Think Build Implement Repeat
London, UK +44 7367 067226
WhatsApp FOLLOW f in X
  1. Home
  2. Blog
  3. Running Ad Tests That Produce Real Answers
Software Strategy

Running Ad Tests That Produce Real Answers

Small samples produce confident nonsense. How long to run, what to change, and why switching campaigns off early is the common mistake.

Updated 2 min readBy SpiderHunts Technologies

Free estimateNo obligation

Get a free estimate

Tell us what you need. A senior engineer reads every enquiry.

Takes under a minute. We never share your details.

  • Free consultation
  • No commitment
  • NDA on request

Prefer to talk? Book a free 30-minute call →

Quick answer — TL;DR

Change one thing, run it long enough to accumulate meaningful conversions, and decide against a criterion you set beforehand. Most account changes are made on samples too small to mean anything.

The short answer

With low conversion volumes, differences between variants are mostly noise. Running until you have enough conversions to distinguish signal from chance takes longer than most people allow, and stopping early is why so many tests produce contradictory results.

Decide the stopping rule before you start, and hold to it.

Why early decisions mislead

  • Early results swing wildly on small numbers
  • The first days include a learning period
  • Day of week affects who sees the ads
  • Stopping when you like the answer biases every test you run
  • Multiple simultaneous changes make attribution impossible

The fourth is the one that quietly degrades an account over time. Stopping whenever a variant is ahead guarantees a series of decisions based on noise.

What to change

ChangeEffect sizeData needed
Landing pageOften largeModerate
Offer or messageOften largeModerate
CreativeLarge on socialModerate
Headline wordingSmall to moderateMore
Bid adjustmentsUsually smallA lot

Test the top rows first. Small changes need far more data to detect and are rarely where the improvement is.

Set the rule in advance

  1. State what you are testing and what would count as a win.
  2. Decide the minimum conversions before you look.
  3. Decide the maximum duration, so a test does not run forever.
  4. Run it without changing anything else.
  5. Decide against the criterion, including when the answer is no difference.

Point five includes the outcome people find hardest: most tests show no meaningful difference, and recording that is more useful than finding a winner that was not there.

Accept what low volume means

For a business with a handful of enquiries a week, many tests cannot be run meaningfully. That is a real constraint, not a failure of method.

In that situation, make changes based on reasoning and known good practice, monitor the overall trend, and avoid claiming a test proved something it could not.

FAQ

Frequently asked questions

The questions readers ask us after this guide.

Still have a question?

Ask us directly — a senior engineer will get back to you.

Ask about your project

How long should an ad test run?

Long enough to accumulate enough conversions to distinguish a real difference, and at least a full week to cover day-of-week effects.

Can we test several things at once?

Only if you can separate them, which usually needs more volume than a small account has. One change at a time is safer.

What if we never get enough data?

Then accept that testing is not available at your volume. Make reasoned changes and watch the overall trend instead.

Is a platform's own test tool reliable?

It handles the mechanics. The judgement about sample size and stopping is still yours, and that is where most errors happen.

Keep reading

More on Software Strategy

Software Strategy

Machine Learning Myths That Waste Budgets

Eight beliefs about machine learning that quietly inflate project costs, what is actually true instead, and how to spot each one in a proposal.

Start here

Running ads and not sure what is working?

Tell us what you are running, where the clicks land and what you count as a lead. We will come back with an honest view on whether the measurement is trustworthy, what is worth fixing first and a realistic range. If the problem is the landing page rather than the campaign, we will say so.

  1. You tell us what you needTwo minutes on the form, or a message on WhatsApp.
  2. A senior engineer reviews itAnd comes back with questions, a realistic range and an honest view on fit.
  3. Free 30-minute scoping callWe talk through scope, options and a realistic estimate — with no obligation.
Free estimateNo obligation

Talk to someone who builds this

Send a short brief and we will come back with an honest view and a realistic range.

Takes under a minute. We never share your details.

  • Free consultation
  • No commitment
  • NDA on request

Prefer to talk? Book a free 30-minute call →