The Startup Teaching Software to Explain Itself
How two Dashlane veterans turned the frustration of writing a help article nobody reads into a knowledge platform that answers back.
Somewhere in the archives of every software company sits a help article that no human has read to the end. It was written in a hurry, for a version of the product that has since changed twice, and it assumes you already know the thing you opened it to learn. Alexis Fogel spent years at Dashlane staring at documents like that - trying, and mostly failing, to explain how to install a browser extension across a dozen operating systems and browsers. The gap between what the product could do and what a person could figure out kept nagging at him. Stonly is his answer to it.
Founded in 2018 and based in Paris, Stonly calls itself a knowledge platform for customer service. In plain terms: it helps companies replace static help pages with interactive, step-by-step guides that adapt to whoever is reading them. Instead of a wall of text, a customer gets one question at a time - "What are you trying to do?" - and is walked to the specific fix for their situation. Behind the scenes the same guides feed a searchable knowledge base, an AI answer engine, and playbooks that sit next to support agents while they work.
01The problem it actually solves
Support has a math problem. A customer has a question, the answer exists somewhere, and the two rarely meet without a human in between. Companies respond by hiring more agents, who then spend their days copy-pasting the same answers, or by publishing more articles, which pile up faster than anyone reads them. Both are expensive. Neither feels good to the person stuck at 11 p.m. trying to reset a thermostat.
Stonly's founders looked at that loop and asked a narrower question than "how do we fix support?" They asked: what if the help content asked you a question first? A refund, a broken install, a device that won't pair - each has a decision tree hiding inside it. Map the tree once, and software can walk a million people down the right branch without a single agent touching it.
answer accuracy
success rate
Peer Insights
02From FAQ to flow
The clearest way to understand Stonly is to picture the before and after. The old help article is linear: read paragraph one, hope it applies, scroll to paragraph nine, discover it was written for a different plan. A Stonly guide is a branching path - you click, it responds, and the irrelevant 90% never appears.
Old: static article
Stonly: interactive guide
Crucially, the people who build these guides are not engineers. Stonly is deliberately no-code: a support manager, a marketer, or an operations lead defines the steps, then adds text, images, GIFs, video and buttons in a web editor. Conditional logic - "if they picked Android, skip to step four" - is set with clicks, not scripts. That design choice is the whole business. The teams closest to customer pain can fix the content the moment they notice it's wrong, instead of filing a ticket with the web team and waiting a sprint.
Imagine if every time you were trying a new service, you could instantly and personally know how you could use it.Alexis Fogel, co-founder & CEO
03The products, briefly
Stonly has grown from a single tool into a stack. Interactive Guides remain the core. On top of them sit Knowledge Base software - a self-service help center customers can search or reach through a chat-style widget - and AI Answers, which generate replies grounded in the company's own structured content rather than the open internet. That grounding is the pitch: an AI that quotes your refund policy correctly because it can only draw from guides your team wrote. The company also ships checklists for onboarding, conditional contact forms, surveys, and agent-facing playbooks that live inside the help desks support teams already use.
04Who's paying for it
Stonly sells to mid-market and enterprise support, success and operations teams - the people who feel every deflected ticket in their budget. Named customers span consumer hardware and heavy industry: SimpliSafe, Devialet, Tonal, Fastned, Siemens, TELUS, Carta, King, RemoteLock and the French energy retailer ekWateur, among others. The common thread is a product complex enough that customers get stuck, and a support volume large enough that answering each person by hand doesn't scale.
The business model is textbook B2B SaaS - subscriptions priced by usage, seats and modules, with room to expand as a customer moves from a single self-service knowledge base into AI answers and agent playbooks. The land-and-expand motion is helped by the integrations: Stonly plugs into Zendesk, Freshdesk and Front, so a team can add interactive guides without ripping out the help desk they already run.
There's a quieter reason the model works. Support content is never finished. Products change, edge cases surface, a wording that confused thousands gets caught and fixed. A tool that makes editing painless earns its keep every week, not just at renewal. Stonly's early customers tended to start with one stubborn, high-volume problem - the install that always failed, the setting nobody could find - build a guide for it, watch the tickets drop, and then reach for the next one. Deflection is measurable, which makes the return on a Stonly subscription unusually easy to put on a slide.
05How it's different
The customer-support tooling market is crowded, and most competitors approach knowledge from one angle. Digital-adoption players like Whatfix and Chameleon focus on in-app walkthroughs. Traditional help desks - Zendesk, Freshworks, Intercom's Fin - bolt a knowledge base onto a ticketing system. Newer entrants like Knowmax, Featurebase and Supademo circle the same self-service space. Stonly's wedge is the interactive guide as the atomic unit: the same decision-tree content powers the customer-facing help center, the AI answers and the agent's screen at once. One source of truth, three audiences.
The knowledge base isn't dead. It just stopped being a document and started being a conversation.The Stonly thesis, paraphrased
06The people and the money
Stonly's credibility starts with its founders. Alexis Fogel co-founded Dashlane, the password manager used by millions, and served as its chief product officer before leaving - by his own account not out of boredom but affection, wanting to start something he genuinely believed in while still loving the company he was leaving. He built the first version of Stonly as a side project, then recruited Krzysztof Dabrowski, who had led UI and UX design at Dashlane, as co-founder.
Investors followed the pedigree. Accel led a $3.5 million seed round in early 2020. Two years later, Northzone led a $22 million Series A with Accel returning - bringing the total raised to roughly $25.5 million. The angel list reads like a founder's group chat: Miro chief executive Andrey Khusid, Y Combinator group partner Nicolas Dessaigne, Personio's chief product officer, and Dashlane's former CEO Emmanuel Schalit.
07Where it fits
The expertise on display is less about machine learning than about a craft that's easy to underrate: breaking a scary, ambiguous process into one calm step at a time. That's a product-design discipline, and it's the same one Fogel and Dabrowski practiced at Dashlane, where the whole job was coaxing ordinary people through the deeply unglamorous chore of managing passwords. Stonly is that instinct pointed at a bigger surface - every moment a company has to explain itself to a customer.
Zoom out and Stonly sits at the intersection of three trends every software company is living through: rising support costs, customers who expect instant answers, and AI that promises to deliver them but often invents facts instead. Stonly's answer to that last risk is almost old-fashioned - keep humans in charge of the knowledge, let AI retrieve and phrase it. In a market rushing to hand everything to a chatbot, a platform that insists the source material be structured, correct and human-authored is making a specific wager about what customers will actually trust.
Whether that wager pays off depends on execution and on a category still being defined. But the underlying observation is hard to argue with: most software still explains itself with a wall of text, and most people still can't find the answer. The company that closes that gap - one clickable step at a time - has a real business. Stonly is betting it can be that company.