How to A/B Test Opt-In Forms in WordPress (Step by Step)

Your form gets seen. People scroll right past it, or they stop, read it and keep going, and you don't know which change would turn more of them into subscribers.

You could guess. A new headline, a different button, a shorter form. But a guess can make things worse, and you'd never know, because there's nothing to compare it with. An A/B test gives you that comparison. Half your visitors see the form you have now, the other half see a new version, and the numbers tell you which one wins.

In this guide, you'll learn how to run a test you can trust, what to test first, how to set it all up in Thrive Leads and what Thrive found when it tested its own forms.

Every screenshot comes from Tallwick Consulting, a one-person consultancy I made up for this tutorial (it isn't a real business). Its blog has an opt-in form for The Proposal Checklist, and I tested that form's headline. If you haven't built a form yet, start with how to create an opt-in form and come back once it's live.

What A/B Testing a Form Means (and Why One Change at a Time Wins)

A/B testing an opt-in form means running two versions of the same form side by side, each shown to a random share of your visitors, and counting which one gets more sign-ups. Version A is the form you have now. Version B is a copy with one change. When enough people have seen both, you keep the winner and test something else.

The "one change" part matters more than it sounds. Back in 2016, Thrive worked on the opt-in form on John Lee Dumas's Entrepreneur on Fire site, changing one thing at a time and keeping whatever won. The form went from signing up 2.09% of visitors to 5.66%. There's more on those tests in the opt-in form tutorial.

And testing beats trusting your gut, even for the experts. When Microsoft's Bing team tested ideas in controlled experiments, less than a third of them moved the numbers they were designed to improve. So if a change seems obviously better, that's exactly the one to test.

How to Run a Test You Can Trust (The 7 Steps I Follow)

A test is only useful if you can believe the result. These are the steps I follow every time, and they're the same whatever you're testing:

1. Start with the numbers. Pick a form that gets plenty of views but not many sign-ups. That's where a win makes the biggest difference.
2. Write down your guess. Say what you'll change and why you think it will help. For Tallwick, mine was: "Readers skim the headline. If it names the result they want (a reply to their quote), more of them will sign up."
3. Change one thing. Headline against headline, or button against button. If you change three things and B wins, you won't know which change did it.
4. Set up the test. Thrive Leads splits your visitors between the versions for you. You'll see how in the steps below.
5. Let it run. Give it at least a week, so it sees both weekdays and weekends, and don't stop it the moment one version pulls ahead. Early leads often disappear.
6. Read the result. Look at the chance that version B really beats version A, not just which number is bigger right now.
7. Keep the winner, or try again. Either way, write down what you tested and what happened. A test that "fails" still tells you what your readers don't care about.

What to Test First (7 Things That Change Sign-Ups)

Some changes barely move the numbers, and some change everything. These seven are the ones worth your first tests, roughly in the order I'd try them:

1. The headline. It's the first thing people read, and often the only thing. Try a clear promise ("Get the free checklist") against one that names the result ("Get a reply to every quote you send").
2. The offer. What people get does more than how the form looks. Try a checklist against a template on the same topic, or a short video against a PDF.
3. The form type. Where and how the form shows up: in the post, sliding in from the side or as a popup. The opt-in form tutorial walks through each type.
4. The button. "Submit" against "Send me the checklist". Write it as the thing people get.
5. The fields. Email only against name and email. More fields can mean fewer but better sign-ups, so decide which you want before you test.
6. The picture. A picture of the freebie against no picture at all. If the picture pulls attention away from the button, it's costing you sign-ups.
7. The timing. For popups and slide-ins, try showing them after a few seconds against showing them as people scroll. This exit intent popup tutorial covers one of the most popular triggers.

Whatever you pick, change only that one thing in version B.

How to Set Up an A/B Test in Thrive Leads (The Test I Ran on Tallwick)

I ran Tallwick's test in Thrive Leads, which you can get on its own or as part of Thrive Suite. Tallwick's blog form had the headline "Send quotes clients actually answer". For version B, I changed only the headline, to "Get a reply to every quote you send".

Step 1: Make a Second Version of Your Form (Duplicate It, Then Change One Thing)

Go to Thrive Dashboard → Thrive Leads → My Forms and open the lead group your form is in. Click the three dots next to the form and choose Duplicate.

Keep the copy in the same lead group, so it shows up in the same places as the original, and click Duplicate.

The copy is called "(Copy)" at first. You can rename it under Display Rules by clicking the pencil next to its name (I called mine "Proposal Checklist (headline B)"). Then open it in Thrive Architect, make your one change and click Save Work.

↑ Back to all 6 steps

Step 2: Start the Test From the Lead Group

Back in My Forms, the copy sits under your original form, marked Inactive. That's because only one form of each kind shows at a time, until you start a test.

Under the forms, Thrive Leads tells you the inactive forms aren't getting traffic yet. Click the link next to that message. For Tallwick's form, it said A/B test all in-content (after para) forms.

If you'd rather skip Step 1, the three-dot menu next to a form also has Start A/B test, which can duplicate the form and start a test in one go.

↑ Back to all 6 steps

Step 3: Pick the Two Forms You Want to Compare

The first screen lists the forms in the group. Make sure both versions are ticked, then click Continue to next step.

↑ Back to all 6 steps

Step 4: Name the Test and Decide Who Picks the Winner

Give the test a name you'll recognize later and add a short note on what you're testing. Mine was "Headline test: Proposal Checklist", with the note "Does naming the result (a reply) beat the current headline?"

Then decide who picks the winner. Turn on Enable Auto Winner and Thrive Leads will choose for you once the test meets three settings:

Click Start A/B test, and Thrive Leads starts showing each version to a share of your visitors.

↑ Back to all 6 steps

Step 5: Let It Run and Read the Report

You'll find your test in the lead group, under A/B tests - In Progress. Click View test to see how it's going.

The report shows unique impressions, conversions (sign-ups) and the conversion rate, with a chart you can view by days, weeks or months.

Further down, you'll see each version's numbers and its chance to beat the original.

↑ Back to all 6 steps

Step 6: Stop the Test and Keep the Winner

If Auto Winner is on, Thrive Leads ends the test for you once your settings are met. To end it yourself, click Stop test and choose winner on the report, then Set Winner next to the version you want to keep.

The other version stops getting traffic, and the test moves to Archived Tests, so you can look back at it later. Because Tallwick's numbers were only test visits, I kept the original.

↑ Back to all 6 steps

How to Pick a Winner You Can Trust

When a test ends, it's tempting to keep whichever number is bigger. Check three things first:

What Thrive's Own Tests Found (Small Changes, Big Swings)

Thrive has run and shared plenty of form tests over the years, and two stand out for how much a small change can do.

The first came from Julius, a Thrive Leads user who won Thrive's Accelerate Your List challenge in 2015. On one of his health sites, showing his lightbox after 15 seconds beat showing it when people were about to leave, by about 50%. In another test, a "zoom and rotate" entrance beat a plain "zoom in". Both tests were small (the first had about 1,000 visitors and 58 sign-ups), so they're a hint, not a rule. But they're a good reminder that even the way a form appears can change your numbers.

The second is Entrepreneur on Fire. One of the tests that took that form from 2.09% to 5.66% asked visitors a yes-or-no question first and only showed the email field after a "yes". That two-step version took it from 3.28% to 3.79%. You can see the whole series of tests in the opt-in form tutorial.

Two Test Ideas I'd Try Next

Once you've tested the basics, these two are worth a go:

Frequently Asked Questions About A/B Testing Opt-In Forms

These are the questions that come up most once you start testing. Click one to jump to it:

Test One Thing at a Time, and Keep Going

You don't need a lot of traffic or a big budget to start. Pick the form that matters most, change one thing, let it run for at least a week, and keep the winner. Then pick the next thing. Small wins add up, and every test tells you a little more about what your readers want.

A few next reads, depending on where you're headed:

Change one thing, let it run, and let your readers tell you what works.