27 Real Ways Founders Can Use AI Today
Founder Therapy x The Signal
If you were forwarded this newsletter, join 8,604 weekly readers—some of the smartest GTM operators & founders—by subscribing here:
Presented by Clay.
Bring AI agents, enrichment, and intent data together and turn insights into relevant, timely action. Start building for free.
& Supported by:
Attention → AI Agents for RevOps
Clarify → The Autonomous CRM
Nooks → Agent Workspace for Intelligent Outbound
Sumble → Buying signals from tech stacks, people data, & active projects
11x.ai → Accelerate growth with a digital GTM workforce
Hey, y’all! 👋
Today’s piece is a collaboration with Patrick Thompson, who writes Founder Therapy and is the Co-founder & CEO of Clarify. More on his newsletter at the end.
As a founder, you’re under real pressure to set the pace on AI inside your company (and scale with AI & automation instead of headcount). Your team watches how you use it. Your board asks what you’ve automated. And it’s hard to tell a real workflow from demoware on LinkedIn.
The most AI-pilled companies I see have a very AI-pilled CEO. The pace starts with the founders.
So here are 27 specific ways founders can use AI today, across:
Revenue and customer-facing
Marketing and content
Product and the feedback loop
Running the company
Hiring
Fundraising
Alright, let’s get into it.
Revenue and customer-facing
1. Build your ICP from your calls.
Feed an agent your prospect and customer call transcripts and have it build a specific ICP and buyer-persona sheet from what buyers actually say, not what you guessed at an offsite.
2. Signal-based outbound, founder-led.
An agent watches your ICP for the moments that predict a deal (social listening, a key hire, a competitor in their stack, a telling job post), then hands you a short list + a pre-drafted email to review. You still do founder-led outreach, but the signal-watching agent runs the research for you. (More ideas here: the best AI-native GTM plays you’re not running.)
3. Champion tracking.
Your best next customer is the one who already knows and loves your product. This is especially true in the early days. So, when a past champion changes jobs, an agent flags it and drafts the note to reach them in their new seat (this play will often lead to your fastest new logos).
4. Lookalikes of your best customers.
An agent builds a target list that mirrors your closed-won accounts, so you start from who already works, not a search using finger-in-the-air firmographic filters.
5. Warm intros.
Point an agent at your collective network (LinkedIn, Gmail, calendar, plus investors, advisors, and team), have it find who can introduce you into a target account, and draft the ask.
[Pro tip: export your LinkedIn messages and have Claude score them for sentiment, and wire up the Google APIs for calendar and email. I’ve done both, and it’s magic.]
6. A pre-call brief on every inbound.
The moment someone books a demo, an agent turns their funding, stack, recent news, and anything else relevant for you, into a simple brief, so nobody is scrambling to research the account three minutes before the call starts (or three minutes into the call… we’ve all been there).
7. Post-meeting follow-up that moves the deal.
After a call, an agent drafts the recap and the explicit next step toward closing, sent within the hour instead of three days later.
8. Closed-lost, analyzed and reopened.
Monthly, an agent clusters why deals actually died and drafts the re-engagement note when something changes.
9. A battlecard that updates itself.
An agent watches competitors’ pricing, changelogs, and G2 reviews and keeps your battlecard current instead of stale since March.
(The person who builds and owns most of these is usually a GTM Engineer, and most teams scope that role too narrowly. Here’s why. But in the early days, the founders should personally be building these agents.)
Marketing and content
10. Build your founder brand (idea-miner and editor, not ghostwriter).
If your buyers live on LinkedIn or X, use AI to mine post ideas from call transcripts and internal docs, and as a custom editor on your drafts (I wouldn’t let it ghostwrite yet). The same setup turns one post, talk, or call into a week of content in your voice. (More here: How to build a founder brand with the Rick Rubin of LinkedIn, Alec Paul.)
11. Personalized assets and decks.
Use Claude Design, Gamma, etc to spin up tailored sales-demo decks, marketing and event assets, even investor decks and internal memos.
12. Case studies from your first customers.
The moment you have a couple of paying customers, interview them and have an agent turn the call into a personalized case study with pull quotes. If you want to go deeper here, my friends Noah and Adrian recently published a great walkthrough: Building a Social Proof Engine.
13. Partner activation on autopilot.
When a partner goes live, an agent builds their landing page and drafts the welcome email with the link and promo code. (Patrick’s team runs this inside Clarify, on Sanity and Stripe.)
Product and the feedback loop
14. Feedback into a ranked roadmap.
Every ticket, sales objection, and churn note gets tagged by feature request and ranked, so the roadmap follows aggregated data, not the loudest customer.
15. A PRD from a voice memo.
Talk an idea through on a walk, and an agent returns a structured spec with edge cases and open questions. (Capture it with something like Superwhisper and it’s close to free.)
16. Support triage and first-draft answers.
Conversations sync from your help desk hourly, are deduped, and an agent drafts answers from your docs for a human to approve.
Running the company
17. The morning brief.
A daily brief across your inbox, calendar, CRM, and Slack, so you show up prepared by default.
18. A daily “what did the agents do” report.
Each morning, a Slack summary of the last 24 hours of agent runs and time saved. Visible, repeated wins are what make AI adoption actually stick on a team. (As you scale agents, this step will become more important.)
19. Decision memos on demand.
Talk through a hard call and get back a structured memo with the options, the tradeoffs, and a recommendation to react to.
20. Board updates and the metrics narrative.
An agent assembles the monthly board update from your metrics, CRM, and product data, and you query the numbers in plain language (”why did net revenue dip?”) instead of clicking dashboards.
Hiring
21. Sourcing, not just screening.
Talk to Claude about the exact person you want (eg: the companies they came from, their current role) and have it run deep research to source specific candidates. You can do this instead of using a recruiter or to augment a recruiter’s efforts.
22. A work sample that allows requires AI.
Let candidates use AI in the coding round, and its equivalent for every role. You’re hiring for real problem-solving in 2026, not for who can pretend ChatGPT doesn’t exist. Hire AI-natives.
23. Role briefs and screening, drafted.
Draft the internal role brief that defines what great looks like, then have an agent screen inbound applicants against it for consistency.
Fundraising
(Note: Treat fundraising as sales and these map cleanly onto a sales workflow.)
24. Record the calls, sharpen the pitch.
After each investor call, an agent recaps what landed and drafts a tighter version of the pitch for the next one.
25. Cluster objections into a battlecard.
An agent groups objections across calls (valuation, market, traction, team, competition, moat) and maintains a living battlecard with your best responses and proof points.
26. Run the raise like a sales pipeline.
A lightweight pipeline of targets, status, and next steps, with a weekly summary of what you’re learning, so your pitch and process improve as you go.
27. Prep and follow up for every meeting.
Before the meeting: get info on current portcos, where the company is investing, the partner’s thesis, tailored talking points and questions. After the meeting: a follow-up draft with the materials and deadlines they asked for.
Where to start
The founders pulling ahead have moved a dozen of these off their plate, so the company runs when you’re heads down, on a flight, or out for a week.
Don’t build all 27. Start with three: the workflow you hate most, the one that always gets skipped, and the one closest to revenue. Watch them for a week, then add three more.
Once data capture stops being a human job, the interesting questions become: what is the human best used for, and what does your software even look like when agents do more of the operating? I think that reshapes the tools we live in, starting with the CRM. More on that in a post that will be published later this month (“How SaaS UI is evolving”).
If you got something out of today’s piece, half the credit is Patrick’s. He writes Founder Therapy, a weekly, honest read on what building a company actually feels like. It’s a phenomenal Substack (I read it). If you’re a founder, subscribe here:
Thank you for your attention and trust. I do not take it for granted.
See you next time,
Brendan 🫡
 | A guest post by
|