What AI GTM leadership actually means

I spend most weeks working inside portfolio companies, and recently I’ve been thinking about one question more than any other.

Will the GTM model these companies have today still be viable in two or three years?

AI is obviously part of that. But I don’t think the answer starts with which AI tools a company is using. It starts much earlier than that. Does management really understand what is working? Is the company focused on the right customers? Does it know why those customers buy? Can management tell whether the pipeline is getting better or worse? And when something isn’t working, how quickly does the leadership team make a decision and move forward?

I had a series of conversations across portfolio companies this week that made me think about this in much more practical terms.

The first was a GTM reporting session. The company had started changing its approach, so I asked whether the board properly understood what was changing and what that meant for the numbers they would see. I see this quite a lot when companies change their GTM motion. The company starts doing something different, but the board continues looking at the measures that made sense under the old approach.

The classic example is MQLs. A company deliberately moves away from a volume-led approach, MQLs drop off a cliff, and somebody at the board meeting starts digging into why MQLs are down. The problem isn’t necessarily the drop in MQLs. The problem is that management hasn’t clearly explained what has changed, what should happen instead, and how everyone should now judge whether the new approach is working.

That matters to an investor because changing a GTM motion creates a period where the old indicators become less useful before the new model has enough data behind it. If management doesn’t handle that transition well, the board can spend months asking the wrong questions. Worse, management can start changing direction again because the old metrics look bad, even though those metrics no longer reflect what the company is trying to achieve.

We got into pipeline visibility in another conversation that week, and I realized I was pushing on a similar problem. I said that I had never really been able to answer a very basic question from the information I was seeing: “Are things going well? Are they flat? Are they going bad?”

For a post-investment company, I think management should be able to answer that question quickly. If the investment case assumes growth accelerates, the board needs to understand whether pipeline creation is increasing, whether the important opportunities are progressing, and whether the new GTM investments are actually changing the trajectory. A large pipeline number in a CRM doesn’t tell you that. Neither does a beautifully produced board dashboard if nobody can explain what is changing underneath it.

This is one place where I think AI should materially raise the standard for GTM leadership. Companies now can analyze CRM data, call transcripts, customer conversations, and account activity much faster than before. I’m less interested in whether a company has added AI to its sales stack than whether the leadership team can use all of that information to understand the business better. If I’m still asking “Are things going well, flat, or bad?” three months later, the technology hasn’t solved very much.

A second conversation was an ICP session. We spent a lot of time getting much more specific about where the company should focus. My view was to keep the ICP as slim as possible—ideally one. Sometimes there are genuinely two or three, and that’s fine, but I wouldn’t want more than three at the absolute maximum for a company at this stage.

We then got into the information that sits underneath the ICP. What actually triggers the purchase? Why did the customer choose you? Why did they buy now? I suggested embedding “Why us?” and “Why now?” into the process every time a new customer signs and recording the answers. If you continually do that across new deals, marketing and sales start getting a real feedback loop from customers rather than relying on the internal view of why the company thinks it wins.

I think this becomes much more important in an AI GTM model because the cost of going broad is disappearing. A company can now research thousands of accounts, create enormous amounts of content, and personalize outbound at a scale that would previously have required a much bigger team. That sounds like an advantage, but it creates a new risk. A company with a vague ICP can now waste time across a much larger market, much faster.

I’d therefore look for the opposite in a portfolio company. Can the leadership team describe its best customers precisely? Does it know the situations that cause those customers to enter a buying process? Does it know why it wins? And is that information actually changing where marketing spends money and where sales spends time? If those answers are weak, giving the company more AI capability probably increases activity before it increases effectiveness.

The third conversation was with a sales leader who had only been in the role for a few weeks. The company is post-investment and going through the normal transition that comes with trying to scale. As I said to him on the call, post-investment and scaling is always a transition phase. Things take time to settle, organizations are always in some degree of flux, and at some point you have to ride the wave, focus, and get on with what you need to do.

So we focused on the immediate challenge around building pipeline. We went through where the gaps were, where marketing could better support sales, and what needed to happen next. We agreed to catch up again after an important customer meeting because that would give us more information to work from rather than trying to solve everything in one conversation.

I think there’s an investor point in that which gets missed. Post-investment companies often add people, tools, and activity because the plan requires more growth. But adding capacity doesn’t automatically create a scalable GTM model. A new sales leader still needs to work out where pipeline will actually come from, which deals matter, where the gaps are, and what support they need from the rest of the organization.

AI makes it possible to add even more apparent capacity without adding people. That makes the distinction between activity and progress more important, not less. I’d want to know whether the extra output is creating a better pipeline, moving important opportunities, or shortening the time it takes the company to learn what works. Otherwise the company may have found a cheaper way to produce activity.

The fourth conversation was with a marketing leader I work with every week. I had added some questions to his weekly document because one thing I’ve noticed in high-performing CMOs and CROs is good decision velocity. I see the same thing when I look across portfolio companies.

When I go back into companies that are performing better, there has normally been a series of decisions made since we last spoke. Whether every decision was perfect isn’t really the important thing. They made decisions, saw what happened, and moved forward. The companies that concern me more are the ones where I come back months later, and we’re still discussing the same problems.

An old CEO of mine ingrained this into me when I was a CMO. He basically said that if I were missing the plan, we would meet every week until I hit it. If I were hitting the plan, we would meet once a quarter. And when we met, he asked the same questions: what got done, what worked, what didn’t work, and how do we know what worked?

I still think those are useful questions for investors because they tell you something that a KPI dashboard often doesn’t. Is this leadership team learning? If something didn’t work last month, what did they change? If something did work, what did they decide to do more of? If the company missed the plan, what decision followed from that information?

AI should compress that loop considerably. Research that took days can happen in minutes. Customer calls can be analysed automatically. Data can be interrogated without waiting for someone to build another report. Leaders can get from a question to useful information much faster. If the company can produce the analysis in an hour but still takes six weeks to make the decision, the constraint isn’t AI. It’s leadership.

That was also part of a broader conversation I had this week about what GTM leaders now need to learn themselves. We talked about using AI to build decks, write reports and improve documentation, but the bigger point was much simpler: this can’t sit as some future project you get to when you have more time.

I said on the call that “I don’t have the time to learn this stuff” is the wrong question. The right question is: when do I learn it and embed it into my daily routine?

That matters because AI GTM leadership isn’t really about having access to the tools. Everyone will have access to the tools.

It’s about whether the people running GTM actually change how they work because those tools now exist. Put those conversations together, and I think the investor question around AI becomes much more interesting.

I wouldn’t spend too much time asking portfolio companies for their AI strategy. I’d look at how their GTM leaders actually operate. Can they explain what is working and what isn’t? Can they tell you why customers buy? Have they made real choices about where to focus? Can they tell whether the pipeline is improving? When they learn something, do they make a decision? And are the people running sales and marketing actively changing how they do their own jobs as AI gives them new capabilities?

That, to me, is what AI GTM leadership actually means.

The fundamentals haven’t disappeared. If anything, AI makes them more important because companies can now execute bad GTM decisions at much greater speed and scale. A broad ICP can become thousands of badly targeted personalized emails. Poor customer understanding can become hundreds of pieces of irrelevant content. Weak management information can become a very impressive automated dashboard. Slow decision-making can survive even when everything around the decision now happens ten times faster.

So the risk I’d be looking for across a portfolio isn’t simply whether companies are adopting AI quickly enough.

I’d look for companies where AI is increasing output, but management quality isn’t increasing with it. Then I’d look for the opposite: leadership teams that know where to focus, understand what is actually working, make decisions quickly, and are learning how to use AI inside the work they already do.

I think those companies have a much better chance of building a GTM model that still works three years from now.

That’s all for this week.

See you next Saturday.

Cheers,

Edwin & Josh

Demand Karma runs structured GTM roadmap builds for investor-backed B2B SaaS and AI portfolio companies. We produce fast, board-ready assessments that align investors, CEOs, and GTM leaders around the facts, then stay to build the leadership and operating capability the transition requires.

We work across the critical GTM failure points: maturity before you commit budget, leadership readiness across the team, and diligence before or after the deal.

→ AI GTM Acceleration Plan maps your AI GTM maturity across Market, Engine, and Organization, and hands you a prioritized roadmap of the highest-value use cases you aren’t running yet. Days, not months.

→ AI GTM Leadership Kickstart gets your GTM leadership team aligned on what’s changed, what still works, and what the new rules of GTM leadership demand, applied to your live challenges.

→ GTM Due Diligence is an investment-grade audit of every function and every pound of spend, at Premium or Red Flag depth.

This is GTM risk, not technical due diligence. We audit the whole engine because disruption enters through marketing and exits through retention: marketing, sales, RevOps, and customer success. Auditing one station and calling it a GTM review is how risk stays unpriced.

After the readout, we stay alongside the transition: the AI GTM Leadership Kickstart to start the change as a team, one-to-one Leadership Development to build the leaders, and Private Board Advisory to keep the board confident the change is holding.

Know the risk. Build the operators. Then win the decade.
demandkarma.com