Matin Movassate and the art of getting out of the way

A founder usually writes a succession announcement to explain why somebody else should take over. Matin Movassate used his to ask why the company needed him in the first place. In March 2021, after eight years as Heap’s CEO, he announced that Ken Fine would succeed him. His test for leadership was pointed: “how can I make myself as redundant as possible?”

Redundant is an odd aspiration in a business that sells indispensability. Founders are encouraged to become the face, voice and occasionally the furniture of their companies. Movassate had been thinking about an organization that could operate beyond his daily involvement. He would become chairman. The business would carry on without requiring its original author to approve every paragraph.

There was a familiar shape to the decision. Heap, the analytics company he co-founded with Ravi Parikh in 2013, had been built around removing a dependency. People wanted answers about their customers. Getting those answers meant waiting for somebody else. The founder’s recurring interest was what happened when that somebody could step aside.

A good reason to finish the chores

Long before product analytics, there were files on his father’s computer. Growing up in the Bay Area, Movassate was enlisted to move and organize them. He wanted to return to his video games. Learning to automate the tasks offered a pleasing compromise: the work got done, and he reclaimed the time.

His parents had immigrated from Iran and ran a house-building business. He has described himself as a beneficiary of their difficult move. His own route into making things began with the immediate reward software offered. Change the code, see the result. For a child with games waiting, a computer that could perform the boring work was a useful discovery.

At Stanford, he studied computer science and economics. He also entered a master’s program and left it, later joking about the Silicon Valley prestige attached to dropping out. That joke fits the story better than a solemn account of destiny. His early incentive was quite practical. A tedious task had met someone willing to learn enough to avoid repeating it.

Before Heap, a question about tweets

There is an earlier public record of Movassate and Parikh working together. A Stanford course report dated June 4, 2009 carries both their names. Its subject is sentiment analysis of Twitter updates: whether a short post expresses a positive or negative opinion.

The paper considered how businesses might use those opinions to understand reactions to products or campaigns. Twitter still imposed its 140-character limit. The researchers were dealing with brief, messy expressions of human preference, and asking whether software could turn them into something useful.

It would be too tidy to treat a student project as a business plan in disguise. What it establishes is a collaboration, years before Heap, around extracting meaning from digital behavior. The two names on the report would later appear together on a company built to help other people investigate that behavior.

The queue inside the data company

Movassate’s route through technology included a software engineering internship at Mozilla, an associate product management internship at Google and a product manager role at Facebook. Those roles put him close to the work of developing products and deciding which changes mattered.

Facebook supplied the irritation that became Heap. When he wanted to investigate how people used a feature, he needed engineers to add tracking code. Then that code had to reach users. Then enough new information had to accumulate to make analysis possible. A question could be perfectly sensible and still arrive too late to have its evidence collected.

The frustration was especially striking inside a company with substantial analytics resources. If this was cumbersome there, what did it feel like at an ordinary online business? Movassate and Parikh pursued a different sequence: capture interactions first, then allow people to define and examine the events they cared about afterward.

The change gave curiosity more room. A product manager could discover an unexpected question without immediately discovering an empty historical record. That was the appeal of retroactive analysis. Yesterday’s activity could help answer today’s question, provided the relevant interactions had been captured.

One-third of a cent, before the first line

There was an awkward practical question beneath this idea. Collecting a broad stream of interactions might be useful. Storing and querying it might also make a young company spectacularly expensive to run.

Before coding, the founders estimated their infrastructure costs. Movassate later described starting with the size of an individual event, estimating activity during a visit and making conservative assumptions about storage. They narrowed their attention to useful signals such as clicks, submissions and page views rather than treating every movement of a mouse as equally valuable.

Their calculation produced an estimated memory cost of roughly one-third of a cent per visit per month. They took the economics to prospective customers and asked whether eliminating manual instrumentation would be worth a particular price. The replies helped justify building. The estimate later needed adjustment as compression, processor demands and infrastructure choices changed the arithmetic. Its purpose had been to make uncertainty small enough to act on.

Click here to understand what happened

The early product announcements give the idea a more human scale. In February 2013, Movassate wrote about graphing events over time and defining groups of users by what they had done. The language was direct, including an admission that the queries did not yet read like natural English. This was software being worked on in public, with its rough edges still visible.

That June, Heap brought its approach to native iOS apps. Movassate laid out the old routine for investigating an untracked interaction: change the app, submit it for approval, wait for release and then wait for information to accumulate. His list lingered on the waiting. Anyone who has watched a perfectly reasonable request disappear into a release schedule can appreciate the comic timing.

The Event Visualizer followed in February 2014. Users could click around their own website to define events for analysis. It addressed a specific imbalance: the people who needed the numbers often lacked the coding skills or engineering access to collect them. An interface that connected a visible action to its event also offered relief from bewildering internal labels. Naming a button should not require an archaeological expedition.

Hiring for the next version of a person

Heap joined Y Combinator’s Winter 2013 batch. By an October 2018 founders’ Q&A, it had 100 employees. Its hiring philosophy borrowed a mathematical metaphor inspired by Stanford professor John Ousterhout: “Hire for Slope Over Y-Intercept”. The emphasis was on how quickly someone could develop, alongside what they already knew.

The interviews tried to make that idea observable. Engineers spent a day building a feature. Sales candidates practiced and revised customer calls. Prospective managers met the team, looked for patterns and proposed ways to improve execution. A polished account of past accomplishments had to share the room with actual work.

In December 2018, Heap announced its first-place ranking among US small and medium businesses in Glassdoor’s 2019 Employees’ Choice Awards. The recognition came from employee feedback. It offered a public marker of the workplace during Movassate’s tenure, at a time when the staff was growing and the company had expanded into New York and Salt Lake City.

“Hire for Slope Over Y-Intercept”Heap’s hiring value, described by its founders in 2018

A customer before a sales pitch

Movassate’s business advice often returned to the distance between making something and finding out whether anybody wanted it. He urged technical founders to test willingness to pay before disappearing into development. In his own account, he had been too eager to build and should have spent more time talking to prospective buyers.

His approach to selling Heap also tried to reduce the delay before usefulness. An OpenView case study described a sales process intended to help prospects obtain insights while evaluating the software. Instead of leaving implementation until after a contract, the product could begin demonstrating what it did during the conversation.

He also argued for hiring customer success before sales, a subject of his 2017 SaaStr interview. These choices put attention on whether people could get value from the product and keep using it. A purchase was one event in a longer relationship. For someone building a company around understanding behavior over time, that was a consistent place to look.

Finding a successor on purpose

Ken Fine joined as president and chief operating officer in 2019, following a $55 million Series C. His responsibilities covered sales, marketing, operations and customer success. The appointment gave Movassate an experienced operating partner as Heap prepared to expand.

The succession became official in 2021, with Fine taking over as CEO on April 2. Heap said it served more than 6,000 companies at the announcement. Movassate’s chairman remit would cover strategy, the board and selected initiatives with the leadership team, rather than daily operations.

Later that year, under Fine, Heap announced a $110 million Series D at a $960 million valuation, taking total funding to $205 million. Those are company financing figures, belonging to a particular moment. They also show the succession taking place while the business continued to attract capital and invest in its product. Movassate’s eight-year turn as CEO had ended; the company’s work had not.

The questions keep changing

Contentsquare announced its agreement to acquire Heap in September 2023 and subsequently completed the acquisition. Heap’s capabilities now sit within Contentsquare Product Analytics. The company Movassate helped start has become part of a broader effort to connect different views of a customer’s online experience.

His public thinking had already moved beyond collecting more information. He wrote about bias in the questions people choose, the need to communicate uncertainty and the danger of treating artificial intelligence as a decorative promise. A large dataset still leaves the difficult work of interpreting it responsibly. The interface has to help people understand what an answer actually warrants.

That concern gives his story a useful ending without requiring a final verdict. The boy automating files, the product manager waiting for tracking code and the founder preparing a successor were tackling different problems. Each involved a piece of work held up by an avoidable dependency. Movassate kept looking for ways to shorten the wait. Eventually, that meant making room for someone else to lead.