Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Conversion Rate Optimisation › Funnel & Checkout

Sign-Up and Trial-Start Flows

How to design sign-up and trial-start flows: what to ask now and later, verification, card or no card, consent, the first screen and what to measure.
Guideintermediate13 min readUpdated

A sign-up is a funnel of its own

Most sign-up flows are treated as a form at the end of a landing page or a pricing page. In practice they are a sequence of screens, each with its own reason to stop: a request for information the person does not have to hand, a verification email that does not arrive, a card form that feels premature, a setup question that seems irrelevant, a first screen that leaves them wondering what to do. Each step has a rate, and the flow as a whole has a result that only shows up later, in whether the new account goes on to use the product.

That second point is what separates sign-up work from ordinary form work. A lead form's job ends when a good lead reaches sales. A sign-up's job ends when the new user reaches value, and a flow that produces many accounts that are never used has not done its job. The comparison of product-led and sales-led growth names this as the characteristic failure of self-serve models: many sign-ups, few activated users. The AARRR framework puts the same idea in its stages, with activation, the first moment a customer receives the value they came for, standing between acquisition and everything that follows.

This guide sits between neighbours. The forms guide covers fields, labels, errors and multi-step layout on any form. The checkout optimisation guide deals with accounts forced on shoppers before an order, where the right answer is usually guest checkout. Here the account is the product, so the question is not whether to create one but how to create it with the least cost to the person and the most use to the business.

The steps from call to action to first screen

The steps vary by product, but most flows contain some or all of those in the table. Mapping the actual flow, screen by screen, is the first piece of work, because teams often do not know how many screens a new user passes through or what each asks. The funnel and checkout section explains how to find the step that loses the most people who should have continued; a sign-up flow is read the same way.

StepWhat it asksWhat it is forCommon failure
Call to actionA click on a plan, a trial or a free accountCarries the chosen plan into the flowA generic button that loses the plan choice
Account detailsA way to sign in and a way to reach the personCreating the accountFields for sales, marketing or reporting nobody uses in the first session
VerificationConfirmation that the address worksProtecting the account and later messagesA block on the product until an email arrives, with no help if it does not
Setup questionsRole, team size, use caseShaping the first experienceQuestions asked for reporting rather than to change what the person sees
PaymentCard details, where the trial or plan takes themStarting billingA payment step that appears without warning or with unclear renewal terms
First screenNothing; it givesStarting the first useful taskAn empty dashboard and a tour

Every step should be visible in measurement, so that its rate can be read. Where steps are spread across the website, an identity service and the product, that usually means agreeing which system records which step, which the measurement section below returns to.

Ask now, ask later

The rule for sign-up questions is the forms guide's rule for every form: each field is a cost the person pays and a reason to stop, so ask only what the next step needs. For a sign-up, the next step is the first session. Anything that does not change the first session can be asked later, inside the product, once the person has seen enough to want to answer, or it can be inferred from what they do.

Information given at sign-up is First-Party Data, and it is only worth collecting if something uses it. The questions that most often fail that test are the ones added for other teams: company size for sales routing, job title for segmentation, how the person heard about the product for attribution. They are not worthless, but they are rarely worth their cost at the point of sign-up. The conversion research hub works through an example of a trial sign-up where the company-details step loses the most people and a poll shows visitors do not know why the information is needed; the team does not simply remove the field, because sales uses it for routing, and writes a hypothesis instead. That is the right instinct. When a question must stay, say why it is asked; when it can move, move it to the moment it is used.

As general practice, these are the usual candidates for asking later:

  • Profile details that do not change the first screen, such as job title or phone number.
  • Team and company details that matter for sales routing or billing, asked when the person invites a colleague or approaches a plan limit, or inferred from a business email domain.
  • Preferences that the product can default sensibly and let the person change.
  • Attribution questions, asked once the account exists, where they cost nothing at the decisive moment.

A setup question that does change the first session, such as which of two main uses the person has in mind, can earn its place, because the answer makes the first screen more useful. Keep such questions few, make them quick to answer, and show the result straight away so the person sees why they were asked.

Verification and sign-in

Two decisions shape the start of the flow: how people sign in, and when the address is verified. Both are described here as general practice; the features of particular sign-in providers are outside the scope of this guide.

Offering an existing account from a widely used provider as a way to sign in can remove the need to invent and remember a password, and many people prefer it. Offering an email and password as well serves people who do not want to link accounts. Whatever the options, the form should make clear which one the person used last time, so returning users do not create a second account by mistake.

Verification protects the account and the messages that follow, but its timing decides whether it stalls the first session. Blocking the product until a verification link is clicked turns an email delivery delay into an abandoned sign-up. Letting the person start, and asking them to verify before anything that depends on the address, such as inviting colleagues, sending messages from the product or starting billing, keeps momentum. Where the product cannot work without a verified address, the waiting screen should say where the message comes from, offer to resend it and allow the address to be corrected.

Verification of an account address is not the same thing as double opt-in to a marketing list. The first confirms that the account belongs to the person; the second evidences consent to marketing email. Combining them in one message confuses both, which the next sections take up.

Card at sign-up or not

Whether a trial takes payment details at sign-up is one of the largest decisions in the flow, and it is a trade-off rather than a best practice. The glossary entry on trial-to-paid conversion rate describes it: a trial that takes a card converts automatically unless the person cancels, so it usually shows a higher rate on fewer starts, while a trial without a card converts only when the person actively chooses to pay, so it usually shows more starts and a lower rate. Comparing the two rates directly compares trial policies, not marketing.

The decision belongs to a test, and the card-required trial test sets it out: visitors randomised before the card step, both arms with identical renewal wording and the same reminder before the first charge, and the result read on paying customers per visitor, net of refunds, at a fixed horizon. Its guardrails are refunds and chargebacks, early churn and cancellation complaints. It also notes that regulators treat clear terms and easy cancellation for subscriptions as consumer protection matters, so the clarity of renewal terms is held constant and never tested. Trial length is a separate question with its own design, the free trial length test.

For the flow itself, the practical points are these. If a card is taken, say so on the call to action and the pricing page, not first on the payment screen. Build the payment step to the forms guide's standard, so that a test of asking for a card is not a test of asking for it badly. And state the trial length, the price that follows and how to cancel in the same place the card is asked for.

A sign-up is a natural moment to ask whether the person wants marketing email, and the most common mistake is to ask it badly. The list building and consent guide is clear about the pattern that works at account creation: one clear, unticked box that adds marketing consent, separate from acceptance of the terms, with an honest description of the mail that will follow. Pre-ticked boxes, consent bundled into the terms and vague wording are the failures it names.

The legal side is summarised only as that guide states it, and nothing here is legal advice. The UK Information Commissioner's Office publishes direct marketing guidance that explains the UK's rules on electronic mail marketing, including when consent is needed, the conditions under which existing customers can be contacted about similar products without fresh consent (often called the soft opt-in), and the requirement to offer a way to opt out in every message. Other countries have their own regimes. The list building guide's practical advice holds everywhere: record the source, the wording the person saw, the choice they made and the time, so the business can answer how each person came to be on the list.

Two further points keep the sign-up clean. Messages needed to run the account, such as verification and password resets, are a different stream from marketing; the SaaS industry guide notes that trial onboarding and upgrade prompts are commercial email when they promote the product, and that transactional messages are treated differently, so keep the streams and the records apart. And the marketing question should never be a condition of creating the account.

The first screen after sign-up

The first screen is where the sign-up flow ends and onboarding begins, and it decides whether the account will be used. A new user who lands on an empty dashboard with a tour has been handed a second task before receiving any value. A first screen that starts the first useful task, using what the person said in setup, and shows a result quickly is the hand-over the rest of the funnel depends on.

Onboarding itself is owned elsewhere. The customer retention strategy guide treats it as the first retention lever, because early loss is the largest and cheapest opportunity in most businesses, and frames the questions as what first value is, how long it takes to reach and what happens to customers who stall. The sign-up flow's part is narrower: carry the plan and the setup answers into the product, avoid any screen that stands between the account and the first task, and record when the person reaches the activation moment so the flow can be judged on it. Onboarding and welcome emails belong to the email marketing root and are not designed here. In a mobile app, the account often moves after the first useful task, and the guide to app onboarding and permission requests sets that order alongside the operating-system permissions the first session may need. The deferred account creation test is the design for checking that moving the account later raises activation without costing the accounts the business needs.

Measuring sign-ups by what happens next

Google's documentation of events lists sign_up among the recommended events, and the events and parameters guide explains that recommended events use the documented names and parameters, with the sign-up method as a parameter, and that the sign-up event is offered as a Key Event candidate. The key events guide recommends counting sign-ups once per session, so a double submission or a refreshed page is not counted twice. The subscription measurement guide assigns each lifecycle event to the system that sees it reliably: the sign-up to the product's database, the trial start and the first payment to billing, and activation to product data, each sent on to analytics.

With those events in place, read the flow as a funnel and judge it by its outcome.

MeasureWhat it showsRole
Step rates through the flowWhere people leave between the call to action and the first screenDiagnosis
Sign-ups per visitor to the sign-up pageHow many accounts the flow createsContext
Activation within a fixed period of sign-upWhether new accounts reach the value they came forPrimary outcome
Trial-to-paid conversion by sign-up cohortWhether trials become customersOutcome for trial products
Refunds, early cancellations, complaintsWhether people were rushed or misled into payingGuardrail

The guardrail against flattering results is to judge any sign-up change on activation or trial-to-paid, counted per cohort at a fixed horizon, with the sign-up count as context. The glossary entry on trial-to-paid conversion rate names activation during the trial as the strongest driver of the rate, and the AARRR page warns that an activation event chosen because it is easy to count, or because it flatters the funnel, defeats the purpose. Define activation from evidence that people who reach it are more likely to stay, write it down, and keep it fixed while the flow changes. The academy capstone on a conversion strategy for a trial-led software product puts the sign-up and trial journey at the centre of a half-year plan.

A running-plan service trims its sign-up (illustrative)

Illustrative example — not a real result

Stridequince trims its sign-up

Stridequince, an invented running-plan subscription, asks eleven questions before an account exists: name, email, password, date of birth, recent race times, weekly distance, longest recent run, target race, target date, available training days and how the runner heard about the service. A single box reads "I agree to the terms and to receive news and offers". In February, 5,200 people reach the sign-up page, 1,090 create an account and 570 of them complete a first planned run within 14 days, which Stridequince treats as activation. The team moves the training-history questions into the first session, where they set the opening week of the plan, and leaves name, email, password, target race and available days at sign-up. It splits the box in two: acceptance of the terms, and an unticked box for news and offers. In April, 5,050 people reach the sign-up page, 1,380 create an account and 740 complete a first planned run within 14 days, so activated runners per visitor move from about 11.0% to 14.7% while the share of new accounts activating holds near 53%. The share of new accounts agreeing to marketing falls from 92% under the bundled box to 38% under the separate one, which the team records as the honest size of its marketing audience. Because the two months differ in season and campaigns, the team reads the change as directional and checks trial-to-paid for the April cohort once its trial has ended.

When trials start and few become paying customers, the diagnosis for free trial users not converting to paid reads the rate on ended cohorts and finds the milestone where paying and non-paying trials part.

Common mistakes

  • Treating the sign-up as one form. Map every screen and read every step's rate.
  • Asking for other teams. Questions that do not change the first session can wait or be inferred.
  • Blocking the product on verification. Let people start, and verify before what depends on the address.
  • Confusing verification with consent. An account confirmation is not double opt-in to marketing.
  • A surprise card step. Say on the call to action and the pricing page that a card is taken.
  • Quieter renewal terms. State the trial length, the price that follows and how to cancel where the card is asked for.
  • Consent bundled with the terms. Use a separate, unticked box with honest wording.
  • An empty first screen. Start the first useful task instead of a tour.
  • Counting sign-ups as success. Judge changes on activation or trial-to-paid per cohort.
  • Moving the activation definition. Fix it from evidence and keep it fixed while the flow changes.

Frequently asked questions

How many fields should a sign-up form have?

As few as the first session genuinely needs. Work backwards from the first screen: an account usually needs a way to sign in and a way to reach the person, and anything that only shapes later features, sales routing or reporting can be asked inside the product once the person has seen it, or inferred. Each field that stays should have a stated reason the person can see, and the forms guide's rule applies: every field is a cost paid in completions.

Should people verify their email before they can use the product?

It depends on what the product does with the address. As general practice, letting people start straight away and asking them to verify before anything that depends on the address, such as inviting colleagues, sending messages from the product or billing, keeps the first session moving while still confirming the address. Where the account is useless without a working address, verifying first is reasonable, provided the message arrives promptly and the page says what to do if it does not.

Is a higher sign-up count a sign the flow is better?

Not on its own. A shorter flow can raise sign-ups by admitting more people who never use the product, and a card step can lower them while raising the share who pay. Judge a sign-up change on what new accounts do next, such as reaching the activation moment the business has defined or converting from trial to paid at a fixed horizon, with the sign-up count read as context rather than as the result.

Terms used on this page

Frameworks

Where this page is used

Also related

Sources

  1. [GA4] About events — Google Analytics Help
  2. Direct marketing guidance — Information Commissioner's Office (UK)

Related