The VCs All Asked Who Else Had Done It. HubSpot’s Answer Was Intuit.

Sarah Guo opens by calling HubSpot one of her hero companies: an unlikely SMB marketing-software startup, founded in Boston in 2006, that became Salesforce’s primary contender through strategy, execution, and a community that is as much a movement as a business. Brian Halligan, co-founder, longtime CEO, and now executive chairman, tosses the compliment back. Why, he asks, were they an unlikely success?

Because of Boston, Guo says, and because of the vintage. In 2006 the markets venture capitalists wanted were high-end enterprise in a small number of categories. Selling software to small and midsize businesses to manage an online presence did not look like a path to a $25 billion-plus company. “It was clear to us,” Halligan says. Guo’s reply: that is what makes you the founder. Her own hero list, under pressure, starts with Figma — Dylan Field and Evan Wallace changing how people work in a way that was not obvious at the beginning. Then she puts him to work.

Halligan’s path to HubSpot has three chapters. Chapter one was PTC, where he started as the first sales-development hire at headquarters in Boston, moved into sales and channels, then spent years in Japan and Hong Kong building Asia. Ten years, a hard-charging, eat-your-young, sales-driven culture: a terrific company and a tough place to work. Chapter two could not have been more different. He started the sales organization at Groove Networks for Ray Ozzie, the father of Lotus Notes. Groove got to about $20 million in revenue, Microsoft bought it, and it became a little part of SharePoint. Ozzie called himself a social anthropologist: watch where technology is going, imagine how human behavior will change, and build something in the future to match. Groove was a bit of Dropbox, a bit of Slack, a bit of Notion, on peer-to-peer platforms, just a little early. Ozzie was also early on freemium and product-led growth in B2B, when almost nobody else was.

Chapter three was MIT Sloan. The right-wing intensity of PTC, the left-wing product idealism of Groove, and back-to-basics business school still influence him, and HubSpot. The company itself started from two ahas, one his and one his co-founder Dharmesh Shah’s. Halligan was an entrepreneur-in-residence at a going-nowhere Boston firm called Longworth, teaching portfolio companies the PTC playbook. They all bought lists, cold-called, spammed, hit other people’s trade shows, hired PR agencies, and bought ads. It did not work. Humans had gotten good at blocking traditional marketing: caller ID, ad blockers, DVRs. His playbook was wallowing.

Shah blogged his way through business school on OnStartups.com, writing up interesting lectures. Halligan looked at the Google Analytics and compared them to his wealthy, venture-backed startups with money, teams, and expertise. The “crappy little blog” was crushing them by two orders of magnitude. He started naming the split: his people were doing outbound marketing; Shah was doing inbound. You have to match the way you market with the way humans actually shop. People had started using Google, unsubscribed from magazines, subscribed to blogs, and shown up on social media. The foundational idea was a platform to help marketers move from the old-school outbound world to the new-school inbound one.

The INBOUND conference is the temple that survived every later chapter. Early on they described themselves to VCs the way that sticks: Salesforce is to sales as HubSpot is to marketing. They sat in the Dreamforce audience every year praying Salesforce would not announce a marketing product. User conferences felt too narrow, so they built a community event people would attend even if they were not customers. The first one was a room of 400 at the Marriott in Kendall Square, with Seth Godin, David Meerman Scott, and Brian Solis — early acolytes who were already saying similar things, with or without the word inbound. It sold out in about five minutes. They got on the treadmill. Today it is 12,000 people live and a couple hundred thousand listening, with speakers from Barack Obama to Reese Witherspoon. Halligan’s favorite interview is Andrew Huberman, who is on his Mount Rushmore. Guo calls it a pilgrimage. Halligan’s better metaphor is the parking lot outside a Grateful Dead show: the community is the point, and the sessions people value most are the small ones, like what the hell is AI and how do marketers get their arms around it.

Inside the company there was a fight over the name. Half the room said internet marketing software, which it was. A couple of people said inbound marketing software, and everyone asked what the hell that was. They chose the harder road and tried to create a category around inbound versus outbound. What resonated, Halligan thinks, was an enemy (outbound always helps), a catchy word that stuck, and the fact that Godin, Scott, Solis and others were already talking about the same shift — HubSpot just stuck a word on it. Then they worked it: each founder wrote two blog posts a week, they spoke at flea markets if invited, and they wrote a book nights and weekends. It was a ton of work and one of the keys to success. Inbound sales, by contrast, failed. Sales leaders said they did outbound sales, that is what they do. A CMO’s boss might ask why Google could not find the company, or mention a new Facebook account. No CEO was asking the head of sales why inbound leads were up. They had to support outbound selling to build a viable sales business.

Acquisitions taught a harder lesson. HubSpot had been built for the top of the funnel: strangers to website visitors to leads. They were not passionate about the database, segmentation, and email drip campaigns, and did not have deep expertise. After a Sequoia round, some of the proceeds were meant to buy that middle-of-the-funnel layer. They bid for Pardot and could not get to a number. Halligan flew to Marketo’s CEO and proposed a merger of equals, offering to give up the CEO job; Marketo was not interested. They acquihired David Cancel and Elias Torres, known quantities in Boston who were passionate about the middle of the funnel, and that team built it. Cultures are hard to match. If he had it to do over, Halligan says, he would have done a much better job assimilating and partnering with them. He was pretty junior. They have done only very small deals since.

Halligan is a believer in the Peter Thiel line: you have to be right about something everyone else thinks you are wrong about, and you have to be right about it for a long time. HubSpot’s first zag was SMB — really the M in SMB, startups and scale-ups, not big slow enterprises. Every VC asked who else had done it. The answer was Intuit, and then there was no one else. Historically you could not drive down customer-acquisition cost because you needed humans. Inbound marketing opened a path, if you also upsold existing customers, got lifetime value up, and pushed revenue retention over 100 percent. Board members stayed skeptical through the IPO; he still gets asked. The cost was real: they diluted more than they would have by going enterprise like Marketo and Eloqua. Everyone wanted them to be Marketo. The world, they said, did not need a third one of those.

The second zig arrived when the Dreamforce prayer failed. Salesforce bought four marketing companies in one year, and HubSpot went from best partner to enemy. They pivoted underneath, into CRM: not a Salesforce clone, but something incredibly easy, inexpensive, and freemium. Investors were bored. Most employees thought it was a terrible idea. The founders’ evidence was personal: they used Salesforce, it was hard and expensive, so they would make an easy, inexpensive one. Almost ten years later the company has Marketing Hub, Sales Hub, Service Hub, Ops Hub, CMS Hub, and Commerce — a suite on the way to the CRM platform they drew on a whiteboard at an offsite, plus a shift from pure inside sales to what they then called premium and everyone else calls product-led growth. Nine years later they are still filling that in. The sales product grew faster and more profitable than the original marketing business.

Year eight they tried two changes at once: a sales application and a freemium motion. They copied the Apple Mac lore — a separate part of the building, separate login, separate stack. Halligan jammed three internal stars into that isolation: Brian Balfour, Christopher O'Donnell, and Mark Roberge, now a sales professor at Harvard Business School. The three could not agree on the vision. The lesson: people have to want to come together; you cannot mush immensely talented people together and expect alchemy. Some found other gigs. Halligan became, in effect, co-founder with O'Donnell, made a couple of strategy calls, and ran. It took a good two and a half years to get a half-decent sales product and a half-decent PLG motion. They kept a separate sales team and a separate instance of HubSpot. Marketing reps were not allowed to sell it. On merge day they did the counterintuitive thing: the sales business “acquired” the marketing business, and the freemium motion plus a bunch of that leadership moved onto marketing. Dealing with the fallout, finding the strategy, and having patience — he made a lot of mistakes in there. He still considers it maybe his best chapter inside HubSpot.

O'Donnell later told Guo the one question to ask Halligan: how HubSpot kept succeeding after the IPO, when so many Boston companies have not. Halligan had heard Marc Andreessen argue that companies outside Silicon Valley optimize for a local maximum — biggest tech company in Boston or Austin — while Valley companies want to beat Apple and Google. HubSpot’s answer was a slogan they meant: build a West Coast company on the East Coast. Once a year they drove the entire management team west to sit with people who would push their thinking. The other line he said a million times: the IPO is the starting line, not the finish line. Guo had heard O'Donnell quote it in their first conversation. It had permeated.

Going public usually splits a company: name five officers who see everything and get locked into trading windows, and block a lot of information from everyone else. HubSpot flipped it. Every employee, including a frontline support person, was treated as an officer — same information, same trading windows — so the IPO did not feel like a big internal change. The roadshow was still awkward. They had a $200 million marketing business and a $1 million sales business on the last slide. Bankers said don’t even bring it up; nobody gives credit. They brought it up anyway, because it was a big internal initiative and people could see it. Dharmesh and Brian wanted to build something their grandkids would be proud of. Halligan’s father had worked at BBN and at GE when GE was cool; he brags about his dad, and he wants his son Luke to brag about him.

After the Sequoia round — Sequoia, Google, and Salesforce in the cap table — they bought Performable, stopped tracking expenses, and spent “like drunken soldiers.” They had never missed a quarter. They missed that one on revenue and missed expenses by a country mile. Longtime board members Larry Bohn of General Catalyst and Dave, Halligan says, told him he could use some help: hire a COO. He pushed back — “I got this” — then interviewed a pile of people and met JD Sherman. Sherman’s IBM years were an “oof” for a fast company that was the antithesis of IBM, but he had been CFO at Akamai, a Boston scale-up they admired, a trains-on-time person with humor and culture fit. It was a really good call. The board later pushed a Columbia psychology-professor coach, also useful. When HubSpot hit $2 billion in revenue, Halligan did not think the road from $2 billion to $20 billion was his forte, and he was less passionate about it. Yamini Rangan, who had been running go-to-market, became CEO; he became chairman. A lot of CEOs, he says, need a COO. A lot should not retire but move to chairman and let someone else run the day-to-day. They might be surprised how happy they are.

HubSpot had an “AI” team for a long time that was really an ML team. Two and a half years before this conversation, Halligan got on the phone with Sam Altman — “before he levitated and walked on water” — and asked whether to raid MIT for AI PhDs. The talent war already cost about a million dollars a hire; it is two million now. Altman said don’t hire a bunch of PhDs, you will never keep up; OpenAI was already losing people to Google. Second, AI was progressing linearly, slowly. At some unknown future point it would go asymptotic. Keep chugging and wait. When ChatGPT arrived, Halligan’s verdict was: the man was right. Shah got early looks. They tried to get into the Microsoft round and were not let in, which Halligan still calls a bummer. Blockchain, crypto, and mobile were interesting. This shift is fundamental to CRM and to how humans shop and buy. It was not something they could half-measure.

Buying is going to change. Google answers a question with blue links that land a prospect on your site. In ChatGPT people keep asking questions about you and may never visit. High-quality content will matter more; using AI to blast out tons of content will not, because the model builders will look at similar signals to Google — inbound links, the power of those links, other marks of credibility. Halligan has been an ungate-everything person, white papers out from behind forms and into the blog. He now thinks gates may come back a little, because you still need a reason to visit the site. PLG remains viable. The jobs of marketers and sellers are going to change a lot.

On the product side they have shipped a pile of V1s and betas with real adoption. Content assistant writes the blog post, the image, the landing page. Campaign assistant looks at what is going on and builds a better workflow than a human clicking through screens. ChatSpot sits on OpenAI, Anthropic, and other models plus HubSpot’s API. Casual users do not want ten clicks to build a donut-shaped report; they say, give me last month’s leads by geography. That new UI is interesting. They are just starting to build agents. Customers already use bots on their sites for support and marketing. Support orgs and BDR/SDR orgs, he thinks, will get incredibly efficient: more questions, 24 hours a day, not tired, not hung over. They will still make mistakes. The world in CRM and go-to-market shifts mostly in positive ways.

Guo asks the incumbent question: a dozen startups in each of those domains, and HubSpot still feels like the rebel to itself. Usually rebels have the advantage. Halligan does not think that is the case here. They have so much data, over so many years, about so many customers, and that is what makes this stuff sing. A brand-new startup has to get data, and data is hard to get. Incumbents like HubSpot and Salesforce have an unfair advantage if they lean in and move fast. There are about a thousand CRM companies; not all of them will. Guo teases him: it is funny to call HubSpot an incumbent at 20 years and about $26 billion of market cap.

Guo brings up Glassdoor wins and an old dinner about culture. For HubSpot’s first three years, culture was a four-letter word. You cannot measure it; it is soft; do not bring it up. About three years in, Halligan joined a CEO group largely because Colin Angle of iRobot was in it — a “major man crush.” Nine CEOs, one topic all day. The topic was culture. He said nothing all morning, then sat next to Angle at lunch to talk product and innovation. Angle clocked him: you don’t like this culture topic, do you? Culture, Angle said, is how you scale your company. Culture is how people make decisions when you are not in the room. Halligan participated in the afternoon. The next day he repeated the line to Dharmesh, who said that sucks, then heard the sentence again. In a weak moment Shah agreed to be culture czar, not a role he usually takes on.

They hired a Sloan professor to survey employees: on a scale of one to ten, how likely are you to refer HubSpot to a friend, and why. Almost everyone said they liked the culture, which shocked two founders who had forbidden the word. Version one of the culture deck described what it was like to work there. They ripped off the Netflix format — Netflix had a unique culture and a famous Culture Code — and wrote a very different document in the same shape, posted it, and watched it take off. It still does. Halligan’s frame is two products: one you sell to customers, a magnet if it is high quality and unique; a second product that is culture, a magnet for employees if it is the same. They edit the document every six months. They still run a quarterly net-promoter survey and publish every answer that is not full of swear words on the internal wiki. They address many of them, sometimes with “we’re not going to do this, but we heard you.”

People will complain about everything. The masterpiece is Smoothie-gate. They put in a smoothie bar, like a proper Silicon Valley company, and for some forgotten reason — cost, hassle — there was no protein powder. A thousand complaints. Eight years later, still no protein powder. Their first people leader was not a career HR hire. Katie Burke was a marketer who was a dog with a bone on people issues. It took weeks and months to convince her to leave marketing. She still runs people ops, and she still will not give in on the powder. Halligan’s hiring logic: recruiting, compensation, culture, DEI, and benefits do not rhyme the way sales and service do. A career HR person usually grew up through one of those tracks. It is the one function where you can take a smart athlete who thinks from first principles and let them figure it out.

Halligan’s second act started on a snowmobile. Two years before the taping he broke 13 bones, had five surgeries, and lay in the snow thinking he was going to die. In that moment he knew two things: he no longer enjoyed running HubSpot’s day-to-day, and he was not as good at this size as he had been at earlier sizes. If he lived, he was not coming back. Shortly after getting healthy, with Yamini as CEO, he wanted into climate. It is the existential crisis of our lives, and he is a mission-driven person. He did a roadshow of founders, VCs, and professors. Everyone in climate, he found, was depressed and not optimistic.

Then he visited Woods Hole Oceanographic Institution on Cape Cod. Oceanographers, surprisingly, are optimistic. The ocean absorbs much of the carbon dioxide from the air and almost all the excess heat we produce; it got us out of the last ice age; it is the only thing of scale that can get us out of the pickle. Trees will not get us there. They die and give a lot of the carbon dioxide back. He started volunteering, then decided research was not enough — the problem needed venture money. He co-founded Propeller, a $120 million fund backing startups at the intersection of climate change and the ocean. A little over a year in, about ten investments, so far so good. The closer is the same as the HubSpot years: know what you are good at, and know when to give the chair to someone who is good at the next size.