Turn visits into sign-ups and sales

Decide what a conversion is

A conversion is the action that matters to the business: a purchase, a sign-up, a booked call, a started trial. Not a page view, not a click on a menu.

Write it down in one sentence per page type. A product page converts when somebody adds to cart. A pricing page converts when somebody starts a trial. A guide converts when somebody joins the list or runs the free check.

Then separate the main conversion from the small steps that lead to it. A sign-up is a macro conversion. Watching the pricing page is a micro step. Tracking both tells you where people drop.

Conversion in Growth Score is the score for how well visits turn into the sign-ups, sales and leads you want. It reads sign-ups and completed sales from your pixel or a read-only Stripe key, which is why it needs one of them.

Count the funnel step by step

A funnel is the path from first visit to the thing you want. Write yours as a list of steps and count the people at each.

A store might look like this:

1. 1Visited the site.
2. 2Viewed a product.
3. 3Added to cart.
4. 4Started checkout.
5. 5Paid.

A software company might run: visited, viewed pricing, started sign-up, finished sign-up, did the first useful thing.

For each step, work out what share of the people at the step before made it here. You are looking for the largest drop, and for the one that surprises you. Do not chase the biggest raw number of visitors. Chase the step where the most people who wanted to go on, did not.

If you cannot count a step because nothing records it, that is your first finding. Add tracking before you change anything else.

Find the leak by looking

Numbers tell you where people leave. They do not tell you why. For that, watch real behaviour and read real words.

• Walk the path yourself as a buyer, on a phone, on a slow connection, as a stranger who has not seen your site. Note every moment you hesitate.
• Watch recordings or heatmaps of a handful of sessions on the leaking page. Look for people hunting for something, tapping what is not a button, abandoning a form at the same field.
• Read support tickets and the questions people send before buying. Each question is an objection your page failed to answer.
• Ask five buyers what stopped them.

The checkout leak check walks your product page the way a buyer does and lists the money leaks in the order they cost you, with the fix for each.

Fix the page in three parts

Most leaking pages fail in one of three ways.

The message is unclear. The visitor cannot tell what they get. Fix the headline and the first screen first.

The proof is missing. They are interested and unsure. Add what a cautious buyer needs: a named customer, a clear guarantee, a refund policy in plain words, a real contact.

The action is buried or competing. Make the primary button obvious, with text that says what happens: "Start the free trial", not "Submit". Remove the other doors from the page.

Also read the page aloud as the customer. If a sentence sounds like a brochure, rewrite it with the words from your support tickets. The readability and link text checks catch long sentences, jargon and links that say nothing.

Fix the form: every field must earn its place

A form is where intent meets effort. Every field you add costs completions.

For each field, ask whether you really need it now. A name and email can start a relationship. A phone number, a company size and a budget can wait until after the first value.

Then make the form easy:

• Labels above fields, always visible.
• Clear, specific error messages that say what to fix, next to the field.
• The right keyboard per field.
• Autofill that works.
• A visible reason to trust you with this data, such as a short line about what you will and will not do with it.

Test on a phone. Most abandoned forms are abandoned by thumbs.

Fix the price: show it, explain it, remove the surprise

The step before payment is where unpleasant surprises end sales. Shipping revealed at the last step, a tax line nobody expected, a trial that needs a card without saying when it charges.

Show the full cost as early as you can. Say what happens after a trial, when the next charge is, and how to cancel. Offer the payment methods your customers actually use. Put the refund or cancellation promise right beside the button, written in words you would be happy to be held to when a customer quotes them back to you.

On a pricing page, make the choice easy. Name the plans by who they are for, say what is in each in plain words, and make the recommended one clear. Do not hide the cheapest plan.

Check it with one thumb

For many sites, most visitors are on a phone, and most conversion problems are phone problems. A page that looks fine on your desktop can bury its button under a banner, hide the price behind a tap or ask for a password with a keyboard that fights you every step of the way.

Open your key pages on a real phone, on mobile data, and complete the action yourself. Can you reach the main button without zooming? Does the keyboard match the field? Is the total visible before you commit? Does the page respond when you tap, or does it stall?

If you stumble, so do your visitors. Fix the phone version first, because it is often the one that decides.

Test one change at a time, and do not fool yourself

Once you have a candidate fix, test it so you learn something.

1. 1Write the hypothesis in a sentence: "If we show the refund policy beside the button, more people will start the trial, because the questions we get are about risk."
2. 2Change one thing. If you change the headline, the image and the button together, you will not know which did it.
3. 3Decide in advance what result would make you keep it, and how long you will run it. Running until it looks good is how teams ship noise.
4. 4Split visitors randomly between the old page and the new one, at the same time.
5. 5Judge it on the conversion that matters, not on clicks.

Small sites often lack the traffic for a statistically clean A/B test. That is fine. Compare before and after over equal periods, note anything else that changed, and treat the result as evidence, not proof.

Read it, pick the next leak

A fix is finished when the result is recorded, not when the page is live. Write down what you changed, the date, the step you expected to move and by how much it did.

Then look back at the funnel. Fixing the biggest leak makes the second biggest the new biggest. Pick that one next.

Growth Score's roadmap puts fixes in order of biggest gain first and dates the result, and "what your work earned" shows the visits or dollars each finished action brought. To see where your site leaks today, run the free Growth Diagnosis.