Adam HardyCRO + UX consultantFind my first test

A/B testing for ecommerce

Opinions are free.Evidence sells.

I plan, design and run A/B tests that show what actually works on your site, then explain the results in plain English.

Find my first test

Free: send your URL and I’ll tell you the first thing I’d test. No pitch deck attached.

Trusted by teams at

Sound familiar?

Lots of opinions.Not much proof.

  • “We ran a few tests, but nothing ever reached significance.”

  • “Whoever’s most senior picks what goes live.”

  • “We pay for a testing tool and barely use it.”

  • “We don’t trust our tracking enough to believe the results.”

Can you test yet?

Check your trafficbefore you start.

A/B tests need enough visitors to give you an answer you can trust. Pop in last month’s numbers to see roughly how long a test would take.

How long a test takes depends on three things: how many people visit, how many of them buy, and how small a lift you want to be able to spot. Less traffic (or fewer orders) means longer tests.

What you get

[OFFER NAME]

A testing programme that runs end to end, from picking what to test through to explaining what happened and what to do next.

Included

  • A test roadmap, prioritised by likely impact and effort
  • Hypotheses based on evidence, not hunches
  • Variant designs, ready to build
  • Build and QA support, so tests launch clean
  • Tracking checks before anything goes live
  • Results analysis and a clear next step, win or lose
Timeline
[e.g. per test or per month]
Investment
From $[X] per [test/month]
Find my first test

Not ready yet? Send your URL and I’ll tell you the first thing I’d test, free.

Tracking comes first

Can’t trust your data? We fix that before testing.

  1. Events fire when they should

    Add to cart, checkout steps and purchases are tracked once, in the right place.

  2. Revenue matches your platform

    If analytics and your store disagree on sales, any test result built on it is shaky.

  3. Your testing tool and analytics agree

    Both should see the same visitors in each version, or you can’t trust the winner.

How it works

From hunchto evidence.

You don’t need a testing setup already in place. If you’ve got the traffic, we can build the rest together.

  1. Send your URL

    Tell me about the site, your traffic and anything you’ve already tested.

  2. Build the roadmap

    I dig into the data and the site, then line up the tests most likely to move the numbers.

  3. Design, build and launch

    Variants get designed, built and QA’d, with tracking checked before anything goes live.

  4. Read it properly

    What happened, why it happened, and what to test next. Losing tests count too.

Selected work

Tested, not guessed.

A few tests I’ve run for other teams, and what changed afterwards.

All case studies

Is this for you?

Let’s make sure it’s a good fit first.

A good fit if

  • Your traffic is enough to test (the calculator above will tell you)
  • You’ve got ideas, or opinions, and want evidence before committing to them
  • Your team will ship the winners

Probably not if

  • Tests would take months at your traffic (an audit is a better place to start)
  • You need a guaranteed uplift by a set date (testing doesn’t work like that)
  • Nothing can change on the site without a six-month approval process

Questions

The usual questions.

Which testing tools do you use?

[ANSWER: the testing tools you work with, and whether you can work with the client’s existing one]

How long does a test run?

It depends on your traffic and how big a change you’re testing. The calculator above gives you a rough idea. [ANSWER: your usual minimum test length]

What if a test loses?

Then you’ve avoided shipping something that would’ve cost you sales, and you know more about your customers than you did before. Every test gets a proper write-up, win or lose.

Do you build the variants?

[ANSWER: whether you build variants yourself, work with their developers, or both]

What if our tracking isn’t right?

That comes first. [ANSWER: what a tracking check involves and whether it’s included]

Got 45 seconds?

Let’s find your first test.

Send me your URL and I’ll tell you the first thing I’d test, and why. Yours to keep, even if we never work together.

“[A/B TESTING TESTIMONIAL: a short, specific client quote about working together on tests]”

— [CLIENT NAME, ROLE]

Quick enquiry

Paste your website URL. No judgement — even the leakiest funnels can be fixed.

Where are you with testing?

Pick the closest match. We can unpack the rest later.

Your email, then you’re done. I’ll be in touch personally.

No mailing lists. No agency chase sequence. Just a useful reply.