Every form on the internet is a guess. Someone sat down months ago, imagined the person who would eventually fill it out, and froze that guess into a fixed list of fields in a fixed order. Then you arrive, and you are not that person.
You type something genuinely interesting into box four. Nothing happens. The form has no way to say "wait, say more about that." It moves you along to box five, which is a dropdown, and none of the options are right, so you pick the one that is least wrong and keep going.
That is the whole limitation, and it is bigger than it looks. A form cannot ask a follow-up question.
What forms quietly throw away
The visible cost of a bad form is abandonment: the person who gets three fields in, decides this is not worth it, and closes the tab. That one at least announces itself.
The expensive cost is invisible. It is the answer that got flattened on the way in.
Someone selects "Website redesign" from a dropdown because it is the closest option, when what they actually need is to stop paying three vendors for a stack that no longer talks to itself. Someone types "flexible" into the budget field because the real answer is "it depends what you tell me is possible," and there is no box for that. Someone leaves the goals field empty, not because they have no goals, but because the field is four lines tall and their answer is not.
None of this registers as a problem. The form submits. The record looks clean. Everyone downstream now works from a version of the truth that has had the interesting parts sanded off, and nobody knows which parts are missing, because the missing parts never touched the database.
The dropdown option "Other (please specify)" is the entire industry admitting this in public.
The follow-up question is the actual job
Sit in on a good discovery call and count the questions. Very few of them are the ones on the list.
The list only gets you moving. "Tell me about the business." Then the person says something like "we do most of our sales over the phone, but the site is where people check us out first," and everything worth knowing is sitting inside that sentence. A good interviewer stops right there. Why the phone? Who picks up? What do people ask when they call? What do you wish they already knew before they dialed?
Four questions that did not exist sixty seconds earlier, and now you understand the project.
This is not a soft skill. It is where the information density is. The prepared questions collect the facts anyone could have guessed; the follow-ups collect the reason the project exists at all. A form, structurally, cannot do the second thing. It is not that forms are badly designed. It is that they are a broadcast medium pretending to be a conversation.
So we built the one that asks back
Intake is our answer to that, and our own studio runs on it.
A client gets a link. They follow it and an assistant is already there, ready to talk about their project. No eleven-page questionnaire, no progress bar quietly lying about how much is left. They talk, it talks back, and if they have to stop and go pick up a kid from school, the same link drops them back into the same conversation exactly where they left off. Forms lose your place. This one holds it.
It covers the ground a good discovery call covers: the business and what they are actually trying to do, then brand and style, then the pages the site needs and what belongs on each one. When an answer opens something up, it follows the answer instead of the script.
Somewhere in the middle, most clients say some version of "we have a brand guide for that." Instead of being told to email it to somebody, they drag the PDF straight into the conversation, and it gets read. The palette, the typefaces, the notes on voice, the tagline they forgot they had already written down: all of it comes back a moment later as things to confirm rather than things to retype. Same for a Word doc of half-finished page copy, or a phone photo of the sign out front with the logo on it.
Fifteen minutes, give or take. Nobody had to guess what we meant by a field label.
Conversation is the interface, not the storage format
This is the part most chat-shaped products get wrong, so it is worth being precise about.
A conversation is a wonderful way to collect information and a terrible way to keep it. Nobody on a project team wants to scroll a transcript to find out what the client said about the booking flow. A transcript is not a brief.
So Intake does both at once. While the conversation is happening, every confirmed answer is written into a structured record: content filed by section, progress tracked per section, phases that advance as the discovery genuinely advances. The client experiences a conversation. The team opens a dashboard and finds the same organized brief a perfect form would have produced, except with the parts a form would have lost still attached to it.
That is the trade everyone assumes they have to make, and they do not. You can have the front end of a conversation and the back end of a form. You just have to build both, and stay honest about which one is the deliverable. The structured record is the product. The chat is how you got there.
Where the pattern belongs, and where it does not
We are not going to tell you every form should be a conversation. Most of them should not be.
If the questions never change, a form wins. Nobody wants to chat with a checkout. A shipping address has exactly one correct shape, you already know all six fields, and turning it into a dialogue would be an insult to everyone's time. Same for a login, same for a date picker, same for anything where the person already knows precisely what they want to say and simply needs somewhere to put it.
The pattern earns its keep when the questions depend on the answers. When the thing being collected is qualitative and messy and lives in someone's head in no particular order. When documents are involved. When the person filling it out does not know the vocabulary yet, and the form is quietly testing them on it. Client onboarding, project briefs, technical scoping, intake for anything professional: those are all interviews that got compressed into forms because a form was the only thing the web knew how to do.
That stopped being true a while ago.
The front door, too
The same idea runs our own contact page. You do not fill in name, email, message, submit, and then sit there wondering whether it went anywhere. You describe what you are after, our assistant asks whatever it needs to ask, and the conversation arrives on our end as a real inquiry with the context already attached to it.
It works in the other direction as well. If you send us a brief through our hiring page, the conversation reads the brief first and confirms what it already knows instead of making you say all of it twice. Few things are more irritating than a form asking you a question you have already answered on the same website.
Intake is live and free to start. If you run a studio, an agency, or any business whose first real step with a new client is a document full of questions, take it for a drive. And if your version of this problem lives somewhere else in the business entirely, that is the kind of thing we build.
Forms were never the point. The answers were.