Building a Beacon in the Fog of AI
What this covers
The conversation, with Nilam, Rick and Paul, follows Ganenthiran from running Instacart to building Beacon, which acquires vertical software businesses serving operators such as campground owners. A good share of it looks back at Instacart. Whole Foods was 43% of GMV when Amazon acquired it, and it both made the brand and nearly sank it through an unprofitable contract. For about seven years Instacart did no consumer marketing, because retailers promoted it to their own customers. The $2 billion Series C, raised at a run rate of roughly $7 million in GMV, left a cost of capital obligation he still says he feels. He also describes a stretch at Procter & Gamble where brand managers carried responsibility without authority.
The Beacon material is where the contested claims sit. Ganenthiran frames vertical software as a bond plus call option, argues that coding agents will not eliminate enterprise software, and predicts that cheaper R&D makes distribution the scarce asset. He treats M&A as a product problem, claiming about 80 percent of deals are proprietary, and says Beacon avoids winner-take-all markets because Instacart needed 30 to 35 percent share in a tight geography to reach profitability.
The last thread is operating philosophy: a wartime model built on clarity of purpose, velocity and single-owner workstreams, a refusal to chase small fires, and impatience with inputs but patience with outcomes.
Ganenthiran argues that AI adoption in the real economy is limited by trust and operating discipline rather than technology quality, so a permanent-capital holding company that owns trusted Main Street software can carry AI into those businesses.
- Businesses block technology because they do not trust it to be used in their daily operations, not because it is not good enough
- Holding companies with permanent capital can stay patient and then accelerate, which funds cannot do
- Software businesses that keep adding customer value can grow faster and become more profitable after acquisition, while buying distribution becomes the scarce asset as R&D gets cheaper
Trust, not competence, is the differentiator guiding Beacon's whole approach.
- As AI makes competence a commodity, trust is what will be left over as the differentiator, and this principle has guided Ganenthiran's whole life and the organizing idea of Beacon.
“At the end of the day, as AI makes competence a commodity, the thing that's going to be left over is trust.”
- Working at D1 Capital gave Ganenthiran high-velocity decision making, investment process design, and a highly transparent, politics-free culture, which he tried to replicate at Beacon and which he says made Beacon possible.
“without D1, Beacon wouldn't exist — not the least of which they were my original investor and are still my big partner, but also just what I learned about high-velocity decision making”
- Years of showing up in retailers' parking lots built trust that allowed Ganenthiran to become the industry's ambassador after the Amazon-Whole Foods deal, as retailers listened to him and felt he was on their team.
“Because of the trust that was accrued over the previous 4 years leading up to this moment, they listened to me and listened to us. And they felt like we were on their team.”
- Modern scheduled childhood reduces serendipitous encounters and free unstructured time, because children's friendships are planned for them and they no longer experience boredom that pushes them toward books, TV, or independent play.
“when you're so scheduled, where is there an opportunity for serendipity, especially on the relationship side? Because everyone they meet is planned for them to meet in some way.”
- Ganenthiran learned early that if you try to be helpful to others, they will be helpful to you, and he grounds his operating principle of trust in the fact that he never thought he was very smart and needed another way to succeed.
“I realized quickly that if you try to be helpful to others, they'll be helpful to you.”
- Ganenthiran's career looked like a zigzag without a plan; what organized it was trusting his instincts at each turn, even though that felt uncomfortable at the time.
“In my 40s now, the organizing thing that has helped me in my career wasn't having a plan. It was trusting my instincts at each turn.”
- Beacon's four values (humility, horizon, hunger, and a fourth) were written by Ganenthiran in one evening and have remained unchanged, with hunger defined as a joint responsibility to each other to try one's best every day.
“The 3rd value is hunger. When you unpack hunger, the language we use is a joint responsibility we have to each other to try our best every day.”
Beacon operates as a wartime company, with focused execution and no artificial urgency.
- When Amazon acquired Whole Foods, Instacart shifted into wartime mode by splitting responsibilities and holding an all-hands, where Ganenthiran ad-libbed that this was the moment the company had been waiting for.
“We then went into wartime mode. We had to split responsibilities. We had to talk to the team.”
- In wartime, leaders must refuse to divert energy to every small fire, because putting out minor problems feels safe but takes energy away from the singular task that could determine survival.
“You let fires burn. Because if you think about a business, there's lots of little fires that pop up all the time.”
- Wartime does not exhaust people because they work faster, which is energizing; it exhausts them because they must constantly filter out distractions, like a windshield wiper, to stay focused on the road.
“Wartime's not exhausting because people are working faster on a task. I actually think that's an energizing part of it. Wartime is exhausting because of the other thing.”
- The clarity of purpose and adrenaline from crisis cannot be replicated by entrepreneurs who try to concoct their own wartime, because real back-against-the-wall pressure produces a feeling that artificial urgency does not.
“I've read entrepreneurs who are able to concoct their own version of wartime. I've tried doing that. It just doesn't feel the same as when you actually have your back against the wall.”
- A wartime operating model rests on three principles: clarity of purpose so the company does only one thing, velocity so tasks that take weeks compress to hours, and clear ownership with a single leader per workstream, which suits wartime better than peacetime.
“First is clarity of purpose. You can't do more than one thing.”
- The best compliment an organization can receive is that it feels fast when you walk in, and a new Beacon hire said everyone walks faster there, which signaled readiness for wartime.
“The biggest compliment an organization can have is how quickly it feels when you walk in their office.”
- Beacon's hunger value is a joint responsibility to try hard every day, because each employee and acquired founder has other options and the company owes them its best effort.
“the language we use is a joint responsibility we have to each other to try our best every day.”
- Proprietary AI systems now let Beacon go from idea to execution in hours rather than weeks, extending beyond code writing to idea generation and curation.
“it has become far easier to go from idea to execution. What took weeks can take hours now.”
- Ganenthiran says Beacon fell into the holding company structure, and it turned out to be the right design choice for a company that is patient and hungry, reconciling a slow-growth promise with fast-growth ambition.
“we fell into the holding company. It has ended up being the right design choice for us.”
Permanent capital and founder-led structure let Beacon build patiently at scale.
- 65 percent of Beacon's headquarters engineering, product, and design team have previously been founders, and about 75 percent of its more than 30 acquired businesses are still run by their original founding entrepreneurs.
“So 65% of the engineering, product, and design team at Beacon headquarters have been founders before. 75% of our businesses — and we own over 30 businesses — are still run by the original founding entrepreneur.”
- A planned long exit from a company is painful because others feel they helped build it, so a departing leader must be careful not to step on toes; this lesson shaped how Beacon treats founders after acquisition.
“What I realized is long goodbyes suck, because there's other people in the house that you felt you had a hand in building, and you've got to be super cognizant about not stepping on their toes.”
- Beacon uses a salad bowl model rather than a melting pot, so acquired founder-led businesses keep their unique cultures and ingredients rather than being blended into one corporate sauce.
“Our idea is not to have them come in here and blend into the sauce that we're making. Our idea is for each of them to bring their own unique ingredients”
- Beacon's stated ambition is to build a $100 billion-plus business, and Ganenthiran says it will, which he presents as clear and compatible with its holding company structure.
“Our ambitions now, we're clear on. We want to build a $100 billion-plus business, and we will.”
- A permanent-capital holding company can stay patient for two or three years without pressure to open stores or do M&A, then fully accelerate when markets open, which a fund cannot do; this is why Beacon chose the holding company structure.
“The businesses I admire the most are names like Cintas, which was a holding company, Danaher, of course, Walmart, Discount Tire ... And I think that's what a holding company allows you to do.”
- Beacon has around 80 people at headquarters, is running at nine-figure ARR, and has been profitable every year since it was founded, a design principle set after the Instacart experience.
“We're 80 people at headquarters. We're doing 9 figures of ARR. We are profitable, and we've been profitable every year the company has existed.”
Instacart's mistakes taught Beacon to avoid hype, broken models, and share-chasing.
- At about 40, Ganenthiran chose to stop relitigating Instacart's mistakes through other founders' businesses and instead build something the way he wished he had built Instacart, applying Bezos's regret-minimization framework.
“It really was a regret-minimization thing, going back to Bezos's framework.”
- Ganenthiran stepped down as Instacart president because he was burnt out after COVID growth, felt the company had lost its cultural way in growth euphoria, and could no longer change how people were treated while absorbing pressure between external audiences and internal teams.
“I felt the company had lost its way in the euphoria of the growth that came from COVID. The way we treated people started to change.”
- Instacart needed 30 to 35 percent share in a tight geography to reach unit-economic profitability, which created aggressive competitive tactics, so Beacon deliberately does not pursue winner-take-all markets.
“until we achieved somewhere between 30 to 35% market share in a tight geography, we would not be unit-economic profitable. Having gone through that, it just leads to a type of aggression and competitive tactics that was inconsistent with the type of business I wanted to run.”
- Companies must avoid believing their own hype during bull periods, because company building is eating glass and leaders who become euphoric are unprepared for the bad periods that follow; Instacart failed at this equilibrium.
“if you start to believe your own hype during the bull periods, whether it's a year, month, or day, you're just going to be really unprepared to eat the glass that comes.”
- Instacart struggled with whether its customer was the consumer or the retailer, and from roughly 2015 it was in practice an enterprise business building technology for grocers, but it did not acknowledge that for a long time.
“there was a moment in time, probably 2015-ish and beyond, where really, if we had introspected, we were more of an enterprise business”
- Instacart's business model was conceptually broken for a long time before it worked, and the company fixed its operations only when something was on fire rather than wiring the model properly early, which was extremely taxing culturally.
“Instead, what we did was we waited for crisis moments to be able to fix a bunch of stuff. So we got our house in order only when something was on fire.”
- Instacart signed Whole Foods on unprofitable terms to win the account against Google Shopping Express and Amazon, losing roughly $15 of margin on each Whole Foods bag delivered.
“made them a godfather offer that was extremely unprofitable to be able to have the right to deliver their groceries. ... Every Whole Foods bag that came to your house, Rick, probably cost us $15 of margin to deliver. Minus $15.”
Beacon's holding model earns value by building software customers keep needing.
- The line between software and services is blurring, so any service can be thought of as a technology ratio of code to human delivery, such as 80 percent code and 20 percent people, and that ratio will evolve over time.
“That technology can be 80% delivered with code and 20% delivered by human beings, or 100% delivered with code and 0% delivery. There's some ratio there, and that ratio will evolve over time.”
- Beacon is deeply impatient with inputs, meaning high velocity and lots of work, but patient with outcomes, because the timing of payoff is unknowable, and the cause of company failure is often the instant feedback loop of capital markets.
“we're deeply impatient with inputs, but patient with the outcomes. So you've got to show up, be really high velocity, do the work and do lots of it, but you don't know when it'll hit.”
- Beacon chose to be a centrally managed technology holding company rather than decentralized like Constellation, because once a solution works it can be deployed everywhere quickly, which was impossible even five years ago.
“We are a centrally managed technology holding company. That design choice was made because we think that allows for higher-velocity AI dissemination, because once you figure out something works, you can get it everywhere.”
- Beacon's customers are going to their software businesses and asking them to build more for them, including campgrounds, youth soccer clubs, and K-12 schools, which Ganenthiran sees as the exact thesis of the strategy.
“We are seeing our customers go to our software businesses and ask them to build more stuff for them, which is exactly the whole thesis.”
- Traditional holding companies treated vertical software as a bond because of subscription economics, but Beacon views it as a bond plus a call option, where the bond coupon is earned only by continually building new technology that adds value.
“When you bought a vertical software business, you were essentially buying a bond. At Beacon, we do not view ourselves as buying a bond. We are buying a bond plus a call option stapled onto it.”
- Software is not dead, but software that fails to add more value to customers next quarter than the previous quarter deserves no right to exist, and Beacon's acquired businesses are growing faster and are more profitable than when they were bought.
“I think software is dead if you don't try to add more value next quarter than you did the previous quarter to your customers.”
Formative career experiences built Ganenthiran's confidence, crisis skills, and growth approach.
- Commuting weekly from Toronto to San Francisco for seven of eight Instacart years was sustainable because Canadian dinner conversations do not center on work, letting him recover from high-stress weeks and return to them.
“Canada is just this fantastic place where when you go have dinner with people, they're not talking about what you do.”
- Working as a discount grocery cashier, Ganenthiran noticed that shoppers loaded up on diapers when government checks arrived, which showed him how socioeconomic conditions show up in everyday purchases and drew him to consumer packaged goods.
“I knew as a cashier exactly when the government checks would go out, because that's when people would come in and load up on diapers.”
- Harvard Business School taught Ganenthiran that there was a much bigger world and that he could hold his own among smart people, which he had previously doubted, and that forced speaking up in class built his confidence.
“Being in an HBS classroom, you're forced to speak up and have an opinion.”
- Selling soap and toilet paper to a struggling mass retailer taught Ganenthiran that moving product between store locations through displays can drive large changes in sales volume, which fascinated him about retail.
“I kind of fell in love with this idea that you could just move stuff from one part of the store to the other from one week to the next, and suddenly people buy a lot more Tide.”
- Managing the largest recall in consumer packaged goods at age 23 taught Ganenthiran how to manage crises under pressure.
“We had a big recall at the time, the biggest recall in CPG you could have. So I learned how to manage crises.”
- Ganenthiran did not join Instacart as a co-founder because he and his wife did not want to move to San Francisco and were starting a family, so he initially helped Apoorva remotely through nightly calls before eventually moving into a full-time role.
“I didn't join as a co-founder. Many mistakes in life — that is one.”
- Ganenthiran negotiated to keep working full-time at P&G while finishing school at night, which showed him he could take on far more work than he thought, and this realization shaped his later work habits.
“That proved to me that I could take on a lot more work than I thought I was capable of.”
- Brand management at P&G is a role of running a business without owning it, with responsibility but no authority over people, so the brand manager must persuade logistics, finance, and sales through data and human skill.
“how do you manage a business but not actually own anything?”
- Parenting and founding are the greatest jobs because both require constantly being wrong and relearning, which is the mechanism of growth.
“The greatest job in the world by far is being a mom or dad. And the reason for that is you're constantly wrong”
- Ganenthiran joined A.T. Kearney because it promised grocery retail work, and he was surprised to discover he loved consulting, working with nearly every major grocer, drugstore, and mass merchandiser in the country.
“Loved consulting, was there 3 and a half years, thought I'd go all the way to partner.”
Beacon's pitch rests on customer-first AI and systematized acquisitions.
- Small-business customers of Beacon's vertical software, such as campgrounds, youth sports clubs, and private schools, are asking their existing software vendors to build AI-enabled features because AI adoption is a survival imperative for them.
“We are seeing our customers go to our software businesses and ask them to build more stuff for them, which is exactly the whole thesis.”
- Beacon's combination of mission and scale, with 3.6 million users across 18,000 enterprises in 22 industries, is a compelling pitch that attracts top technologists to diffuse AI into the real economy.
“today, 3.6 million human beings will use one of our apps to do their job, across 18,000 enterprises, across 22 different industries.”
- The customer-first principle, defined as starting with what the campground or Instacart shopper needs rather than with technology, explains why Ganenthiran and Chris Rogers independently see AI as a tool that deepens customer relationships rather than a threat.
“we both start with the customer. What does the campground need? What does the Instacart shopper need?”
- Beacon's founders built a spreadsheet of 180 rows of firms competing for niche technology acquisitions, but they discovered their pitch was different enough from what entrepreneurs heard elsewhere that they had a right to exist.
“I think we built a spreadsheet of all the people who were doing it, and there were 180 rows”
- Beacon makes every action in high-frequency M&A and integration into a durable system, process, or technology product, since running many acquisitions at scale is impossible without systematized processes, and this mirrors Danaher's approach.
“we try not to do anything without thinking about how we can make it a durable process.”
Founders should stay patient and resist capital-driven overspeed and overraising.
- Founders should be deeply impatient with daily inputs and velocity while staying patient with outcomes, because capital markets' instant feedback loops and rapid valuation jumps distort timing and pressure companies to raise at inflated prices.
“we're deeply impatient with inputs, but patient with the outcomes.”
- Instacart's Series C at a $2 billion valuation, set when GMV run rate was only about $7 million, created a cost-of-capital burden because investors such as Kleiner Perkins expected returns of roughly 25-30% compounding, and that implicit obligation nearly took the company down.
“Every time I took someone's money, I felt I owed them their cost of capital.”
- Outsiders see successful companies as clean success stories, but founders who succeed often do not acknowledge the lessons, such as avoiding over-raising and over-speed, that their own journey taught them.
“I look back and think this is the gift and the curse. All I could see is all of the things we did wrong”
- Ganenthiran asserts that the buyers of small technology businesses have historically behaved predatorily toward founders, which is why they stand out by being modestly founder-friendly.
“the industry that Beacon operates in has for a long time been populated by folks who were somewhat predatory to the selling entrepreneur.”
A founder-centered acquisition process sets Beacon apart from predatory niche-services acquirers.
- Beacon's customer is the selling entrepreneur, and its people take responsibility after closing to make the business successful so the founder's legacy is enhanced, because selling a company is among the most stressful events in a person's life.
“our customer, we make no mistake about it, is the selling entrepreneur. And it is really hard to have empathy for a selling entrepreneur, when you're going through a transaction process where, other than a divorce, it's probably the most stressful point in people's lives.”
- Many acquirers in the niche business services industry behave predatorily toward selling founders through low LOI-to-close rates, last-minute retrading, and promising one thing pre-close while doing another post-close, so being founder-friendly stands out.
“That shows up with low LOI-to-close rates. That shows up in retrading at the last minute. That shows up in promising one thing pre-close and doing something else post-close.”
- Beacon treats M&A as a technology product problem rather than a bespoke investment problem, designing the end-to-end process around the founder's experience so that about 80 percent of deals are proprietary rather than brokered.
“We looked at the space and we viewed M&A as a technology product problem, versus a traditional bespoke investment problem. ... 80% of our deals are proprietary, not with a broker or investment bank in between.”
Trusted hands and distribution, not AI tools, are the real enterprise bottleneck.
- The narrative that AI coding agents will eliminate enterprise software is wrong because real-world businesses prefer to pay trusted people to build and run technology so they can focus on what they do best, such as selling groceries.
“real-world businesses will always want to pay people to build and help them with technology, whether that technology is called software or AI or agents or blah, blah, blah.”
- Technology gets blocked in the real economy not because it is not good enough but because people do not trust it to be used in their business to help with daily life, so the bottleneck is getting technology into trusted hands.
“the part where technology gets blocked is not because it's not good enough. It's because people don't trust it to be used in their business to help them with their daily lives.”
- If R&D costs fall in a competitive environment, the scarce asset becomes distribution, and companies will spend more to win the next point of distribution rather than taking extra margin, so Beacon buys distribution and builds technology on top of it.
“the scarce resource, or the asset that will hold value, is the distribution. It's S&M. My hypothesis at the time, I don't think we are seeing it yet, but I suspect we will”
Genuine belief in the venture makes recruiting, selling, and earning trust easier.
- Organizations where people want to put their name on projects do better in peacetime, while organizations where people play for the team and a single leader owns each stream do better in wartime.
“The second type of organization does better in wartime. The first arguably does better in peacetime, or not at all”
- Recruiting and selling people to join a venture is easy when the leader genuinely believes in it, so figuring out how to associate yourself with things you believe in matters because you will have to sell them constantly.
“in this case with Andrew, with sales, it's super easy to sell something if you really believe in it. Figure out how to associate yourself with things you really believe in”
- Ganenthiran joined Instacart because he envisioned that his daughter would never need to shop for groceries the way families do, a vision of changing daily life that he found compelling enough to trust over the uncertain risk.
“if this works — this being Instacart — our daughter Sita will never need to shop for groceries the way we do today.”
- Talent attraction at Beacon is easier than it was at Instacart because Beacon's people believe in what they are selling, and the pitch to top technologists is a profitable, fast-growing business with top investors and a channel to diffuse AI into the real economy.
“At Beacon, the story we get to tell the best technologists is, you get to come work on a very profitable business that is 9 figures of ARR”
- Being honest about how hard a venture will be, while still inspiring people, earns more trust than hype, and vulnerability is what people relate to.
“I think you can be honest but still inspiring. That level of vulnerability people sometimes relate to, because they don't want to hear all the hype.”
Retailer partnerships, not consumer marketing, both built and nearly broke Instacart.
- Whole Foods Market was 43% of Instacart's GMV when Amazon acquired it, and Whole Foods both made Instacart's brand, through retailer-driven marketing, and nearly killed it through an unprofitable contract.
“Whole Foods Market was 43% of Instacart's GMV at the point that Amazon bought them.”
- For about seven years Instacart did no consumer marketing, and retailers promoted instacart.com to their own customers, which acquired users for free and made Whole Foods the first retailer of scale to invent that model.
“they really made our brand and sort of invented the model where retailers did the marketing for us”
- Family-owned grocers, often third to fifth generation, saw Amazon transform other retail segments and knew they were ill-equipped to handle it alone, which made the grocery industry a wartime situation that opened the door for Instacart's help.
“These were family-owned, family-run businesses that were often 3rd or 4th generation, sometimes 5th generation.”
- Instacart's culture did not support executive longevity, with the next-longest-tenured executive team member staying about three to three-and-a-half years compared to Ganenthiran's eight, which undermined long-term relationships among peers.
“the next longest-tenured executive team member was there like 3 or 3 and a half years.”
Small businesses face existential AI adoption pressure that Beacon's scale addresses.
- Many AI commentators start with the technology, which is a weird place to start, because technology is a means to an end, while Ganenthiran and Chris Rogers both start with the customer's needs such as what a campground or shopper needs.
“a lot of the pontification you hear on X starts with the technology, which is a weird place to start, because the technology is a means to an end, not the end in itself.”
- Beacon's apps are used by 3.6 million people across 18,000 enterprises in 22 industries, which Ganenthiran describes as immense scale achieved within two years.
“today, 3.6 million human beings will use one of our apps to do their job, across 18,000 enterprises, across 22 different industries.”
- For small-business operators like campground owners, AI is not a strategic imperative but a survival imperative, and they readily adopt any tool that helps them manage fixed costs, scheduling, and systems.
“It's not a strategic imperative. It's like a survival imperative for the real economy, which makes it a lot easier for us”
- Ramp data showed AI adoption among SMBs around 20 percent versus the mid-80s percent for Fortune 500 type businesses, and the cost of being a late adopter for small businesses is existential, which Beacon aims to address.
“And it was 20%, versus mid-80s for Fortune 500 type businesses. That is what Beacon's trying to address at the limit, because the cost of being a late adapter on AI, if you are running any sort of business, is far, far greater.”
Instacart's volume and inflated funding shaped its economics and repayment obligations.
- Instacart pursued enough delivery volume with Whole Foods to renegotiate the contract, and by year two it had moved from losing money on each delivery to making money.
“So the first year and a half, we wanted to get sufficient volume such that we would be able to renegotiate that contract, get from red to green. And we did. By year 2, we had done that.”
- Raising at inflated valuations creates an obligation to investors' cost of capital, and Instacart's $2 billion Series C at a $7 million GMV run rate created a tax that Ganenthiran felt obligated to repay, so he treats every dollar taken as owed its cost of capital.
“We were at a $7 million GMV run rate, but that was the market, and we were trying to raise at as high a price as possible. ... Every time I took someone's money, I felt I owed them their cost of capital.”
Immigrant upbringing and Canadian culture fueled Ganenthiran's drive and ambition.
- Canadian culture's emphasis on long-term relationships and its lack of status-driven small talk create an environment that supports good entrepreneurship, and Canada's environment allowed Ganenthiran to manage intense work weeks while staying grounded.
“Canada is just this fantastic place where when you go have dinner with people, they're not talking about what you do.”
- Ganenthiran felt no overt pressure to prove his immigrant parents' decision, but the implicit drive to justify their sacrifice likely provided fuel for his ambition.
“under the surface, proving them right probably gave me a bunch of fuel to try to make something of myself.”
- As a child Ganenthiran subscribed to Fortune and Sports Illustrated and became fascinated with business titans like Jack Welch, which revealed an interest in business before he knew it was a career option, though neither parent had a business background.
“I really loved reading all these stories of Jack Welch and all the other titans of industry”
- Ganenthiran applied to only one school, HBS, to placate his mother and girlfriend, and was admitted, which opened a bigger world and showed him he could hold his own with smart people.
“For the first time in my life, I felt I could hang with all these smart people, because I always felt like I didn't belong in those types of rooms.”
- Ganenthiran's parents uprooted an upper-middle-class life in Kuala Lumpur when he was six to move to Canada for better opportunities, and he believes this story could not have happened anywhere else but Canada.
“That single gift was probably the most important thing in my life. I don't think my story would've been possible anywhere else but Canada.”