The episode opens with a rapid-fire montage of its best ideas: Joe Schmidt lamenting that every startup is chasing the same San Francisco logos, Andy McCall recounting how Meraki competed against Cisco by targeting mid-market customers who just wanted simplicity, and the observation that too few founders are willing to pick up the phone and get on a plane. The narrator then introduces the episode premise: Elena Burger sits down with Joe Schmidt and Andy McCall to unpack two competing go-to-market strategies — lighthouse versus land grab. The framing question is direct: do you win a handful of high-profile customers and use their credibility to unlock a market, or do you find customers with existing budgets, prove the math, and capture as much of the market as possible? The cold open establishes the stakes before any formal introduction begins.

Elena opens the episode proper and prompts Joe Schmidt to explain the origin of his piece — an observation made while driving up Highway 101, where he noticed competing AI companies targeting the exact same San Francisco logos rather than considering the much broader opportunity elsewhere. ^{[1] — Joe Schmidt "Two axes define your GTM destiny: buyer exposure (high vs. low risk of buying wrong) and whether proof travels in your market. High exposur…" 03:40 Play} The framework that emerged is elegantly simple: a 2x2 matrix with buyer exposure on the Y-axis (the risk a buyer faces if they purchase the wrong product, including regulatory consequences and reputational harm) and proof travel on the X-axis (whether a successful customer win generates social proof that cascades through an industry). The top-right quadrant — high exposure, strong proof travel — defines the lighthouse market: regulated industries where getting it wrong can mean legal trouble, and where a few marquee wins unlock a flood of followers. The bottom-left defines the land-grab market: low buyer exposure, established budgets, where the seller just needs to show the math. Joe is careful to describe these as 'proof' versus 'math' — a clean shorthand that will carry the rest of the episode.

With the framework established, Elena pushes for concrete examples and Joe delivers two that crystallize the distinction. ^{[1] — Joe Schmidt "Stuut went after accounts receivable — unglamorous, but with an established budget and a clear ROI story. Their pitch was pure math: AI col…" 13:40 Play} Stuut, the AI accounts receivable company founded by Tarek and Ben, attacked a market nobody glamorizes — collections — but with a decisive advantage: an existing budget and demonstrable math showing AI could outperform human teams at the task. They went to mid-market buyers, showed the numbers, and stacked wins. Harvey took the opposite path into legal AI, a high-exposure market where buying the wrong product could expose a law firm to regulatory or ethical risk. Their playbook was to identify the specific handful of law firm lighthouse accounts whose endorsement would make the entire legal industry feel safe — and once those fell, proof traveled automatically. Andy McCall then adds Pylon, an AI-native customer support company in the a16z portfolio, as another land-grab example climbing the ACV ladder by directly replacing existing workflows with a faster, AI-powered alternative.

Joe Schmidt prompts Andy McCall to share the Samsara origin story, specifically how the company navigated the social proof question in what looked like a potentially regulated, high-exposure market. Andy's answer is refreshingly honest: there wasn't a grand strategy session about lighthouse versus land grab. ^{[1] — Andy McCall "The US government's 2016–2019 electronic logging device mandate forced every trucking company to buy telematics hardware — overnight, an en…" 07:40 Play} When you're an 18-month-old company, the cold calls to the largest trucking firms end with a click before you finish your pitch. So Samsara listened to the market: mid-market transportation customers would actually take the call, buy quickly, and give rapid product feedback because short sales cycles meant short feedback loops. The ELD mandate then acted as a government-issued purchase order for the entire industry — AT&T and Verizon and established players all benefited, but for a new entrant with a modern product, the mandate meant a certain percentage of every company in America suddenly had both budget and motivation to evaluate new options. Andy describes it plainly: rising tide floats all boats, but it helped new entrants most. Joe draws the parallel to today's AI moment — CEOs and AI boards everywhere are mandating AI adoption, creating a comparable urgency without (yet) the force of law.

Joe Schmidt pivots to Meraki, specifically probing how the company navigated ACV thinking in its early days and what lessons apply to AI companies dealing with inference costs. Andy obliges with the full backstory: Meraki was founded in 2006 by MIT PhD students whose RoofNet research project — large-scale mesh Wi-Fi installed on Cambridge rooftops — failed as a municipal business model and pivoted into enterprise networking. ^{[1] — Andy McCall "In 2009, everyone thought Meraki was insane for trying to sell enterprise networking when Cisco and HP had already won the market a decade …" 19:05 Play} In 2009, the conventional wisdom was that the enterprise networking market had been won by Cisco and HP a decade earlier. The breakthrough insight was that the cloud made it possible to configure and manage networking equipment remotely — obvious in retrospect, genuinely innovative at the time — and that this mattered most to mid-market companies without dedicated IT teams trained in command-line configuration. The land-grab play was to target buyers who didn't have the staff to work with Cisco's complexity, making Meraki's simplicity not just a feature but a decisive competitive moat. Andy also introduces the ACV philosophy he carried from these experiences: set a floor based on unit economics, then stop thinking about deal size and just stack wins above the threshold.

Andy describes the mechanism that made Meraki's land-grab so effective: a webinar program where attending earned you a free access point. ^{[1] — Andy McCall "Meraki couldn't beat Cisco on reputation in 2009. So they stopped trying to pitch and started shipping. Attend a webinar, get a free access…" 19:55 Play} The logic was elegant — no amount of pitching could communicate simplicity as effectively as experiencing it. Plug it in, configure it in minutes, and the light bulb goes off. Joe asks how this translates to AI companies, where products are configurable, deployment is complex, and capabilities are evolving daily. Andy identifies the key risk: AI POCs can become open-ended science projects where buyers keep asking 'can it do this, can it do this?' because the answer is almost always 'yes, and we'll add it next sprint.' The discipline required is to resist scope expansion and pre-negotiate two things: a hard end date (30, 45, or 60 days — period) and mutually agreed success criteria defined before the trial starts. Joe adds a nuance: when automating something that has never been automated before, there's a configuration cost and a learning curve that means the product may technically work but still require significant hands-on optimization — and founders must be clear about what they're signing up to deliver versus what depends on customer adoption.

Elena asks who does onboarding and customer evangelism best, and Joe Schmidt offers three examples that span the risk spectrum. Decagon, the AI customer support company, is his lead example: they enter each engagement with explicit, pre-defined benchmarks — 'here is what we will deliver' — and then hit those benchmarks within the agreed timeframe. This disciplined commitment to measurable outcomes is what builds trust in a high-visibility product. Stuut earns praise for relentless founder-led selling — Tarek and Ben hit the pavement better than almost anyone Joe has seen. Then there's Further AI, an insurance-focused AI company where Joe sits on the board: their customers are some of the largest insurance companies in the world, a sector not historically known for technology adoption. Further AI wins those accounts by leading with governance and security, building a compliance-first posture, and then embedding forward-deployed teams directly in the customer's environment to ensure successful rollout. Andy adds that Further AI is a good example of a lighthouse strategy — winning major insurance brands whose social proof then cascades down the long tail of smaller insurers.

Elena asks whether a company can stay lighthouse forever, and the discussion broadens into a rich analysis of how both Samsara and Meraki navigated the transition between strategies. ^{[1] — Andy McCall "No company stays purely land grab or purely lighthouse forever. Both Samsara and Meraki started by capturing mid-market broadly, then verti…" 27:20 Play} Andy's answer is that most companies will eventually deploy both playbooks, and the trigger for switching is usually verticalization. At both Samsara and Meraki, the early land-grab captured mid-market breadth; then as the company matured, the sales org verticalized — identifying the top 5 transportation companies, the top 5 public sector accounts, the top 5 school districts — and those campaigns became lighthouse plays, because the decision-making processes, procurement cycles, and buyer profiles were fundamentally different from mid-market. Andy also notes that school districts are a surprisingly clean lighthouse market: they all know each other, and winning the biggest district in a state cascades immediately to all the smaller ones beneath it. Joe adds the counterexample of Applied Intuition, an autonomous vehicle software company in the a16z portfolio whose market is so constrained — a finite number of car manufacturers — that the lighthouse playbook is permanent, and every account must be treated with extraordinary care because the total addressable universe is small.

The conversation turns philosophical as Elena raises the question of developer-led, bottoms-up product adoption — the PLG model — and whether it's still relevant. Joe Schmidt seizes the moment to deliver the episode's central thesis. ^{[1] — Joe Schmidt "For 15 years, cloud-to-cloud SaaS replacements weren't worth the switching cost. Now AI is forcing companies to rethink entire workflows fr…" 33:05 Play} His earlier piece 'Trading Margin for Moat' argued that PLG dominated the last 15 years because the major cloud platform companies in CRM, HR, ITSM, and security had been founded between 2000 and 2010, locking up the enterprise and leaving challengers with no option but to wedge in with a feature and expand. Going from on-premise to cloud was a big enough shift to justify switching vendors; going from one cloud CRM to another cloud CRM over a slightly better button color was not. But the AI transition is categorically different: it is not a skeuomorphic 1-to-1 replacement. Companies are fundamentally rethinking who does what — humans are moving to higher-value tasks, agents are handling the rote work — and that reconception opens the door to replacing entire platforms, not just swapping features. This is why the moment to sell big software is now. Andy adds that as each technology transition unfolds, buyers become progressively more self-educated, meaning the modern seller's job has shifted from missionary education to differentiation — the buyer already knows what they want, you just have to prove you're the right choice.

Elena asks the pointed question: why do founders misjudge which game they're playing? Joe's answer is blunt — it sounds more impressive to tell investors you're selling to JPMorgan Chase than to admit you're winning mid-market accounts in the Midwest. The prestige bias pulls founders toward lighthouse plays even when the math of their product and market clearly points toward a land grab. ^{[1] — Andy McCall "Strategy's important, but you should spend like 1% of your time on the strategy. Pick it and then spend 99% of your time trying to execute." 37:51 Play} Andy's prescription is radical in its simplicity: strategy matters, but the fatal mistake is spending too much time on it. His formula is 1% strategy, 99% execution — pick a direction and go. There are no extra multipliers on revenue for winning the hardest deal, no bonus points for choosing the prestigious customer over the accessible one. Get out, talk to customers, find the ones who will actually buy the product you have today, and chase that path. After the first year of hitting revenue milestones, reassess — but don't let analysis paralysis masquerade as rigor.

The conversation shifts gears into a rapid-fire exchange where Joe Schmidt queries Andy McCall on lessons learned across his career. Andy reveals he's closed deals in some unconventional settings — fishing trips, shooting ranges, ballparks — a fitting illustration of the land-grab ethos: be willing to sell anywhere. His biggest career advice for early-stage salespeople is counter-intuitive: don't chase the best commission rate, title, or hottest technology — find the best company that's going to grow, and treat it as a career elevator. The ego-driven pursuit of director titles and base salary targets is the mistake he made early. On organizational design, Andy flags sales operations as the function founders hire too late. It doesn't require a big team — one person thinking daily about territory alignment, commission structures, name lists, and a sales constitution is enough to prevent the bottlenecks that emerge at scale. On quota philosophy, Andy is unequivocal: in early-stage companies, close to 100% of the team should be hitting quota. If only 40–50% are hitting, quotas are too high or the hiring profile is wrong. The whole point of the early sales team is to stack wins, and you attract and retain the best sellers by giving them a realistic chance to win.

Elena Burger closes the main conversation by observing that what struck her most about the discussion is how consistent these principles are across different technological eras — the frameworks Joe and Andy described for cloud networking in 2009 map cleanly onto AI in 2026. Joe and Andy sign off warmly. The episode closes with the standard Andreessen Horowitz legal disclaimer: the content is for informational purposes only, should not be taken as legal, business, tax, or investment advice, and does not constitute an offer or solicitation for any a16z fund investment. Listeners are directed to a16z.com/disclosures for more details, including a link to a16z's own portfolio investments in many of the companies discussed.