The B2B Intake Form Has Been Broken Since Typeform Launched

01 /The Concept

Surface Labs is a Y Combinator graduate from the summer 2023 batch. It has raised $500K so far and is still in the pilot phase with early customers – but the underlying idea is sharp enough to warrant attention.

The product: a platform for building and deploying intake forms on B2B company websites.

The target customer: B2B sellers who want to capture information about website visitors – to qualify them and hand them off to sales. The trigger for filling out a form might be a request for a product demo, a content download, a consultation booking, or an early-access waitlist signup.

Surface Labs' stated objective is conversion rate improvement – specifically, maximizing the percentage of people who open a form and actually complete it. A dedicated AI engine manages the form experience in real time to push that number up.

The fundamental tension in form design is well known: more questions yield richer lead data but fewer completions; fewer questions get more submissions but leave sales reps with less to work with. Surface Labs attacks this tension through a combination of adaptive sequencing and real-time contextual nudging. The AI dynamically adjusts which questions get asked – and in what order – based on how each respondent has answered so far. Conditional logic in forms isn't new; what is new is that Surface Labs does this automatically, in real time, optimizing simultaneously for completion rate and information yield rather than requiring someone to hardcode branching logic upfront. In parallel, the platform surfaces relevant product facts at each step – social proof, use case examples, capability highlights – chosen by the AI to maximize the likelihood that the respondent continues. The goal is to keep motivation high through the full flow, not just at the entry point.

Surface Labs claims a 70% improvement in form completion rates compared to conventional tools like Typeform or Google Forms.

02 /Why It Matters

Forms might sound like a sleepy corner of the SaaS world, but the numbers tell a different story. Typeform – one of the category leaders – raised $135M at nearly a $1 billion valuation in a market downturn year when startup valuations were collapsing. By that point it had 125,000 company customers and $70M in ARR, up 3x from 2018.

So the market is real and the money is there. The question is whether AI can open up new advantages that allow challengers to compete with or displace incumbents.

Surface Labs' approach – adaptive sequencing plus real-time contextual nudging – is one answer. But a more radical answer is already being demonstrated elsewhere: skip the form entirely.

The core job to be done isn't "have someone fill out a form" – it's "extract useful information from a website visitor." A form is just one mechanism for doing that. A well-run qualitative interview by a skilled interviewer gets the same information more naturally, often more completely, and with higher engagement.

AI-powered interview platforms are now good enough to replicate that experience at scale.

Outset, [covered here](/review/insajty-dvigatel-biznesa) in November, also graduated from the same Y Combinator batch as Surface Labs – and has since raised $4.9M. Outset built a platform for AI-conducted user research interviews: given a list of questions to answer, the AI interviewer conducts a free-flowing conversation, adapting based on each respondent's answers rather than following a fixed script. The result is the same information a form would collect, gathered through dialogue.

Orson, [covered previously](/review/horoshie-istorii-prinosjat-horoshie-dengi), raised $8M for a related idea: an AI interviewer that conducts video conversations with customers about their product experience, then automatically edits the footage into short highlight reels for use in marketing. The first half of Orson's platform – the video-native AI interviewer – solves the same data collection problem Surface Labs addresses, but in a format that can achieve far higher engagement.

Both Outset and Orson replace rigid form completion with dynamic conversation, and both use digital personas as interviewers – which further improves completion rates relative to either static forms or text chat.

The technology stack needed to do all of this well now exists. The remaining work is assembling it into a product that's simple enough to deploy without a team of engineers, and smart enough to adapt across different use cases.

03 /Opportunities

The direction: build the AI platform that makes today's form tools obsolete.

The goal isn't just better forms – it's collecting meaningfully more information from meaningfully more visitors. The ideal product takes a list of questions the company wants answered, and figures out autonomously how best to get those answers from each individual visitor: structured questions, open-ended conversation, text chat, video, or some intelligent combination.

All the technology to do this exists. What the market is waiting for is a product that wraps it in an experience simple enough for non-technical teams to deploy and iterate on.

The market ceiling is at least $1 billion – Typeform alone is nearly there. And Typeform isn't standing still. Neither are the well-funded challengers already in motion. But the AI-native version of this product hasn't been built yet at scale. That's the opening.