Big Idea

The argument in one line.

AI raises close rate when it automates the grinding, repetitive sales tasks that humans skip, but only on top of a proven manual workflow and never for creative decisions or client conversations.

Who This Is For

Read if. Skip if.

READ IF YOU ARE…

You run an agency or sales team with several closers and want specific AI automations for lead follow-up, call prep and rep accountability.
You sell a high-ticket offer through applications and booked calls, and too many applicants never book a call.
You manage sales reps who ignore the script, skip research or follow up inconsistently, and you want a system instead of more yelling.

SKIP IF…

You sell low-ticket or self-serve products with no sales calls, so none of these call-based workflows apply.
You are a solo founder with no closers, since the leaderboards only pay off with a team of three or more.

TL;DR

The full version, fast.

A pay-per-lead agency with about 15 closers and 80 booked calls a day raised close rate and revenue per live call by 22% in eight weeks using five AI workflows. Instant multi-channel outreach lifted applicant-to-booked-call rate from 50% to 65%. AI-built pre-call research documents and personalized 12-week follow-up replaced work closers never did consistently. Two leaderboards, one scoring adherence to a Straight Line Selling checklist and one tracking common objections, moved script adherence from 55% to about 90%. The rule behind all of it: build the manual workflow first, then automate it, and keep AI away from creative decisions.

Free for members

Chat with this breakdown — free.

Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.

Create a free account →

Chapters

Where the time goes.

00:00 – 03:27

01 · Context and AI philosophy

The agency's size, why AI fails without a manual workflow first, and the rules for where AI is and is not used.

03:27 – 07:41

02 · 1A: Appointment setting for applications

Instant multi-channel outreach, enrichment and staggered LinkedIn touches lift application-to-booked-call rate from 30% to 65%.

07:41 – 08:35

03 · 1B: Calls booked sequence

Booked prospects get a selfie video from their closer, a one-page info sheet and a Loom explaining the offer.

08:35 – 10:30

04 · 2: Pre-call research

AI builds a one-page doc with business overview, client wins, case studies and five quick wins, then tells the prospect 30 minutes before the call.

10:30 – 12:42

05 · 3: Personalized follow-up

A 12-week automated sequence built from banks of case studies, activation points and FAQ answers matched to the call transcript.

12:42 – 15:02

06 · 4: Script adherence leaderboard

Every call is scored against a Straight Line Selling checklist; adherence rose from 55% to about 90% and tracks commission.

15:02 – 16:48

07 · 5: Objections leaderboard

Transcripts are mined for common objections, role-played in a daily meeting and fed back into marketing assets.

16:48 – 17:14

08 · The result

Close rate and revenue per live call up 22% company-wide in eight weeks, on top of an already top-tier in-person team.

Atomic Insights

Lines worth screenshotting.

AI only helps if you already know what good looks like and have a manual workflow; otherwise you just do the wrong things more efficiently.
Use AI to replace low-skill repetitive labor, then roll those salaries into fewer, far better A players.
Stacking thirty small 1% to 5% improvements that were never worth the cost before can transform a business once AI agents make them cheap.
Never use AI for original thinking, decisions or live client conversations unless the output is template-based.
The biggest leak in this agency was applicants who never booked a call, starting under 30% and reaching 65% only after AI-driven follow-up.
Hitting a new applicant with iMessage, WhatsApp, email and a LinkedIn request within seconds is the speed-to-lead play humans cannot match.
Randomizing every message and action slightly was what got the LinkedIn automation past anti-automation restrictions.
AI rewrites of 20 daily posts, each about 5% different and staggered, make blank setter LinkedIn profiles look legitimate.
A pre-call message saying the research is done and what the closer is bringing raises show rate and call quality.
Blocking bookings less than 24 hours out gives the automated pre-call sequence time to play out.
Follow-up emails personalized from the call transcript beat static FAQ emails because they answer the objections the prospect actually raised.
An activation point is a small tip the prospect can try in their own business so they feel the result before buying.
Scoring every call against a Straight Line Selling checklist moved script adherence from about 55% to about 90%.
Putting commission next to script adherence on one leaderboard shows reps a one-to-one link between following the script and earning more.
Manual script reviews stop working past about three closers; AI makes call review scale again.
Feed the top current objections back into ads, the VSL, landing page, email flows and the sales presentation to answer them before the call.

Takeaway

Automate the grind, keep humans on judgment

WHAT TO LEARN

AI lifts close rate when it takes over repetitive sales tasks that reps skip or cannot scale, layered on a manual process that already works.

01Context and AI philosophy

Learn what good looks like and build the manual workflow first, because AI on top of no process only does the wrong things faster.
Use AI for low-stakes repetitive work and move the saved salary into fewer, stronger people instead of cutting cost alone.

021A: Appointment setting for applications

Contact every new applicant instantly across iMessage, WhatsApp, email and LinkedIn, and randomize actions to avoid platform automation flags.
Enrich each lead and have setters call manually; this team found AI calling underperformed so humans still make the call.

031B: Calls booked sequence

Send booked prospects a closer selfie video, an info sheet and an offer explainer so they arrive warmed up.

042: Pre-call research

Generate a one-page prep doc with relevant client wins, case studies and quick wins, then tell the prospect it is ready 30 minutes before the call.
Require bookings at least 24 hours out so the automated pre-call sequence has time to run.

053: Personalized follow-up

Build banks of case studies, activation points and FAQ answers, and let AI match them to each prospect's application and call transcript.
Run follow-up for about 12 weeks on a fixed cadence instead of relying on closers to remember.

064: Script adherence leaderboard

Score every call against a written script checklist so adherence is measured, not guessed, and display it next to commission.

075: Objections leaderboard

Mine call transcripts for objections, role-play them daily and push the top objections into ads, the VSL, landing page and emails.

Glossary

Terms worth knowing.

Speed to lead: How quickly a new lead gets contacted after submitting a form. The faster the first touch, the more likely they respond.
Setter: A salesperson who contacts applicants and gets them booked onto a call with a closer, instead of closing the sale themselves.
Closer: The salesperson who runs the sales call and asks for the purchase.
VSL: Video sales letter. A video on a landing page that pitches the offer before the visitor applies or books.
Clay: A data enrichment tool that looks up extra information about a lead, such as their social profiles, from basic form details.
Straight Line Selling: Jordan Belfort's sales system built around steering every conversation back to a fixed path toward the close.
Script adherence: A score of how closely a rep followed the planned sales script and its order on a call.
Activation point: A tip in a follow-up email that the prospect can apply right away so they see the method work for themselves.
Pay-per-lead agency: An agency that charges clients for each lead it delivers instead of a flat retainer.

Resources

Things they pointed at.

05:35toolClay

08:07toolLoom

13:28productStraight Line Selling System (Jordan Belfort)

00:25productLead Generation Academy School Group

Quotables

Lines you could clip.

01:45

“If you don't know what you're doing and you don't really have this workflow, you're going to be at best doing the wrong things or doing useless things more efficiently.”

Tight warning against AI as a shiny object→ TikTok hook↗ Tweet quote

02:40

“I'm using it to replace low skilled labor that do kind of low stakes, repetitive tasks, and rolling their salaries into A players.”

Clear staffing philosophy in one line→ newsletter pull-quote↗ Tweet quote

14:32

“You can see a direct one-to-one correlation between commission paid and sales script adherence.”

Concrete proof point for following the script→ IG reel cold open↗ Tweet quote

The Script

Word for word.

Read-along

Don't just watch it. Burn it in.

See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.

analogy

Today I'm going to talk about how we used AI to increase our agency's close rate by 22%. And the most exciting part is that is 22 % on total calls booked, not even just live calls or shown calls. And we did it with basically five different workflows that are either fully automated with AI or kind of hybrid AI and hybrid manual.

So I'm just going to go and kind of talk about them one at a time. And then you'll be able to see if you can implement something like that in your agency or your business, big or small. I think all of these things have their place.

So let's get started. So if you don't know who I am, my name is Matthew Larson, and I run the Lead Generation Academy School Group, and I'm a partner in a pay -per -lead agency. So our pay -per -lead agency sells leads to 12 different verticals, heavily in financial services, insurance, lending, different kinds of medical, and so on.

And we'll have basically about 15 sales reps at any given time. And it usually is pretty steady around that number. People get hired, people leave, et cetera.

And we book about 80 calls per day. So 70 of those calls are through paid ads and 10 are from cold outreach and other sources. Majority of the time, like a high percentage of those 10 are from like manual targeted outreach and prospecting, not just like mass cold email.

So we also never really had a close rate or a revenue per live call problem. So this is a huge increase for us on top of an already very high performing sales team. And I think the problem with AI and why it seems just like a shiny object for most people is to me, there's three steps to doing anything.

You need to learn how to do it or just know what's good. You need to create the workflow. Everything has an underlying, very manual workflow, these steps in the process.

And only then can you automate it with AI or just any automation tool. And if you don't know what to do and you don't have the underlying workflow, there's almost no benefit to using AI because those things rely on you. If you don't know what you're doing and you don't really have this workflow, you're going to be at best doing the wrong things or doing useless things more efficiently.

Nowadays, especially, learning should be your number one priority. And then as far as my AI philosophy goes so far, I'm using it to replace low skilled labor that do kind of low stakes, repetitive tasks. I'm getting rid of all those people and then rolling their salaries into A players who are orders of magnitude better that will actually help us grow.

So let's just say I can lay off five people who make 4K each because their tasks have been automated and roll that salary into one 20K a month person, for example. I also love to use AI to do workflows that were kind of just impractical to do before. And those are usually a combination of manual and AI.

And I love to use AI to stack like 1%, 2%, 3%, 5 % improvements that kind of obvious and could be like a good idea, but they're just too minor and not really feasible from a cost benefit standpoint. But now with using AI agents, when you stack dozens of them, you know, you have 30. one to 5 % improvements, it completely transforms your business.

And one thing that I'm not doing is I'm never using AI for any kind of original or creative thinking, writing, making decisions, or talking to clients unless it's some sort of like template -based thing. So let's get into the first thing for what we did to increase our close rate. The first thing, and it's going to be part A and part B, first is part A appointment setting.

And this is to people who submitted an application, But they didn't book a call. And we have a five -minute delay in here to allow them to book a call before this automation triggers.

So in our agency, we've always had a very super solid show rate and close rate. We're very good at ads as that is our primary service and area of expertise. But historically, the drop -off rate between someone submitting an application and booking a call has always been the number one problem in our agency.

And when I first joined, because I'm not the founder of this agency and there's just so many people at it now, it was sub 30%. So think about that. If you got 10 applications, only three booked a call.

So this was always a very important step and focus for us. And we were already putting tremendous effort into this pre -AI, but we basically got it from 30 % to about 50 % manually over the past. two to three years and now with this ai workflow that i'm about to go over it's up to 65 so really orders of magnitude better for us so how it works is you know what you'd expect you drive the paid ads to the landing page they view the vsl landing page they fill out an application asking for name email phone number website and linkedin because linkedin is just the powerhouse platform and the vertical that this is about and then the qualification questions Of course, that application goes to the CRM and a spreadsheet.

So I basically, or we basically set up this AI workflow off the spreadsheet in parallel to the CRM, just so it doesn't break. And then they're matched to me, my partner, the appointment center that they've been assigned to, and the CRO, which is our chief revenue officer. So basically what happens is now they instantly get an iMessage, a WhatsApp, an email, and a LinkedIn connection request from the setter.

And that's all done with AI. So like instantaneously, simultaneously done. Then in the spreadsheet, they instantly get enriched in clay to try and find their Twitter, their Instagram, their Facebook, their YouTube, et cetera, whatever we can do to get more information on them.

Then the setter calls them. The setter still does this 100 % manually. We've tried to use AI for this.

Hasn't really done well for us. And then if clay or any other enrichment tool has found additional information, The setter also follows them from Instagram, Twitter, and messages them there.

And that's done with AI. And then two hours later, they get a connection request on LinkedIn from the chief revenue officer's LinkedIn page. And then a message.

And that's done by AI. 22 hours later, they get a connection request from my partner on LinkedIn, AI. And then seven days later, they get a connection request from me.

And then someone at our company, not me. but they basically create 20 LinkedIn posts per day, which are a mix of just kind of like short posts, long form carousels, like you name it, insights, all that kind of stuff. And we have these in basically a, just like a big, well, it's in a database at this point, but just pretend that these are in a sheet, like a Google sheet.

And then what we do is we use AI to rewrite them. So they're like 5 % different. And then we stagger them.

And then we use AI to post. content on every setter's LinkedIn profile so that they look way more legit than just some blank profile. And every setter has their own follow -up process that they do manually if they don't apply.

And every message here and action is slightly randomized. As we've had, especially when we first started this, a lot of troubles on LinkedIn with their anti -AI, anti -automation detection and getting profiles restricted. But we largely got past that.

And I think like this workflow is likely impossible without AI with the mixed constraints of kind of maintaining proper speed to lead for the setter, the high cost of writing and posting original content for every setter, and then just managing all the LinkedIn accounts. And then if we basically go to part B, so like if they actually booked a call.

It's largely the same, but a bit different. So they book a call, they're still matched, but instead of the setter, they would get an iMessage, a WhatsApp, an email, and a LinkedIn connection request from the closer that they're matched to instead of the setter. And that would send them a selfie video of the closer.

So we have a script where you send like an iMessage video and every single one is like, we just have every, all the 15 closers have filmed it and then it's just like categorized. So it sends them that video. And then a one -page screenshot of all the information they need to know.

And then a loom with a fifth 13 -minute video of me, actually. I'm in this loom explaining the offer and the unique mechanism. And that's all done by AI.

And kind of the rest of that workflow is the same as what I went over before. So I don't really need to talk about that much. The next is pre -call research.

And it is a complete grind to get sales reps and closers to do any form of research before the call. If you do like any kind of sales call, putting effort and preparation into it will make you seem 100 times more professional. And you'll also come across better because you'll be more confident because you're prepared.

So we use AI to do almost all of this now, but the closer will still spend, allegedly anyway, 10 to 15 minutes before the call confirming the accuracy and reading it. So AI does 100 % of this. So we have a one -page Google Doc of...

you know, in preparation for the call. It goes over their business overview. It has their answers to their application questions.

It goes into basically we have a spreadsheet of every client win in like kind of like a screenshot and it pulls the five most relevant ones. And then it also does the same in a different spreadsheet with the three most relevant case studies and it puts them in the document. Then it will basically generate five quick wins for each prospect based on their application.

answers, their website, their profiles, et cetera. And it pulls these from kind of like a bank of 20 potential quick wins that are like pre -made and it makes it so it's very personalized to them. And then 30 minutes before the call, because we actually don't let people in this particular sales process book a call less than 24 hours in the future, just because we want this process to play out.

30 minutes before the call, it sends them an SMS, a WhatsApp, a LinkedIn, and an email telling the person that the closer, just finished up the research process and is bringing the quick wins, the client case studies and all of that stuff to the call. It's like, yo, Matt, like I just finished the research process for company name.

So excited to talk to you in 30 minutes. I'm bringing X, Y, and Z to the call and really, really strong as far as an increase in show rate and close rate and just the quality of the calls altogether. So like.

We should also note that they are receiving an automated pre -call flow during this process too. We also want to note that this probably is not possible without using AI as closers simply like don't have enough time to do all of this in between their already scheduled calls and all that kind of stuff. So you really do need AI for this workflow.

The third is personalized follow -up, but like ultra personalized. And following up is arguably the most important part of sales, but it's very similar to pre -call research. It's a very big grind to get your sales reps to do it like 100 % accurately.

So now all of the follow -up is entirely automated and our sales team does not even do a single action of this anymore. We don't call in our follow -up process. We very much have a one call close and they buy it or they don't.

So that could be some important context. But every closer has... a follow -up system that's connected to their personal WhatsApp, iMessage, email, LinkedIn, and then to our internal operating system and our CRM and all that kind of stuff.

And I basically, long time ago, but written out a bunch of follow -up templates. And we just have a framework that you can see on the screen. But these templates connect to a bank of FAQ answers, a bank of case studies, and a bank of what we'll call activation points.

And then AI personalizes the emails in the context of the framework and then matches them also based on their application answers and the call recording transcript. So the first day is the best case study. Second day is the most relevant case study to them based on their application answers and the transcript.

The third is the best activation point. So it's basically like an activation point is going to be a tip that like, okay, do this in your business. And then they do it and they say, wow, like that actually works fine.

I'm going to hire them. type thing. That's kind of what I'm referencing here.

And then day five, seven, and nine is the answer to the biggest FAQ from their transcripts. Like, well, the three biggest FAQs are the three most important, and it pulls directly from the bank of the FAQ answers and puts them in an email. Previously, these were just static.

Like, this is the three biggest ones in general, regardless if they really had those ones or not. Then we alternate for 12 weeks, Monday case study, Thursday activation point. So we basically follow up for a little over 12 weeks.

And I just, again, don't think that this was possible pre -AI as closers would not have enough time to do all this in addition to their scheduled calls during the day. And then the next one is a script adherence leaderboard. And another huge entry in the things that are a grind department is getting sales reps to adhere to your sales process consistently.

Your sales script or plan exists because it represents the most likely path to closing the deal. But so many sales reps, and this increases the more you have, just will do whatever they want. And the primary purpose of most sales managers is to watch them and scream at them when they do whatever they want.

For more context, we use Jordan Belfort's straight line selling system as the primary foundation for our sales process. So we just have kind of like a checklist. If our sales reps are trained that, you know, if the conversation goes off the checklist, off the straight line, so to speak, you answer it or you do whatever, and you try to get them back to the straight line as quickly as possible.

So we have this straight line selling system checklist, and then we use AI to analyze every single call for how well these sales reps stuck to the plan in the particular order, i .e. adherence. And then we have a leaderboard that shows this.

So two things happen because of this. It holds them accountable. And we've seen an increase of 55 % adherence to the scripts to about 90%, which is orders of magnitude better in the grand scheme of things.

And the close rate has really increased because of that. And then what we have is we have a leaderboard that shows sales team statistics. It's two views on one leaderboard.

So on the left -hand side, it shows like the typical, the close rate, the revenue per live call, the commissions, total calls, et cetera, like the things you would expect on a sales leaderboard. And on the right -hand side, it shows stats about how well they stuck to the script. And then when you look at this thing, you can see a direct one -to -one correlation between commission paid and sales script adherence.

It's just an incredibly powerful example and form of proof to the sales team that it works. And across all 15, there's literally a 100 % without a single deviation, total commission versus adherence to the sales script. And we had this workflow pre -AI.

It kind of works a little bit if you have one closer or two closer or three closers, but then it really becomes impractical without having a full -time person reviewing every sales call. And now it's back. And then the final one is kind of similar, but not.

It's the objections leaderboard. And quite frankly, we've been in this business at a large scale for many years. So there's not really too many surprises here, but things do vary from time to time as the market shifts.

And this leaderboard basically shows like every transcript is analyzed and added to this. It shows basically the most common objections from every sales call in real time across the team so that people are able to share and like be aware and prepare for these objections. And well, to be honest, we only had this for a month, but it'll just show them every single month or like rolling 30 days.

And we have a daily sales meeting across all of our team with our CRO handles that. where basically they role play and they share and they find the best way to overcome these objections so everybody's prepared. And then the top objections based on the current time are rolled back into the marketing and pre -call assets to try to preemptively overcome and address them.

So we're going to add them back to retargeting ads, edit them earlier in the VSL, put them higher on the landing page, earlier in our set email flows, earlier in the sales presentation, and then come up with better ways for our objection handling checklist. And similar to the last one, we had this workflow pre -AI when we had less closers, but it just becomes impractical after about three.

And now it is basically back. So the results, and like all things considered, our close rate and basically revenue per live column, because we sell one thing at a certain price, so it's identical in our case, have gone up 22 % company -wide in the past eight weeks. And this is already on top of a top 1 % or better performing in -person sales team.

And they're all in an actual call center, not even remote. So it's even pretty crazy that this is the case. So that is just how we're using AI to increase close rate in our agency.

I hope it was useful, and I'll see you in the next one.

The Hook

The bait, then the rug-pull.

A 22% lift in close rate is a lot for a team that already closes well. This breakdown covers the five AI workflows that did it at an agency booking about 80 calls a day for 15 closers.

Frameworks