Google tag manager.
It's got step by step process like login to GTM, intercom credentials, etc.
Everything you need to do to do that is in the article.
And then every article in the knowledge base is optimized for one, the human
reader and
two, the AI so it can answer this stuff.
And it's a look at our dashboard with that workflow that I mentioned earlier
when somebody
opens up the chat, they'd have the option ask a team, use our self-service
support, etc.
This is me going, asking the AI how much a thousand credits would cost per
month for the
whatever.
And then it's a source which is the RV to be photo plan page.
That page doesn't actually frame the pricing in that manner.
It's got two columns showing what the monthly price and what the annual price
is, but it doesn't
have any like, verbiage in it to form a sentence in that way.
Like it doesn't say if you choose to pay annually, you get a discounted break,
blah, blah, blah.
Based on the numbers we put in on that page, it has determined that it's
actually discounted
break, which is true.
It's what Adam Gipper takes 17%.
So yeah, it has taken our data and contextually determined that it's a discount
to get the
annual plan, which is absolutely true and correct.
So if you feed at the right info, it will give the right answers.
And then this one is just an example of me using that self-guided workflow.
I picked like updating my privacy policy.
It asked me a bunch of questions like which topic do you want?
Where do I need to add?
And then from there, it gave me the verbiage that we recommend to people add
into the privacy
policy.
So again, if I was this user looking for privacy policy info, I could find that
answer right
there.
It's all available.
Okay.
So that's the last of like the contextual slides.
I just have nothing else to kind of show out there.
I'm just sharing button.
There we go.
Awesome.
Yeah.
So yeah, that should again, just get some context to some of the stuff that
Adam talked about.
I'm going to talk about and answer questions about.
So great.
I have a few questions.
So I'm sharing questions that I'm here.
All right.
Let's go.
So Rob, is there anything special about our use case or ICP that makes this
automation easier?
What do you think what you just did is applicable to all tech or non-tech
businesses?
Yeah, I don't think there's like this general setup could be, it doesn't
necessarily need
to be used by a tech company.
Like you don't have to be a high tech company to use an AI support agent or
anything like
this.
There's a ton of value in reducing like it could be a mom and pop shop that
they get a lot
of questions with their office hours.
When you guys open or how much does this product cost?
It could be an e-commerce or how much does this product cost, etc.
And using this answer a lot of those questions will save your support team.
A lot of time and effort because you can, the way I initially set it up on day
one is back
when everything was running through my inbox and I was answering every question
, every
question before we had all this.
I went and looked at all the questions that I was starting to answer in every
day and started
setting up articles and workflows and stuff for that.
And once I wasn't answering all those questions all the time, it freed me up to
start
building this more.
And a lot of exponential growth with the more time I had free, the more time I
had to commit
to improving this and making it better, which then freed up more time.
So it's definitely doesn't have to be anybody.
Right.
So if that starts up or not.
Excellent.
So if somebody is out there listening to this and thinking they want to
automate their
support, right?
When is the right time to start and what kind of team are team members do they
need with
what kind of background?
Well, I mean, the right time to start is today.
There's that old, I've used this a few times, like feel that it's one's the
best time to
plant a tree.
Well, 20 years ago, the second best time is today.
Same applies.
Like if you haven't already started, the best time to get started is today.
The type of person that you want, like setting all this up, you don't need to
be able
to need a developer to do this.
Like you don't need a development background, you just need.
You train the AI the same way you would train a human employee.
So you're onboarding a new support person, train your AI the same way, give it
the same
information.
I didn't develop this AI at all.
That's it.
Intercom did it.
I was running on open AI's infrastructure.
All I did was train it on our stuff and how to answer these questions.
So you don't, again, you don't need a development background.
You're not going to be developing an AI chatbot.
You're going to be training a chatbot the same way you train a human member of
your team.
Having a bit of an analytical mind helps because you can see where things are
going wrong
and kind of fix it.
You kind of ask like what resources you would put behind it.
It's not something initially.
Like day one that you're going to be able to say, I'm going to put five hours
into this
a week and see success.
Like you're going to, at the start, you're definitely going to have to dedicate
somebody
to like building that out.
But over time, their involvement in it can lessen.
We've been lucky enough to, I've been lucky enough to be able to have this as
like the core
part of what I've been doing for the last little while.
But it's not my, like the main part of my job.
But yeah, just having somebody that at least initially is kind of dedicated to
helping build
it and get going goes a long way.
But realizing also that a lot of five hours a week isn't going to get you over
the finish
line any time soon.
Right.
So Rob, once you build this and it's working, how do you cope with change?
So if the pricing changes or the process or workflow or even product changes.
So how do you go back?
What's your cadence to keep this up to date?
Assuming I have heads up when changes are happening, which is in those cases, I
will try and
re-edit a bunch of articles and just not publish the changes.
So like we on June 6th, so about two months ago, we changed, you know,
completely changed our
price structure.
Twice.
Twice.
Yeah, that was far.
June 6th, we released new pricing and then June 6th by about 10 a.m. we
realized that we were
wrong.
So for June 6th, part one, I actually updated all the pricing articles and
stuff, leading
up to it.
And then when I got the green light that we flipped the switch, I went and hit
publish on a couple
of those articles.
Workflows, I wasn't that, no, sorry, workflows.
I did the same thing, I kind of pre-edited them, saved them as edits and then
published them.
Now the good thing is with the AI, it will go and update its knowledge as soon
as you update
an article.
So as soon as I published all those articles, the AI started ingesting them and
going, everything
I knew about the 495 plan is now gone.
Here's our new pricing structure and it expunged everything about 495 out of
its knowledge
and started going on the new structure.
And then later that day when we changed it again and I added the articles again
, it
undid everything I taught at that morning.
So as soon as I had published, it starts ingesting that content and updating
its knowledge and
evolves like that.
Nice.
So how do you know when you hit publish, it's not going to just start spouting
off a bunch
of nonsense?
Do you do any testing yourself by asking it or do you just wait for customers
to do it?
There's a test function in there where I can have the test conversation.
You're testing it and sort of dynamically changing and getting to an MVP level
of quality.
Yeah, it's absolutely not a spray and spray.
When I'm updating those articles, I'll go and like when we changed all the
pricing, I went
and started asking the AI with the test feature, asking common questions that
people would
have, especially when we got into our credit usage, right?
Like that is a completely different model.
So I started asking questions that I caught.
The users would ask and if it kind of, it stumbled, I went and added the
article.
So it wasn't just like a set it for get it walk away.
I again did my testing on this end, but also very much pay attention to the
conversations
that the AI is having with people and make changes daily based on those
interactions.
So a few months ago, I was reading, I think Air Canada has had some bad
experience with chat
bot on multiple instances, right?
So the one that was recently news is they taught their chat bot.
I think it was some variant of chat GPD, but they taught it to give incentives
and discounts
to disgruntled passengers and some passengers figured out how to trick it into
giving a lot
of discounts and this thing went to quote an AI Canada of course lost.
But how do you ensure that like other do you think about how much power you
should give
this chat box?
Yeah, well that's why I don't have it tied into our building platform.
It is somebody who's going to be sneaky.
It's not that there's a distress of people, but knowing that these are flawed,
like there's
some AI's, they are programmed to provide you with an answer that it thinks
will appease
you.
So often they hallucinate answers and you can kind of bend that to your
advantage like
the Air Canada case.
So I don't have it tied into a building system.
I don't have any knowledge in there about possible discounts or like free
credits or anything
like that.
It doesn't have that mold so it couldn't hallucinate that info.
I try to avoid it giving, I try to avoid giving it partial info like that.
So it can't.
I haven't run into anybody trying to do that yet so that's good.
I've known that somebody's rushing in there right now to do it.
So we will see.
Yeah, I again, I avoid that by just keeping it as far away from that as
possible.
Like I said, it doesn't have access to our building platform.
There's no way that if somebody asked for a restart on their trial, it goes to
a human
member of our team and then we handle it.
Because otherwise if you kind of programmed it to like restart somebody's trial
, if they
ask for it, well, somebody somewhere could be like, I'm just going to be free
forever
because I'm going to ask the AI and it's going to make me happy by giving me
that.
And while I very, very, very rarely tell somebody no one to restart or the
trial, we do, you
know, it's almost always in our best interest.
Really?
Yeah, you didn't get the script installed until day 29.
No problem.
We'll hook you up.
But if we gave anybody carte blanche to restart the trial just simply by asking
the AI, that
would be a terrible business practice.
We get around.
That was the way to use our BDB for free forever.
Didn't I just see a post about somebody going, hey, I stole leads from Adam
Robinson?
Oh, yeah.
Well, I know that would actually do.
I just thought it was hilarious that the guy's title, the guy's title on his
LinkedIn post
was, I have better hair than Adam Robinson.
That's his job.
And he does, by the way, it has like this curly like shoulder length locks.
It's really hilarious.
So, yeah, so I just want to ask the audience, is this, do you feel as listening
to this?
Is this something that you could like turn around and go attempt to do or does
it seem like
there's some distance between what we're saying and where you're at?
I'm just curious.
If you can chime in and give me your opinions, is this like, are you feeling
compelled to
go download Intercom and try to do this?
Yeah, seems pretty cut and dry.
That's what we're hoping for because we think it is.
And we think it's super valuable.
So that's what we're doing this.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
And that's what we're going to do.
So Adam, we are out of questions.
Look,
you know, if anybody else wants to ask Rob or us anything, you can.
I don't mind.
wrapping early, because we've never done that,
giving people 15 minutes back.
I do want to plug next week.
Next week, Sanchez and I are going to sit in the woods
and meditate and try to come up
with our next great product idea separately,
but on the same trip.
I did it last year, Sanchez sort of loved the mood
that I was in when I came back,
so we're gonna fly to Aspen,
we're gonna go into the woods
and then separate and then join up six days later.
And figure out what we learned.
So we're not gonna be here.
Pete Crowley is gonna be the host.
Pete Ron's growth for us, he's our head of growth.
He is going to have the understory guy overstory.
(laughs)
Is it understory, overstory?
Our LinkedIn thought leadership ad agency,
they're going to be on our show
and they're gonna show you everything about our ad account.
So what, yeah, understory.
What are the ads we're running?
What are the audiences we're running?
What are, how are they doing?
What are the metrics?
What's our row as?
How to think about row as with thought leadership ads
'cause it's not quite the same as a direct response
with thought leadership ad.
They'll show you the actual posts.
They're going to do what no one else would do,
which is actually show you inside of our account.
So don't miss that one.
It'll be great.
I'll drop the registration link in to the chat.
It's the same registration link is this week.
It's always the same.
So, yeah, I mean, that's all I have to say.
Let me just drop the same work link.
So you guys have it.
Whoops, invalid meeting ID.
Sorry.
Rob, thank you very much for coming on.
- No problem.
- People are saying that it was valuable.
By the way, Rob Clark,
Rob with 2B's Clark with an E.
He writes about this stuff on LinkedIn, if you want to follow him.
- Just to say that,
that's following in Adam's very big footsteps,
a little bit here and trying to work in public
with a lot of this.
Anytime I have a win, I share it, any stats,
any misses, I share those two.
Like, hey, I screwed up on this or whatever.
And it's not just AI support stuff.
It's also just talking about how I'm using AI.
Like, our home life, I mentioned,
they using it for homework with my daughter.
Super useful in that regard too.
Like, anytime I find something useful with AI,
I'll kind of share it.
But also our B2B support success and public.
- Well, everybody, thanks for coming today.
This will be published live on our YouTube channel tomorrow.
There will be a freebie that you can get,
which are the slides that Rob published.
So you can see how we're setting everybody up.
And different kind of stuff we're doing.
And thank you very much.
Hope to see you next week, even though Santer Shennel
will not see you next week.
Next week could be the best show of all.
Nothing against this show, Rob.
But people want to know the metrics, you know?
- Oh, no, it's okay.
- They want to see what's going on.
BIA, Pete says that it will.
- Pete's the man.
So I'm not competing with Pete.
- They want to see what's going on
with the numbers behind the curtain.
Cool, with that, everybody have a great rest of the day.
Enjoy this 10 minutes you got back.
And we will see you next week.
Oh yeah, there's Rob just in like,
I'll just let people, I'll wait 60 seconds to close out
if you want to sign up for the thing.
- My bad.
- Sorry.
- Bye everybody.
Bye everybody.
Great show Rob, thank you very much.
- Thanks Adam.
- Thanks on touch.
[BLANK_AUDIO]