The useful question · What happens before checkout?
Sampo Parkkinen has spent much of his career looking at the awkward few seconds before a purchase. A shopper enters a store, confronts a crowded shelf and hesitates. A browser opens six tabs, reads three contradictory reviews and quietly leaves. Retailers tend to obsess over the final click. Parkkinen is interested in the uncertainty that comes first.
His thesis can survive without a slide deck: people do not enjoy being sold to, but they often need help buying. The distinction sounds small. In practice, it changes where a company puts its attention. The hard sell begins with inventory. Helpful commerce begins with a question.
That idea now sits inside Revieve, the enterprise software company Parkkinen co-founded in 2016. Its platform lets beauty brands and retailers build guided digital experiences using artificial intelligence, computer vision and augmented reality. A shopper can share preferences, use an image-based experience, explore products or try a look virtually. The retailer can respond with a narrower, more intelligible set of choices.
“Nobody wants to be sold to, like, I don’t want to be sold to.”Sampo Parkkinen, 2026
The second draft of a retail obsession
Revieve was not Parkkinen’s first attempt to make shoppers legible to software. In 2007, he co-founded RapidBlue Solutions with Gavin Weigh and Samuli Siivinen. RapidBlue worked in shopper analytics, translating movement inside physical retail spaces into information a business could use. Parkkinen served as chief operating officer. ShopperTrak acquired the company in 2013, and he stayed on as a director of product management responsible for mobile and location-based products until 2016.
The chronology matters because Revieve did not arrive as a tourist in retail wearing an AI badge. Parkkinen and his co-founders had already spent years studying how physical stores work. They had built together, sold together and remained close enough to choose another company together. Startups celebrate novelty, but familiarity can be an asset when it means colleagues already know how one another behave after a missed target or an inconvenient customer request.
The seed of Revieve came from a personal shopping frustration after Parkkinen moved from Helsinki to Chicago. He found it needlessly difficult to understand which skincare products suited him. The interesting part was not simply that the shelves held too many choices. It was that the guidance available in a store had not made the journey online.
Parkkinen saw a distribution decision hiding inside the problem. Revieve could have become another consumer brand. Instead, the founders chose to supply technology to existing beauty companies and retailers. Their reasoning was practical: if useful personalization should be available wherever people shop, the infrastructure needed to travel through many brands rather than demand that everyone visit a single destination.
A beauty counter made of software
The old beauty counter did something a product grid could not. A good adviser listened before pointing. They reduced the shelf to a manageable number of possibilities and explained why each one belonged there. Ecommerce made inventory searchable and purchasing frictionless, but it often left the thinking to the customer.
Revieve tries to put that missing conversation back into the journey. Its experiences can appear on a retailer’s website, inside an app, on a store screen or across other digital touchpoints. The customer supplies context. The system turns that context into recommendations or virtual experiences. The brand gains consented information about what customers are trying to achieve, rather than merely recording what they bought last time.
Parkkinen’s language around this is strikingly unromantic. He describes AI as a tool that should solve a business challenge. This is a useful position in a market where the acronym sometimes arrives before the problem. The customer does not wake up longing for computer vision. They want to make a confident choice without donating an afternoon to research.
The business model had to learn too. Parkkinen has explained that Revieve began with fees connected to recommended-product sales, then moved toward an enterprise SaaS structure. The evolution reflects a company discovering where it creates value and how customers prefer to buy that value. The polished consumer interaction is only the visible layer. Underneath sit catalog work, integrations, configuration, measurement and the patient business of onboarding an enterprise client.
Parkkinen treats that onboarding as strategic work, not a handoff after the contract is signed. A beauty company brings its own assortment, language, goals and customer journey. Software that behaves sensibly for one catalog cannot simply be dropped onto another and expected to perform with equal grace. The provider and client have to decide what the experience should ask, where it should appear and how its recommendations fit the brand. White-glove attention may sound quaint beside machine learning, but enterprise software often succeeds through precisely this kind of human patience.
His example of RoC Skincare makes the point concrete. The brand uses a Revieve-powered experience called AI Skin Insight to help visitors explore relevant products and advice. Parkkinen describes tools like this as serving two jobs: bringing in new customers and giving existing customers a reason to return. They can also reveal, with consent, what shoppers are looking for before purchase. A receipt tells a company what left the shelf. A guided conversation can explain the intention that sent the shopper there.
“Show the customer value and show them that you care about them.”Sampo Parkkinen, on why guidance can precede loyalty
Remote before remote required a policy
Revieve has operated as a remote-first company since 2016. It was an unusually early choice, made four years before millions of office workers met the mute button. The company built teams across countries while serving customers on multiple continents, with hubs in Helsinki and Chicago and product development in Valencia.
A distributed company makes an apt companion to a distributed product. Revieve’s software has to appear wherever a shopper happens to meet a brand. Its people have to collaborate wherever they happen to live. Both arrangements depend on reliable connections, explicit processes and enough context to prevent distance from becoming confusion.
Parkkinen remains an eager user of new tools. In a 2026 podcast, he said Claude had become central to his workday, including presentations and internal research. He praised connected workflows and teased software companies that merely place an assistant on top of an old product. The complaint reveals a product operator’s impatience: a feature should change the work, not decorate it.
The investor keeps asking for focus
Alongside Revieve, Parkkinen serves as a venture partner at Gorilla Capital, an early-stage investor in Nordic technology companies. The firm describes its preferred companies as camels: resilient businesses that put customers first and aim for sustainable profitability. The animal is less theatrical than a unicorn and considerably better equipped for a long dry spell.
His advice to early founders follows that temperament. Before product-market fit, he emphasizes customer learning, focus and capital efficiency. He has warned founders against treating polite encouragement as validation. A prospective customer saying an idea sounds nice proves very little. A specific problem, a committed user and repeated behavior are firmer material.
Begin with uncertainty. Find the moment when a customer becomes confused, delayed or poorly served. That is where useful software earns its place.
Learn before optimizing. Early pricing, channels and features are hypotheses. Treat customer behavior as the sharper editor.
Make the system helpful. Remove gates, explain recommendations and regard onboarding as part of the product experience.
This also explains Parkkinen’s preference for focus. A young company cannot learn clearly while solving five problems for five audiences. Narrowness is not a lack of ambition. It is a way of hearing the signal before the company has enough money, customers or time to survive noise.
The transaction is the middle, not the ending
Revieve now spans a wider set of beauty experiences than the skincare question that started it. Yet Parkkinen’s central argument has remained steady. A retailer should become useful before checkout and remain useful after it. The transaction sits inside a relationship, rather than serving as its entire plot.
That ambition gained a tangible marker in 2025, when Revieve and Paula’s Choice received a CommerceNexty Award for an ecommerce experience built around an AI-powered analyzer and personalized product matching. Parkkinen credited the partnership and its shared attention to the customer. Awards are pleasant punctuation. The more revealing outcome is that the original thesis survived contact with a large brand and an actual shopping journey.
His official biography calls him a lifelong learner, a father of three and someone who enjoys sports. The phrase “lifelong learner” can become corporate wallpaper, but in Parkkinen’s career it has a visible trail: shopper analytics became product management, physical retail became digital guidance, an acquired company became a second founding team, and operating became investing without ceasing to operate.
The shelf will keep getting larger. AI will add answers, agents and, inevitably, fresh confusion. Parkkinen’s contribution is a sturdy test for whatever arrives next. Does the technology help a person choose, or does it simply give the seller another way to shout? A useful system listens first.