Growth Lessons from the AI Heat Wave
What we learned about growth engines in the age of AI from a rockstar B2B company 3x'ing ARR at 180% NRR.
Good morning, and welcome to a special edition of Growth Stack Mafia.
Todayâs post is co-written with my colleague Brett Browman. Over the last few weeks the two of us went down the rabbit hole with one of our companies that is, to use the technical term, crushing it.
Theyâre a B2B business 3xâing ARR with net revenue retention around 180%, and theyâve made the hard transition most technical companies never do: from early, developer-led traction into a real multi-channel growth engine.
Weâre keeping the company anonymous - theyâre one of ours, and a lot of this is their live playbook - but hereâs what matters for you: they sell a deeply technical, developer-first product, and their buyer is about as skeptical as they come. Engineers and technical leaders, people with finely tuned BS detectors who tune out anything that smells like marketing.
Weâre writing this in the middle of an AI heat wave. Every deck has an AI slide, every founder has an AI shortcut, and half the growth advice on your timeline is some variation of âpoint a model at it.â
So itâs worth saying up front what this teardown is not. This company is not winning because of a clever AI trick.
Theyâre winning because they do the unsexy fundamentals extremely well, and then they use AI to remove drag in the exact places it helps.
Before we start a quick thank you to this drops sponsor - Passionfroot!
Every sponsorship I run, the Growth Stack Mafia newsletter and my LinkedIn posts, goes through one piece of software called Passionfroot. Today they announced a Series A led by Insight Partners. Congrats to Jennifer and the whole team.
Iâve been a creator on the platform for a while, so this is an easy one to write.
Before Passionfroot my âsponsorship systemâ was a hot mess.
Iâd get random linkedin messages, Iâd have emails, and apple notes. Iâd have to create Quickbooks invoices, and even with my basement AI servant Gary, it was not working. I just stopped doing them effectively in March because it was too much work.
What sucked about this is a big part of my JOB is to know about and use tools and then talk about them. I just didnât want to deal with all the fluff of getting paid to do it.
Passionfroot has been pretty rad because it consolidates the storefront, the brief, the contract, the schedule and the payout in one place.
AND THEY HAVE AN MCP??? Yes they do [ wink wink wink iykyk ]
Most of the companies I talk to are trying to figure out ways to scale creator marketing, and passionfroot is the first real answer Iâve found.
With that ⊠Hereâs the playbook.
1. Know exactly who youâre selling to
Start with your data before you build a growth engine.
Before you pour money into channels, figure out who youâre actually winning with and why: the segment where you close fastest, retain best, and expand most. They assumed their sweet spot was scrappy AI startups.
When they actually looked at the data, it was large enterprise. That one correction reshaped their entire ICP, and everything downstream got easier - messaging, channel mix, sales motion - once they were aiming at the right buyer.
You cannot out-execute pointing the engine at the wrong person.
Score accounts on real signals, not vibes.
Once you know who you want, stop treating every account the same.
They score accounts on actual buying signals: website activity, documentation consumption, product signups, engineering engagement, and company fit.
BDR and AE effort concentrates where intent already exists instead of on cold ground. On the inbound side, leads are automatically segmented by company size, traffic, industry, geography, and ICP fit, then routed across SMB, Mid-Market, Enterprise, and EMEA teams. Faster response times, balanced workload, and higher conversion, because the right lead gets to the right rep quickly.
2. Manufacture real relationships at scale
Cold outreach into a skeptical technical buyer is a losing game. Their second insight is that almost every channel they run exists to manufacture a warm relationship before a sales conversation ever starts.
Now youâll know from reading my substack that I hate the concept of âwarm outboundâ largely because itâs done poorly. People are wrapping up cold outbound with ai signals and calling it warm. Itâs not warm just because you know somebodyâs DMA.
What makes it warm is by doing deep research, being specific, strategic, and trying to connect with people over time. Often this means asking for nothing. ESPECIALLY NOT A â15 min call to tell you to pick your brain!â
Here is the real sauce.
Run a podcast that is secretly a lead gen tool.
Donât try to build a popular podcast. The download count is irrelevant. Use it to get your exact ICP persona - the product, engineering, and data leaders you sell to - on a call to talk shop about best practices and how they build.
The format flatters them with a feature, builds a genuine executive relationship, and doubles as discovery. The AE walks away understanding the prospectâs business and has a natural reason to follow up and convert to a demo. Investors rolled their eyes at this one. It works.
(Note: This also works for dinners, and doing pre-dinner interviews to talk shop, not about your product - more on this below).
Treat webinars as a pipeline engine and an upsell engine.
Invite prospects to monthly product and engineering leadership webinars. Attendees who engage get BDR follow-up to see whether a discovery call makes sense. The same motion converts free and open-source users to paid: hire customer success people to actively expand usage, and use the webinars to surface features people didnât know existed.
(At one point a customer team had built a custom workaround for something the company had shipped a year earlier - they just never knew it existed.)
Pair that with gating the right features to create natural upsell moments, and you get net revenue retention around 180%.
In-person is massively underrated.
Dinners that deliberately mix existing customers with prospects work absurdly well, because the happy customer sells for you in the room.
Conferences and field events become warm entry points into target accounts: targeted outreach before a major industry event, then invitation-only executive dinners around it.
The founderâs move is to email a big target saying âIâll be in your area next week,â and only book the flight once they say yes.
At $60-100K ACV, a $1K flight is a rounding error. This is the most underpriced channel in B2B right now precisely because it doesnât scale cleanly, so most teams skip it.
PS remember, human relationships are ultimately at the heart of almost all business, so it shouldnât be surprising that this works.
3. Build a personal brand engine, not a logo
Post from your face, not your logo.
For a technical audience, a founderâs personal account beats the company page every time.
A person can take a hotter take and show real personality.
A corporate page canât, and engineers scroll right past it. The whole thing lives or dies on authenticity.
Their founder blocks 2-3 hours a week to write, works with an agency for a monthly brainstorm and filming session, and personally reviews everything before it ships. Do not post AI slop to this audience. It does more damage than posting nothing.
Lower the quality bar and just ship.
One YC founder went from zero to 100K followers on X by posting constantly and deleting whatever didnât land.
Volume plus fast pruning beat careful curation. Stop overthinking novelty: your audience spans a huge range of sophistication, and âbasicâ content routinely outperforms the clever stuff you agonized over.
If you arenât putting reps in, you never learn what actually resonates. Layer more deliberate thought leadership on top to position yourselves as category leaders (which also feeds answer-engine optimization, since the models increasingly cite the people who show up consistently). Keep it authentic, always.
4. Run marketing like a portfolio
Budget backwards from the deal.
For any big spend - say a $50K conference - start from the outcome: how many deals do we need to close for this to pay off?
If the answer is one or two, itâs probably worth doing.
â Then treat the entire marketing budget as a portfolio: try things, track whatâs working, cut what isnât, and keep reevaluating instead of setting it and forgetting it. For them, LinkedIn and Google ads paid off while X and Reddit did not, so they hold a consistent presence primarily on LinkedIn to drive awareness inside target accounts and generate demo requests from qualified buyers.
5. Build the team and the tooling right
Hire sales reps in pairs.
A single hire gives you no baseline. You canât tell whether the numbers reflect the rep or the territory. Hire two at once and you get a comparison point, which matters even more because strong inbound can mask a weak rep for a long time.
To get real signal before you hire, use structured mock interviews with a distinct pitching phase and a negotiation phase rather than relying on a resume and a good vibe.
Point AI at seller productivity, not seller replacement.
This is where the AI actually makes a grand appearance, and itâs deliberately boring. They lean on AI across the sales workflow: account research, pre-call prep, automated presentation generation, MEDDIC capture, and general workflow automation.
The payoff is less administrative drag and a single centralized view that pulls together HubSpot and Notion, so sellers spend more time selling and less time on data entry.
AI didnât replace the rep. It deleted the parts of the job the rep hated.
6. The channel stack, at a glance
Put the whole thing together and itâs a system, not a list. Every top-of-funnel channel exists to create a warm relationship; scoring and routing decide where the effort goes; and expansion (CS plus webinars) turns closed customers into the 180% NRR that makes the whole model work.
Hereâs the channel breakdown in their own words:
Webinar-led pipeline generation: monthly product and engineering leadership webinars; engaged attendees get BDR follow-up toward a discovery call.
Podcast-led prospecting: invite product, engineering, and data leaders onto the podcast; build authentic executive relationships and a natural path for AE follow-up.
Engagement-based prioritization: score accounts on buying signals (website activity, docs consumption, signups, engineering engagement, company fit) and prioritize the high-intent ones.
Conference and event-driven outreach: conferences, executive dinners, and field events as warm entry points into target accounts.
Inbound scoring and routing: segment inbound by size, traffic, industry, geography, and ICP fit; route across SMB, Mid-Market, Enterprise, and EMEA.
AI-assisted sales operations: AI for account research, pre-call prep, presentation generation, MEDDIC capture, and workflow automation; centralizes HubSpot and Notion into the sellerâs flow.
Thought leadership: authentic content to establish category leadership and feed AEO.
Digital advertising: a consistent, mostly-LinkedIn presence to drive awareness in target accounts and pull demo requests from qualified buyers.
The meta-lesson
Itâs an AI heat wave, and the temptation is to follow all the Linkedin Hype out there, buying tools and vibe coding your way to some fake growth paradise. But the company we watched actually pull this off didnât win with a magic AI play.
They won by knowing exactly who they sell to, manufacturing warm relationships at scale, shipping reps until they learned what lands, and then pointing AI at the boring parts of the job.
Said this for a while: AI is helping solve the boring parts of work that nobody wanted to do anyways.
None of this should feel new or surprising. In a summer where everyone is chasing the flashiest AI story, the boring fundamentals - done with real discipline - are what count in an era of people chasing easy wins.
Austin & Brett
(PS if you got this far, I am not surprised if you arenât AI because this is some swweeettt sweet growth gravy.)
Additional Reading
Elena Verna - 9 ways growth is different in AI companies - why founder LinkedIn posts out-perform paid, and why growth in AI companies is about big bets over optimization.
Kyle Poyar - The compounding startup - how retention and NRR compound as you scale, the math behind a number like 180%.
Lenny Rachitsky - The ultimate guide to founder-led sales (with Jen Abel) - the canonical playbook for founders selling in the early innings.
Harsh Truths of Sub-$10M ARR Marketing - the GSM companion piece on where early-stage marketing dollars actually go.
 | A guest post by
|