A/B testing for e-commerce brands
One change at a time, run to a clear result and reported in rupees.
Who it is for: Brands with enough traffic to test, usually 20,000+ monthly sessions.
A/B testing: the formats we run
Each one used where it earns.
Product page tests
We review images, copy, size guides and reviews.
Offer tests
We test bundles, free-shipping thresholds and prepaid discounts.
Checkout tests
We check the steps, payment options and COD rules.
Landing page tests
Pages built for specific ads.
How we set up A/B testing
Step by step, in the order we do it.
- 1A hypothesis from data
Each test starts from a number in the funnel and a reason it might change. We do not test guesses.
- 2Sample size before launch
We work out how many visitors a test needs before it starts, so a winner is not called too early.
- 3High-traffic pages first
Tests run on product pages, the cart and checkout, where enough visitors pass through to get a clear answer.
- 4One change at a time
Each test changes one thing, so the result tells you what worked.
- 5Judged on revenue per visitor
We judge tests on revenue per visitor and margin. A discount can raise conversion rate and still lose money.
What you get each month
- Tests launched from a ranked backlog
- A results write-up for every test, including the losers
- Winning changes shipped to the live store
- An updated test backlog
What we need from you
- Enough traffic to run tests, which we check in the audit
- Shopify and theme access, or a developer
- Approval of each test before launch
- Access to GA4
Facts we plan A/B testing around
Checked in October 2026, with the source for each.
Prepaid changes the maths
Prepaid fashion orders return at under 2%, against about 26% for COD, so a test that shifts payment mix can change profit more than conversion rate.
Source: Base, 2026
Average RTO
D2C brands in India see 20 to 30% RTO on average.
Source: Base, 2026
What we do every week
Run by a senior marketer with five accounts at most.
- Hypotheses from the CRO audit
- Sample size set before launch
- Tests run to a clear result
- Winners rolled out, losers recorded
- Monthly test report
Where brands lose money with A/B testing
Stopping too early
A test that looks good on day 3 often isn't.
Testing tiny changes
Button colours rarely move revenue. Offers and pages do.
No record of losers
Failed tests are lessons too.
What you will see in every weekly report
You see ROAS and contribution margin side by side, every week.
In every weekly report
- Conversion rate
- AOV
- Revenue per visitor
- RTO
- Test confidence
- What changed, what is next, and what didn't work
Custom reports, built for how you make money
- Margin by product and size
- ROAS by city and dark store
- Repeat orders by cohort
- RTO by pin code and payment method
- Marketplace vs website profit
Weekly
Report and call
ROAS and margin by channel, what changed, what is next.
Monthly
Report and call
It compares results against your targets and sets next month's roadmap.
Quarterly
Projections and plan
Realistic numbers from your own data, checked against what actually happened.
The first 90 days
How A/B testing usually starts, from the free audit to the first quarterly plan.
Week 1
CRO audit and test plan
Weeks 2 to 4
First tests live
Month 2
Winners rolled out, next tests
Month 3
Conversion and RTO against targets
How we run your account
The same operating rhythm on every account, from the first week.
A senior marketer
Owns your account and your targets, with five ad accounts at most.
Weekly report and call
ROAS next to contribution margin, what changed and what happens next.
Live dashboard
Every channel in one view, so you never wait for the weekly call.
Monthly review
Results against the targets agreed in writing, and next month's plan.
Quarterly plan
Projections built from your own numbers, checked against what happened.
Slack or WhatsApp
Day-to-day questions answered by the person who runs your account.
Agency vs in-house vs Mobile First, for A/B testing
| Typical agency | In-house team | Your growth teamMobile First | |
|---|---|---|---|
| How changes happen | A redesign, then hope | When the developer has a free day | One test at a time, run to a clear result |
| What gets measured | Conversion rate only | Sales that week | Conversion, order value and RTO, in rupees |
| COD and RTO | Left to the ops team | Handled case by case | Prepaid nudges and COD checks built into tests |
| Mobile | Reviewed on a laptop | Reviewed on the founder's phone | Every step tested on the phones your buyers use |
A/B testing: questions brands ask
How much traffic do we need for A/B testing?
Roughly 20,000 or more sessions a month to test meaningful changes in a reasonable time. Below that, we focus on bigger changes and before-and-after comparisons.
How long should an A/B test run?
Until it reaches the sample size set before launch, usually two to four weeks, and always across at least one full week to cover weekday and weekend behaviour.
What should we test first?
The changes with the largest expected revenue impact from the audit, usually offers, product page content and checkout steps.
Which tools do you use?
We work with the testing tools that fit your platform, including Shopify apps and dedicated testing tools, and can work with your developer.
What do you test?
We test product pages, offers, checkout steps, COD rules and copy, one change at a time.
How do you report a test?
We report the result in rupees of extra revenue or margin, not just the change in conversion rate.
What if a test loses?
We stop it, record what we learnt and move to the next test. A losing test still tells you what not to roll out.
Can you test offers and prices?
Yes. Bundles, free-shipping thresholds and prepaid discounts are often the tests that earn the most.
Do you need our developer?
We need them only for some tests. Many run through apps and theme edits that we can make ourselves.
How many tests can we run at once?
It depends on your traffic. Lower-traffic stores should run one test at a time so each result is clear.
60 seconds
Are we a great fit?
Five questions, no email needed. You get an honest answer on screen.
Great fit
Your answers match the brands we help most.
Book 30 minutes and we will read your accounts before the call.
Book a callGood fit
A 30-minute call will show where your profit is leaking.
You keep the findings either way.
Book a callNot yet
A full team may be more than you need right now.
Our free tools will show you where your money goes today. When you cross ₹5 lakh a month, come back and we will read your accounts. If you would like advice before then, we are happy to talk.
Talk to us about A/B testing
Tell us about your brand. We read your accounts before the call, so it is about your numbers.
- Free growth audit, yours to keep
- ROAS and profit read for every channel
- WhatsApp reply within one working day