At the companies where Sam Clemens worked, the sales demonstration had a way of becoming an engineering problem. Someone needed a convincing version of the product. Someone else needed the real product to keep working. In one version of the recurring trouble that led to Reprise, sales representatives deleted material from production environments. A meeting intended to win a customer could leave a little housekeeping for everybody else.
Clemens arrived at that problem from the building side of software. His co-founder Evan Powell came from sales. Together with Bryan Stevenson, they founded Reprise in 2020 around an irritation they had encountered before: demonstrating a product required an unreasonable amount of work on the machinery behind the demonstration.
There is a modest comedy here. A company can employ people capable of building complicated software, then struggle to show another person what it does. The product has to work. The example data has to make sense. The salesperson needs freedom to explore, preferably without accidentally rearranging the furniture in somebody else’s account.
For Clemens, now Reprise’s co-founder and CEO, this became a business. It also fits a longer preoccupation: arranging the work so that ordinary people can succeed without an extraordinary rescue.
The mathematician’s objection
Clemens studied applied mathematics and economics at Yale and earned his Harvard MBA in 2004. The word “applied” matters to his account of himself. He describes his interest as systems operating in the real world, where people introduce variation and designs must accommodate it.
His argument about mistakes follows from there. When somebody fails, he wants to examine the conditions surrounding the failure: training, incentives, hiring, the arrangement of the work. Blaming the individual too early closes the investigation precisely where a product designer might want it to begin.
This is a demanding position for a manager. It moves responsibility back toward the people who set up the system. A seller who cannot prepare a useful demonstration may need a better environment, rather than another reminder to prepare harder. The distinction has consequences for what a company decides to build.
He carries a similar caution into automation. A process can become efficient while producing disappointing work. Make that process automatic and the disappointing work can arrive faster. For someone selling software that removes manual preparation, it is a useful complication: speed requires a destination.
A résumé that keeps returning to the product
His working life has taken him through Elance, the business now known as Upwork, product leadership at BzzAgent and HubSpot, and the co-founding of analytics company InsightSquared, where he was chief product officer. Alongside those operating jobs came periods in venture capital at Greylock and Accomplice.
That combination places him on several sides of the startup table. A product leader has to decide what to make. An investor has to decide which company deserves the means to make it. A founder eventually discovers that the table also needs assembling.
In January 2019, appearing on the Giant Robots Smashing Into Other Giant Robots podcast as an Accomplice partner, Clemens discussed company stages, minimum viable products and leaving a business he had founded. Those subjects put his later return to operating in context. He had already lived through the question of when a company needs a different kind of contribution.
His teaching adds another setting. Harvard Business School named him among its Rock Executive Fellows for 2022-2023, an appointment for practitioners contributing to entrepreneurship teaching and programming. He has also taught Launching Technology Ventures, a course concerned with building and scaling young companies. The classroom asks him to explain choices that operators usually make amid competing demands.
The pitch deck arrived after the product
Reprise’s first financing had an unusually small requirement for presentation materials. The founders raised $3 million from Accomplice and Glasswing Ventures with prototype technology and no signed customer. Clemens later recalled conducting the fundraising by telephone, without a pitch deck.
Pendo became the first enterprise customer. Others followed. Then customers began showing their own Reprise demonstrations in board meetings, putting the product in front of investors without the founders having to arrange every introduction. When a deck finally appeared, there were actual customers and demonstrations to put in it.
In February 2021, Bain Capital Ventures led a $17 million Series A. That November, ICONIQ Growth led a $62 million Series B, bringing announced funding to $82 million. The investors also included WndrCo and Trevor Noah. Enterprise demo software is perhaps an unexpected place for the entertainment business to turn up, although both professions appreciate a presentation that survives its audience.
At the Series B announcement, Reprise reported annual recurring revenue growth above 900 percent year over year. That figure belongs to the young company in November 2021. The useful part of the financing story is the sequence: an experienced team proposed an answer to a familiar problem, customers used it, and a larger group of investors backed its expansion.
A refresh button with a job to do
Clemens described two practical needs in 2022: something reliable to present during a live call, and something interactive to send before or afterward. A slide deck could explain a product, but buyers increasingly expected to try their way through it. A shared application could provide access, but it also brought shared consequences.
One revealing detail in his account of Reprise is what happens when a demonstration ends. In the code-capture approach he described in 2023, a salesperson can explore a recreated application without leaving changes behind for the next presenter. Refreshing restores the original state. The next meeting need not inherit the last meeting’s experiments.
He also distinguished between the needs of a junior account executive following a prepared route and an experienced seller who wants to roam through features. Reliability has to leave room for both. A tightly guided tour and an exploratory demonstration serve different conversations, even when they begin with the same software.
The market widened beyond the software companies the founders first imagined. Banks, retailers and networking hardware businesses also made software they needed to show. Clemens’s observation was straightforward: the relevant customer could be any sufficiently large company making software. A product demonstration turns out to have a fairly broad family tree.
The crew gets larger
In August 2022, Clemens described a team that had grown from seven people to more than 130 in two years. He called rapid growth a challenge of understanding which skills a company needed at each stage. Hiring was only part of the work; the organization had to keep changing shape.
“No matter what the business it’s always the team.”
Sam Clemens, 2022
His advice to his earlier self was to concentrate on assembling that team. It is a fairly unsentimental answer from someone whose work includes investing and teaching. There is little room in it for the founder as a solitary source of brilliance. Colleagues determine what the company can actually do.
He has also argued for remote working as a way to reach talent. His 2022 advice to a fellow entrepreneur acknowledged the difficulty of managing without walking around an office, while making the case that access to people justified the trade. The recurring managerial question is practical: what arrangement gives this business a better chance of finding the people it needs?
When the demo starts thinking for itself
By October 2025, Clemens was discussing a new source of unpredictability: AI products whose behavior could change in front of a customer. A demonstration designed to show intelligence could produce an answer the presenter had never intended to discuss. Preparation had acquired another variable.
His response keeps a place for both software and people. Reprise can help create controlled environments; the person presenting still has to understand the buyer. Automating technical preparation leaves the work of judgment and relationship building with the humans in the meeting.
He has a compact phrase for the desired effect: “take the prayer out of the process.” It is an ambition that permits a little humility. Sellers should be able to rely on the demonstration while they attend to the conversation happening around it.
The company’s current tools extend into AI-assisted data preparation, application clones and demo creation through agents. That development gives his old concern a new setting. As more preparation becomes automatic, somebody still has to decide whether the resulting demonstration explains the product honestly and answers a customer’s actual question.
And then there is the tractor
Outside the software business, Clemens has described helping his wife on her commercial farm, driving a tractor and using a chainsaw on weekends. It is a pleasingly concrete addition to a career full of products, portfolios and classroom cases. Some work leaves a result that requires no dashboard.
He also found room in an otherwise technical 2022 interview to ask whether England’s young football team could win that year’s World Cup. Even a conversation about enterprise software can accommodate an entirely different set of hopes.
The farm detail is a good place to leave him, without turning it into a theory of everything. He builds software, teaches entrepreneurship, has invested in other founders, and sometimes helps with work outdoors. Across his business career, he has returned to the moment when a product meets the person expected to use it. Reprise grew out of one such meeting going wrong too often. His answer was to give the next presenter better machinery.