Signing up is a trade. The visitor pays in effort, personal data, and a small amount of trust, and receives — eventually, maybe — something worth having. When a signup page underperforms, it's rarely because people didn't want the thing. It's because you priced the trade badly, and they did the arithmetic.
So the useful question isn't "how do I raise conversion." It's: what am I asking for, what am I offering in return, and does the exchange look fair at the moment I ask?
What you're actually charging
List every cost your signup imposes. Not fields — costs. They're different.
Effort. Keystrokes, decisions, page loads. A dropdown with forty options costs more than a text field.
Data. Not all data is equally expensive. An email is cheap. A phone number is expensive, because the visitor is pricing in the sales call. A company size or budget range is expensive, because they're pricing in being qualified out.
Commitment. A password creates an account, and an account is a relationship. Credit card details create a future obligation the visitor now has to remember to cancel.
Risk. Every field they don't understand the purpose of gets filled in with a worst-case assumption.
Now list what you offer in return, at that moment. Not what your product does eventually — what the visitor gets in the next thirty seconds. On most underperforming signup pages, the honest answer is "a confirmation email," and the trade is obviously bad.
The four asks that break signup
The password before the value. You've asked someone to create a credential for a service they haven't experienced. Most flows can let people do something first — configure, preview, build — and create the account when they have something worth saving. The account then protects their work rather than gating your product.
The credit card on a free trial. This is the single largest, most defensible drop in most SaaS signup flows. It filters aggressively, which is sometimes intentional. But be honest about what it filters for: not the people most likely to love the product, just the people most confident up front. If you require a card, say plainly what happens at the end of the trial and how to stop it, on the same screen. Doubt about cancellation is the real cost, not the card.
The qualifying questions. Company size, role, budget, "how did you hear about us." Each one tells the visitor that a human is going to evaluate them, and some decide not to be evaluated today. Every qualifier you place before signup buys you lead quality at the price of lead quantity — a legitimate trade, but you should be making it deliberately, not because the form template had those fields.
The verification wall. Email confirmation before any value at all. The visitor leaves your product, enters a different application, hunts for a message that may be in spam, and returns — if they return. Deferring verification until it's needed keeps the momentum you just earned.
Reading your own signup drop-off
Three readings, in order.
Do they arrive and never start? Then the page failed before the form. The value isn't clear, or the page is asking them to commit before it has earned it. The form isn't your problem.
Do they start and stop? Then a specific field is your problem, and it is almost never the first one. Watch where the typing ends.
Do they submit and never come back? Then the failure is after signup — verification, an empty first screen, an onboarding that doesn't produce anything worth the trade they just made. This is the leak most teams never look for, because their metric stopped at "signup."
If mobile lags desktop badly, suspect the keyboard: a numeric field opening a text keyboard, an input hidden behind the on-screen keyboard, autofill silently failing because the inputs lack `autocomplete` attributes.
Re-pricing the trade
Delete every field that doesn't change what happens next. Purpose is the test, not count.
Move the account to the moment it protects something. Let people do first, save second.
If a field is expensive but necessary, buy it: explain what it's for in one line, next to the field. "So we can size your plan" costs you nothing and prices the ask honestly.
Put the value at the ask, not above it. "Start your project" beats "Create account" because it names what the visitor receives, not what your database does.
Then change one thing, and look specifically at the transition you were trying to move.
The part you can't reason your way to
Everything above narrows the search. It won't tell you which field on your form loses people, because that depends on your visitors, your wording, your device mix. Field-level abandonment is invisible from the outside — two signup forms with identical structure can bleed at completely different inputs.
Defrixa's free scan will score your signup page's structural friction and name its biggest issue without you installing anything. With the snippet live, it identifies the field where people stop typing, and confirms whether removing it recovered them.
Common questions
It's a deliberate trade — fewer signups, higher intent among them. If you do it, remove the cancellation doubt on the same screen. Most of the drop comes from uncertainty about getting out, not from the card itself.
It removes password and verification friction in one step, which is meaningful. It also introduces a new decision ("which account, and what will they see?"). Offer it alongside email rather than instead of it.
As many as change what happens next, and no more. A three-field form with a phone number can cost more than a five-field form without one.
It's a trade problem. They paid, and what they received in the first minute wasn't worth it. Look at the screen immediately after signup before you touch the form.