Meraki and Samsara Sales Veteran Andy McCall: Most AI Startups Pick the Wrong Sales Strategy

McCall's central tactical point is that the two playbooks are phases, not identities. Every small company wants to grow, and almost every large successful company has deployed both strategies at some point.

"Every small company wants to become a big company. I don't know too many very large successful companies that at some point in time haven't deployed both strategies," he said. "You might start off with land grab but then you mature and you have a lighthouse strategy, or you start with lighthouse and then you get big enough that you can go broad into land grab."

The transition happened at both Meraki and Samsara when they verticalized. At Meraki, the first lighthouse pull was public sector — specifically school districts, which all talk to each other. Winning the biggest district in a state converted every district underneath it: "Oh, what did that one buy? Great." At Samsara, the same logic applied when entering cities, counties, and states.

The reason to build a formal lighthouse motion at that point is structural, not stylistic. Sales cycles, decision makers, and procurement processes all differ, and asking the same team to sell to both markets is asking for a different job. The trigger is when a vertical's unit economics and buyer dynamics justify dedicated focus — then you build a list of the top fifteen accounts in finance, transportation, or the public sector, and take them down one by one. Applied Intuition is the exception that proves the rule: its buyer set never grows, so it remains a flagship-account business for life.

If the framework is the hard part, it shouldn't take much time. "You should spend like 1% of your time on the strategy. Pick it and then spend 99% of your time trying to execute," Schmidt said. McCall's operating rules are deliberately unglamorous, and they are the most transferable part of the conversation.

On annual contract value, the discipline is to decide the threshold once and stop agonizing: "You don't want to be taking deals that are negative to your unit economics. But if it passes the threshold, then the answer is you don't think about it." If you can build a go-to-market engine that lives off $15,000 ACV deals, take as many as you can — just don't take $8,000 deals — then stack wins and inch up the ladder into larger companies. Pylon is executing exactly that trajectory today.

On trials, the danger is the proof of concept that never ends, because an AI product's capabilities improve every day and the answer to "can it do this?" is always yes. McCall's discipline is an end date, always: "It's a 30-day trial, it's a 45-day trial, it's a 60-day trial. Period. End of story." Success criteria must be defined up front, and duration must match complexity — if deployment takes two weeks, a two-week trial is absurd.

Schmidt adds a subtle complication: a product working and a product being used correctly are two different parts of the equation. When a startup takes on both risks — product performance and customer configuration — it needs to educate the buyer on exactly what is being signed up for. The best evangelists he has seen, including Decagon, Stu, and Further AI, commit to specific benchmarks, hit them on schedule, and use forward-deployed teams to get customers running.

The two motions also hire differently. Land grabs want "very aggressive" sellers — attitude and aptitude, often earlier in career, people who will work a large market and stack wins fast. Lighthouses want seasoned enterprise sellers who understand long sales cycles and procurement. And sales operations should be hired before it feels necessary: not a giant RevOps organization, but one person who owns territory alignment, name lists, commission disputes, and the "sales constitution." "When you get into scale mode, you want all that stuff largely figured out. You don't want it to become speed bumps," McCall said.

On quota, the early-stage target is 100 percent of reps hitting: "Sales teams run off momentum. You want to hire winners and give them a chance to win." A team where only 40 to 50 percent of reps make quota is doing itself a disservice — either the quota is too high or the hiring profile is wrong. And early-stage cost of sales barely matters: "Nobody looks back and says, gosh, six years ago your cost of sales was really terrible. Nobody cares about that." His career advice to sellers runs the same direction — chase the best company, not the biggest commission, because a great company is a "career elevator," and titles and base salaries are noise by comparison.

The framework matters more now than it did a decade ago because of what changed underneath it. For roughly fifteen years, product-led growth — where users sign up for free and the product sells itself — was not a philosophy but a byproduct of the cloud platform era. The big categories — CRM, HR, IT service management, security — were founded between roughly 2000 and 2010 and had already been won. Late entrants had only one path: a wedge product that solved one slice and then expanded, a motion known as land-and-expand. It became the default because it was the only path, and switching cloud providers offered no reason to move.

"Going from cloud to cloud for CRM, I don't care if the button is green or blue. I don't care if there's one little feature difference. I'm not going to switch," Schmidt said.

AI topples that logic. Enterprises are no longer comparing green buttons to blue buttons; they are reimagining the category itself — humans doing higher-value work, agents doing the rote work, entirely different organizational structures. Schmidt's essay "Trading Margin for Growth," published about a year and a half before this conversation, argued that software innovation moves in cycles. This is the moment the cycle turns.

"There's a moment right now to go sell big software again and to go sell platforms," he said. "We're now looking at a different way of doing business entirely. This is not a skeuomorphic one-to-one replacement, green to blue. We're now thinking about humans doing something completely different, way more high value, and agents doing the mundane rote work."

The market conditions even rhyme with Samsara's ELD moment without the mandate: AI boards inside every enterprise are telling buyers what to purchase and by when. "That surely will go away," Schmidt said, "but there is this moment of crazy kinetic energy inside of big companies." McCall's complementary observation is that buyers get more educated with every technology transition, so the missionary sell is obsolete — the job is simply to convince an educated buyer that your company is the right solution. "Every year, and with every technology transition, buyers become more and more educated. If you can hit that buyer when they're in their decision mode, you're going to have a better chance."

That leaves two founder failure modes. The first is analysis paralysis — picking a strategy and then refusing to move. The second is logo-chasing vanity: "There's no bonus points for hard-earned revenue. You don't get extra multipliers on your revenue if you get the big logo." Get out, talk to customers, find who will buy what you have today — and after the first year, if milestones are hit, reassess and adjust.

The market implications are simple to state and hard to execute. The current AI-buying urgency inside big companies is a window, not a permanent condition, and the startups that win it will be the ones that answered the two questions — buyer exposure and whether proof travels — correctly, then executed their chosen playbook without ego. The unresolved tension worth watching is whether today's land-grab AI companies can replicate the Meraki and Samsara arc, climbing from $15,000 contract-value replacements into platform-scale lighthouse accounts, before the AI boards' mandates expire and the kinetic energy Schmidt describes dissipates. For investors, the signal is not the most impressive logo on a pitch deck; it is whether the company picked up the phone, got on the plane, and went where the math already pointed.