#8: Using Automation & AI to Supercharge Your Sales Pipeline
Welcome back to another episode at The Junction.
I'm really excited about today's episode, Jace, because we're
going to get a little bit more specific.
Some use cases that we are
utilizing in house for yeah, internally. Internally.
To help our sales teams and Salespeople
do what they do best and maybe
automate the things that they're repetitive.
That are repetitive. Yeah.
So let's talk through how AI? I just want to go out here
and say AI alone is not enough.
Having an AI, a sales AI tool is not enough.
We believe in pairing that with the automation.
If you have multiple AI tools but none
of them are connected, you're still manually probably
importing or exporting that data or manually jumping
in, doing the swivel chair looking between systems.
Right.
So tell us a little bit, how have we bridged the gap
between systems with automation and then we can kind of get into
some of the value of adding AI on top of that.
Well, we talked about data in the last episode
and what a lot of these tools are doing
are accessing the data, but only from one platform.
Right.
Salesforce just had a big announcement with all
the GPT cool things that they're really doing.
And no lie, those are going to be really great,
but they're only going to access what is in Salesforce.
Right.
Where you can get even more benefit
is if you can pair automation with
the AI to now access different databases. Right.
Let's connect to HubSpot salesforce
and your ERP system.
And now the AI has the ability to look across all three
systems and not just what's in that one database or not just
what you fed it to train it from one database. Exactly.
Well, you take a train model, right.
And you give it access to more context and it's naturally
going to be better at what you ask it to do.
What I'm really excited about
is something we're doing internally.
And I'll make this disclaimer.
A lot of what we're going to
talk about today are things that people
and organizations are already doing there's.
Gong is we were talking about
this right before we started.
They do a lot of what we're going to talk about.
Chat Spot is a new HubSpot tool that
will do some of what we're talking about.
And just like I mentioned, Salesforce is going to
do some of what we talked about and if
those are up your alley, go for it.
What we're talking about is maybe one step beyond that.
Each of those tools has maybe kind of like
this, not necessarily a specific use case, but like
a specific area that it focuses on. Right.
Gong focuses on call recordings.
Salesforce GBT is going to be looking at the
data and the records that it has access to.
Chatspun is going to be looking
at building out or writing content. Right.
Write me an email. Right.
Let me look at the competitors that are
associated with my company, things like that.
What we're trying to do is actually take all of
those, build some automation to basically kind of connect those
three areas and then utilize AI to sit right on
top of that and give us insights that we could
go get with each of those three different tool sets.
Or we could just give it all in one spot.
So the first place that we're going to
utilize this is with these call transcripts.
We actually do pull these out
through a lot of automation.
We back everything up into Google
Drive and then we pull yes.
From Zoom, we pull that out, we put it in
Google Drive, then we send that over to Salesforce.
Salesforce now has a record of every meeting that we've
ever had right on the account, on contact opportunity.
That is actually coming.
That is our hope right now.
We've got a record being
created for the meeting recording.
We upload those files or we effectively link to them.
We do upload, I believe, the transcript
to the file object inside of Salesforce.
And what we're going to end
up doing is giving people access.
And we've been doing this, to be
clear, for a while, many years. Right.
So we've got lots of historical recordings.
Oh, yeah, we've got a ton.
Because Zoom doesn't keep them, right.
After like 30 days.
No, I mean, they'll charge you
more money if you want to. I see.
Store it for longer. Okay.
We actually went way over that limit and they
didn't tell us or didn't contact us and try
to charge us more money for a long time.
And then they finally did and we paired
that back because it was kind of funny.
It is interesting though, because if you have people who
don't have a salesforce license but they need access to
a recording for some reason, or not only attaching it
to the record in salesforce, but then also having a
backup in Google Drive serves more than one purpose.
It's backed up.
But then again, not everyone in every has a salesforce.
So you now have well, speaking to what we
are already doing or what we have been doing,
a meeting recording message now goes in slack to
inform you, which I love, by the way.
Hey, the meeting is finished. Yeah. And here's the link.
Here's the link.
You can then send that out to other team members.
Team members, prospects, partners. Yeah.
So we're going to extend that. Right.
We've got a link to the recording.
It pulls up a record in salesforce.
We actually utilize a lot of salesforce privacy and
security settings to make sure that only the right
people have the right access to those meeting recordings.
That's important.
But the next step that we're going to take
is add this layer and this is more what
kind of gong does, right, gong analyzes the recordings.
It gives you the ability to kind of ask
these questions through AI and we're going to do
something very similar and give folks the ability to
ask questions of that specific meeting recording within Slack.
Within Slack within the meeting
recording record inside of salesforce.
So you can easily say, what were
some action items from this call?
And it will spit those out.
You could ask it to summarize what's the summary of the
call, how was the tone, I mean, you name it.
And the benefit here is that we can ask any
question that we can dream up of and those questions
and those answers get pulled up right in front of
you the moment that you ask them again.
Gong has done this for a long time.
We're big fans of Gong.
So if you already have Gong, then
maybe that won't help out so much.
But it's not necessarily just the AI piece.
It's the junction of AI and automation that gives
you much more efficiency that your team will be
able to utilize and they'll just be able to
do a whole lot more because they're not spending
time where's the meeting recording, right.
Or I can't remember, what did we say?
Or one of the things we talked about earlier
was we're going to call it Ask Bjorn, right?
Hey, where's the meeting recording from Thursday of
last week when I talked to Joe Schmo? Right.
And we'll be able to go find those things.
So when you combine that way, I'm not going back
to my calendar, clicking on copying pasting the meeting.
We do this all the time.
It takes me like ten minutes maybe to have the
first thought to pulling up the transcript and reading it.
Ten minutes for every call over the course of a year?
Yeah, let's say five, let's
just be conservative, right? Five minutes.
Let's say I do that maybe, I don't know, we'll
say ten times a week, which is probably a little
much, but on average across all the employees, right?
And let's say you do that for ten employees, right?
So five minutes, ten times that's 50
minutes, that's 500 minutes for ten employees.
That's a decent amount of time that your
team is just going around looking for answers.
And if you can provide that automation that just
pulls that up in front of them and then
they can go ask these questions, you're naturally going
to get a ton of efficiency out of it.
Well, and you can feasibly deliver a better customer
experience because you're back to back to back. You say?
Hey, Mr. And Mrs.
Prospect, I'm going to send you that call
recording after at the end of the day.
Oh yeah, you run a few meetings.
All you got to do, pop into
Slack, see your recordings, copy paste feasibly.
If you're not back to back, then
you're doing it right when it surfaces.
And so I just think it's not just a value
add internally and saving me time or you time we're
now delivering on what we said we would do for
the prospect or the customer or the partner faster.
Well, you actually just gave me a great idea.
And this is one of the things we're already doing.
Right before the meeting happens.
Literally 1 minute before a meeting happens in
Slack, it will pop up and say, hey,
you've got a meeting starting here soon.
Here's the title.
Here's some of the attendees right click
the link to join the meeting.
Makes it super easy so you don't
have to try to where's my calendar?
Okay, I've got 8000 tabs open, right?
Just click the button, boom, you're in the meeting.
And for those of us that are not always watching
the clock, I do this all the time, right?
And I'm joining at 859, right?
And 59 seconds.
At the last possible moment, I click the button.
What I really think would be awesome is if
in that same message, it pops up links to
previous recordings that I've had with this person.
And now even above and beyond that, a summary
of what we talked about in that last meeting.
Because I'm telling you, every time I get
on that meeting, I'm like, oh, shoot.
What did we talk about?
Hey, guys, sorry, I'm just hopping from back.
Can you remind me we're meeting for meeting date.
Oh, yeah.
It's good to see you guys.
Since we met last Thursday, here are
the top three things that we covered.
Yeah, can you do that? Oh, for sure.
Can you go build it right now?
Yeah, we'll pause this commercial break.
Yeah, but I can't tell you how many times
now I'm super pumped about this, how many times
I've gotten on the phone and I literally say,
hey guys, I'm just hopping off another call.
Sorry I'm a minute late.
Like, I don't have anything pulled up.
Can you remind me what we're talking about?
Not only I'm on your back foot. Yeah.
I won't say that it's maybe slightly unprofessional.
It's not the best experience. Right.
For the person that you're meeting with, whether it's
a partner, customer, prospect, you want them to feel
like you spent time exactly going over your notes
in between, and you're ready to roll, and boom.
Within 15 seconds, you have all
of that data that you need.
And to be fair, these are some things that
you could do with AI by itself, but that
level of automation is something that we are slowly
building out to give people what they're looking for.
I do think that we have done some really
unique, cool things with Slack between our other apps.
We're huge.
We live in slack. There's no question.
I've actually never I think it's one of
the first companies I've worked for in my
career that there is a one singular channel.
Prior to I had roles where there was no chat, right?
Email was king.
But you also had phone calls, and then there
was kind of the in person things that happened.
But truly everything for us then happens in Slack.
And so if you need to get hold of someone
or that's our like we talk about content is king.
If I want to know the answer to something
or pull something up, I'm not searching my inbox.
Now, granted, I am not as in much
of a customer facing role, you and Brenton
folks in sales, delivery, you're also managing clients.
But we have shared channels with
clients and partners in Slack, right?
So I do think we've done
some really unique things there.
And the more information and visibility that
we can get from these various systems
pulled to Slack, the better.
I'm getting really bad about making this correlation
of the intern, right, and the AI, right?
But in reality, we're talking about how much
better AI gets when you provide more context.
Well, AI is really just kind of a replica
of an individual person that is doing work.
And the more context you give AI,
the better it's going to get.
The more context you give somebody, the better they're
going to be prepared, better they're going to perform.
And so that's what we're trying to do. Right.
It's a people process, technology.
And if you can bring all those into really
tight knit circles that kind of look like our
logo, then you're going to be very successful, right?
You're going to set that person up for success.
So we've talked a little bit about the
automations that we're doing today, and now we're
actively discussing internally what layering AI on top
of that would look like.
Let's talk about the one specific use case
around zoom recordings that we've been exploring lately.
So the way we've broken this out for listeners, if
you're doing your own building out use cases internally, we
put it on a slide, what's your problem?
Statement call recordings.
We can access them, right?
We've talked about all the cool
different places we dropped them into.
We even send them to you in Slack.
We make it so easy, but they take too long to digest.
We don't have that summary today. Yeah, right.
Well, I mean, you run a 60 minutes
phone call and on average you probably speak,
I don't know, 60 words per minute.
I'm totally making that up.
But there's so much content and data
from a raw data standpoint that gets
not transcribed transpired between two individuals.
If you're trying to digest that in 30 seconds,
it's probably impossible to digest the entire phone call
in 30 seconds and be ready to roll and
hop onto the next phone call.
And where we can really benefit is asking
all of these questions potentially in advance.
Not only give folks like the ability to have a
chat area where they can ask questions, but let's provide
that high level summary on a sales call.
Here are the five questions or things
that we want to know, right?
So you're saying that we would feed this model
a template of sorts and then with every call
that is tagged, this if it's a sales call. Yeah. Right.
It's an opportunity in the demo stage. Exactly.
We want it to summarize next steps,
sentiment, questions, concerns, all in Slack. Yeah.
Well, this goes into you were talking about the slide.
It goes into this idea of there's
utility approaches and there's insightful approaches.
Right.
The utility gives me the ability to utilize
it and do let me type some stuff. Right.
The Insight is kind of maybe proactive in a sense.
Like it happens before something.
I'm not necessarily typing anything, but if you
do both, you're in a really good spot.
If you can just do one of them, you're in a good spot.
But we're going to focus on both.
The utility side would be the chat piece. Right.
Let me ask questions of this transcript.
And the Insight piece is where we're going
to preemptively ask questions of that call recording.
Right.
I'm hoping that as we go build this, at least
on the marketing side, if we're collecting that and putting
that into Slack, we're summarizing that in some form.
Oh, yeah.
And then we can go look at it in mass. Exactly.
Well, you kind of stole my
thunder, but that's where I'm going.
And we broke this out. Right.
The best way to probably go about
this is write out use cases. Right.
Think of areas where you might benefit from this
because you are going to have to ask the
question of yourself if this is something we want
to proactively or preemptively do, you're going to have
to kind of go program it, if you will. Right.
Like with the chat idea, I can just instantly
think of a question and type it out.
But that's what we did.
We wrote down a bunch of use cases here.
One of them was the summary. Right.
Let's just summarize the call.
That's pretty straightforward.
A lot of tools can do that.
You do have to be a little bit worried
about the context that you provide it because sometimes
the call transcript is a little too long.
We can talk about that more.
We were thinking maybe we can
get next steps or action items.
There was an old tool out there that would hop
in on your zoom recordings and you could be like
action item and then say notetaker or something.
You'd have to do the keyword to get
it to key off of, well, Chat, GBT.
These AI systems, they don't need the keyword.
They can just say they can hear us
talk about like, oh yeah, that's something that
I'll do when we get off the phone.
Well, that sounds a lot like an action item. Right.
And these AI systems can pick up on that.
The other one, or another one at least
was bulleting out requirements and scope, at least
as it relates to that one particular call.
Maybe the customer is like, hey, I just
really need it to show me all of
the, I don't know, the best opportunities.
Like I want to be able to see my opportunities,
my opportunities that are going to close this month.
Okay, that's a requirement, right?
They need to be able to do that.
It's not necessarily how we would go do it,
but they're saying this is what I want. Right?
And it can pull that out.
Sentiment is another one.
And I'll be honest, this one's probably maybe not cliche
to an extent, but it is relevant not necessarily to
this one phone call, but Mel, like what you were
saying is, well, if we have a thousand calls and
we have this sentiment across some kind of score right?
Across a thousand calls, well, now that starts to
become really powerful because I can say, hey, AI.
Or Ask Bjorn.
Ask Bjorn.
Tell me what the sentiment is.
Like when Mel is on the phone with customers
and it's going to send back and say, hey,
the average score for Mel is a ten. Right.
But maybe you have somebody that
is new and needs some help.
Well, you can ask what's the sentiment of those
phone calls when Chase is on the phone and
it's going to be like, yeah, that's a three.
But if you have these scores or this data that's being
recorded for every call and you've got thousands of calls now
you can start to get a ton of insight, right?
You can even have the AI and this is something we
will probably end up doing, right, is saying across all of
the sales calls, who has the happiest sentiments and let's see
if that correlates too, by the way, let's check it to
close rate or we run into this.
Any business runs into this.
Every salesperson knows when a deal is like I've got
them, I've got the customer, they're going to sign this
week and you know, I never do that, never. Right?
And so we talk a little bit about push
rate and so any company kind of reports against
a pipeline, it can be subjective, right?
Yeah, you're doing your due diligence.
Also things just happen like people ghost people.
It happens in the sales cycle.
So it's not meant to be weaponized, but more to
understand how of what we say and feel like we're
going to go close this month, what actually books and
then what pushes and oh by the way, let's pair
that with sentiment and let's pair that with and then
we can identify training gaps.
We can identify maybe there's even as you're
talking through requirements, I'm starting to think upsell
cross sell opportunities that might not be as
obvious to someone, especially if they are new.
You've got a new account manager or a
new salesperson online and they're not differentiating between
that's in scope, out of scope.
This could be upsell crosssell opportunity
based on what you provide.
Well, you're talking about that crossover
from AI and the automation.
I was just thinking, wouldn't it be great if
the AI automatically pulls up or sends out notifications
when there's a call that is worth quoting, right?
Like, hey, this was a really great call.
Sentiment was really high.
Customer said these really great things. Boom.
Let's send that out to the marketing team automatically
give them access to the transcript, give them the
quote that they should quote in their social channels.
Right.
Like that automation provides that kind of efficiency that
we don't have right now because we still need
Brent work chase to be like to remember which
it's really a challenge for me, right.
But to remember and say, hey Mel.
Hey Randall, I just had this great sometimes, you know,
we get really excited about those things but boom.
Now I'm back to okay, we got to close the deal.
Yeah, even just after post deal, right.
We've delivered or we're in delivery or we
have delivered a solution project and we get
folks on the phone saying the same stuff.
Gosh, I'm so happy this integration
saved me so much time.
Of course we're always trying to remind
folks internally, hey, if you have a
great call, send me the recording.
But to your point, tool like this can flag it
and surface it without marketing relying on sales or delivery.
Yeah, well, one of the benefits that we're going to
get here because we are already doing all this automation
when we layer in the AI, not only for sales
and I know our episode is mostly focused on sales,
but we are pulling in all of our calls and
we're not just doing sales, right.
We're doing recruiting.
We're having internal meetings.
We're talking about how we should market. Right.
We've got these calls or recordings, right.
Basically anything we can put in here.
We're now going to benefit from some level of
AI and being able to ask those questions.
I think this will probably come across really well or
work out really well for us on the recruiting side.
I love the recruiting use cases.
I can't tell you how many times people are like,
oh, I read every article, I watched every video.
Those are things that we should be putting
on the like, well, it signals to marketing.
Let's keep doing it.
I can obviously look at views
and impressions and other metrics.
I can ask, I always ask Qualitatively when
we have a candidate in the office, how'd
you find know, Sam does a great job.
She does all of our phone screening of
asking if they've been out on the website
and looked at any of our content.
But that additional insight automatically tells us
double down on those things or we're
not getting much traction in this.
Maybe we stop it's.
That start stop mechanism. Yeah, exactly.
Without the manual analysis by a person. Exactly.
Why don't we talk about what other people are doing.
We've got some good use cases internally.
We've talked about gong a little bit.
What about Chatspot and what HubSpot is utilizing?
How is HubSpot utilizing AI to
do things inside of HubSpot?
I'm thinking about content creation.
They've got the ability to basically
write your emails for you.
What are you doing internally at Venn to utilize
some of the tools that HubSpot has pulled mean
they're rolling out new features all the time.
I feel like every time I go back into it and do
a refresh, it's given me another tour, which I mean, props.
I'm a huge fan of HubSpot.
I have played around a little bit
with some of the content rewrites features.
We talked about this in an
earlier episode around marketing use cases.
I would love to do more analysis around okay.
Of the emails, it tells you this
email is performing poorly, but why?
I don't know. Does it?
I haven't spent enough time with it, but
I do think that it would be interesting
if they took it that direction.
To say this email and this Drip that you have
email three of four super low clicks, low engagement.
Here is what?
Our AI content right.
Or here's like an example rewrite or here's. Exactly.
Yeah.
Based on the activity and
the engagement or lack thereof.
And it's looking at all four emails and
the content, the call to action, the tone.
At some point, a human wrote
all of these emails, right?
I can't say that we have, since they've rolled
out the email generation, kind of go, write me
this based on these three things, we're not running
Drips today that are all AI generated content.
I think that's probably where things are going, right?
It would also be interesting for
it to proactively suggest campaigns.
Wouldn't that be cool?
That would be awesome.
Hey, build out the whole camp.
Or template, right?
Build a template for it. We talked about it, like, it's
not going to replace the idea.
We talked about the Venn screen, the original
concept, but actual execution or identifying areas where,
hey, you've got highly engaged people.
They keep coming to the website, they
keep going to these types of articles.
Again, those are things that I'm already
doing today, our team is actively doing.
But to have that predictive or an additional
assistant that intern, so to speak, that says,
I'm noticing a lot of traffic to these
pages and maybe we're getting return visitors.
Can it propose or suggest content around it?
I love that and I think that's one of the
things that we are focusing on internally is not like
something that is just out of the blue.
We've never done it before.
Like, man, this could give us a
lot of bang for our buck.
But what we're focusing on and utilizing AI and
automation are things that we're already doing right.
Like, hey Mel, can you come
up with a really cool campaign?
And now you have to not only think about what we're
going to do, but then you've got to template it out.
You've got to write down all the tasks. Right?
Today let's take it back to HubSpot.
When we go, we roll out, let's just say a
very light campaign or we know we've got one video
we want to send to customers via email and there's
a call to action to fill out a form.
If I could tell HubSpot to
go build me a yeah, triggered.
This is the trigger, this is the thing.
And it goes and builds me a workflow.
Today I'm doing the point and click, which is fine,
I know how to do it, but can it do
it for me faster based on prompts or direction?
I would give that direction to someone on my team
anyways, hey Randall, can you go build a workflow?
The trigger criteria is this list segmentation is a partner
and we want to make them a marketing contact and
we build an if then branch for that.
I mean, these are things that maybe HubSpot's already
talking about it and I've just missed it somewhere,
but that's where my brain goes with it.
Yeah, well, salesforce is they
had a press release, right?
They're going to do all these things with GPT.
One of them was flow, GPT and flows.
If you're not a salesforce user, much like HubSpot workflows
this idea of kind of process automation, do this first,
do that next, so on and so forth.
And what they're going to do is basically you can type
in a chat, build out a flow that will update all
Texas where it's TX, change that to Texas, right.
And then you can type that and
it will so that's been rolled out.
I don't know if it's rolled out.
If you are a salesforce user, you know, it takes some
time from the announcement to it actually being a thing.
But no, that's cool. They do deliver.
So I'm really interested in using that and
I think we will we were talking about
so how does that shift, though?
So we have salesforce consultants on the
team that build flows for clients.
Are they doing it more and faster?
Like, again, I don't see it replacing no, this will
go back to this idea that you've got an expert
that can whip them out very quickly because they have
all the context, they understand how it works.
They've been doing a lot.
They've built thousands of flows.
That individual isn't going to get a lot of benefit out
of the chat bot that builds the flow for you.
But for somebody that is not a salesforce
flow expert, maybe they've never done it before.
They can have this piece go build out
like, I don't know, rough drafts, right? Templates.
It can do simple things.
It's probably not going to go like
solve your super challenging 100 component flow. Right.
I was going to say this sounds.
Like something that might be you have
to take into consideration with this stuff.
A lot of the environments we work
in are highly complex, super configured. Right.
So it'll be interesting to see how it
would respond to a more complex environment or
these kind of more like low level rudimentary.
This is how I roll out a campaign.
I have an email, I have a landing page,
I have a forum, I have a trigger criteria.
And go build me that and then I
can come behind you and tweak it.
Well, there's probably one other spot
where it will be really beneficial.
And that's if you're a salesforce
person, building formulas is never fun.
And if I could in that right.
Or in HubSpot in the workflow, I could say
build me out one action or one component that
does this specific thing that could be really powerful.
I can't tell you I'm pretty good at formulas, but
I'm not going to get them perfect every time.
And if I could just have it kind of bypass
me going and googling the documentation, what is the syntax?
Let me get that exactly right.
And it could just do that for me and
give me a pretty good push in the right
direction, that would save me a ton of time.
Yeah, I think we're getting a little excited, a
little carried away with what we can go do.
I think we should bring it back full circle.
So AI, for the sake of AI is great,
but we're firm believers in the power of automation.
Layering the AI on top.
Yeah, right.
And let's set aside AI and automation altogether.
And really why we started this really the
foundation of then it's people process and technology.
Right.
But take the technology out of it people and then
your process and then go add your technology to it.
So at the end of the day, none of
this works unless you have a clear problem statement,
unless you have a clear understanding of your process.
How will this impact your process? Yep.
You have to have all three of those.
When we first started, Scott gave me that idea.
He was the one.
Maybe he came up with it, maybe he didn't.
But he was the first I heard it from.
And I was like, whatever, we
can just throw technology at this. Right?
And then as I got into the
consulting world, I realized, well, I can't
do the technology piece without the people. Right.
And then as I was getting further into it, a
lot of these folks don't have their processes written down.
They're not documented. They're not documented.
They don't even know what they want.
And we can't deliver a solution
without all three of those things.
And so if you try to do it,
you're naturally going to figure out some of
the things that we've learned over time.
Well, we want to know what you think.
We talked about a lot of different subjects.
Covered a lot of ground, I think, on this episode.
But how are you using these tools?
It doesn't have to do with zoom
or transcripts or even your sales process.
We'd love to know.
Email us your take@thejunctionadventechnology.com.
In the meantime, what do we say? Chase?
Keep it automated.
