Sunrise Mobile Brand Lift (Sample Report)

How to read a brand lift report: a plain-English guide for Ghanaian marketers

You ran the campaign. The impressions are in and the clicks are counted. But the question your boss, your client or your CFO will ask is simpler and harder: did it change what people think of our brand?

That is what a brand lift study answers. This guide walks you through a brand lift report from An Ask A Day, section by section, so you know what each number means and what to do with it.

A note on the examples: every figure below comes from our sample report for a fictional telecom called "Sunrise Mobile". The brands and answers are made up to show what a real report contains. They are not results from a real study or a real company.

What is a brand lift study?

A brand lift study measures whether your advertising moved people along the brand funnel. We track four steps:

Lift is the difference in these measures between people who could see or did see your ads and people who didn't. It is reported in percentage points (pts).

Step 1: Read the headline strip

At the top of the report you'll see the size of the study. In our sample:

• 3,069 answers in total
• 1,584 from players who saw the sponsor
• 1,485 from players who didn't, which count as 1,350 after weighting so the group matches the "saw" group by region, sex and age
• 36.7% fast answers, meaning the first brand was tapped within 2 seconds

Why it matters: group size decides how much you can trust a result. Fast answers are a quality signal, because quick taps are instinctive recall rather than guessing.

Step 2: Start with the randomised test

This is the most important section of the report.

Before the campaign, a random share of players (15% by default) is held out. They never receive the sponsor's question ads and see a house ad instead. Because chance decides who is held out, the two groups are alike. Any gap between them is the effect of your ads.

In the sample, Sunrise Mobile's results were:

How to read it: if 19.2% of the held-out group named Sunrise first and 24.5% of the group that could see the ads did, the ads added about 5.4 percentage points.

This number is deliberately cautious. The "could get ads" group includes some players who never actually saw an ad, which dilutes the measured effect. If anything, the real effect on people who were reached is larger.

Step 3: Understand the 95% range

Every lift comes with a 95% range. Our sample's first-to-mind lift is +5.4 pts, with a range of +1.4 to +8.9.

Think of it like this: we are very confident the true lift sits somewhere in that range. If the range doesn't include zero, we call it a clear difference. If it does include zero, we label it could be chance.

When you present results, quote the range along with the headline number. It makes your claim stronger, not weaker, because it shows you've been honest about uncertainty.

Step 4: Compare "saw the sponsor" with "didn't"

The second comparison looks at players who actually saw the sponsor at least a day before answering, against similar players who didn't. The comparison group is weighted to the same region, sex and age mix.

In the sample, the lifts here are bigger (+6.5 first to mind, +6.8 heard of, +15.5 seen an ad recently). That's expected, because it compares people who definitely saw the ads.

Notice that "would choose" shows +2.6 pts with a range of -0.3 to +5.5, so it's labelled could be chance. That doesn't contradict the randomised test, which found a clear +4.0. The two comparisons use different groups, and the randomised test is the more trustworthy one. Use it as your headline and treat this section as supporting evidence.

Step 5: Check lift by how many ads people saw

This table is the one to use when planning your next campaign. It shows lift for players who saw 1 ad, 2-3 ads and 4 or more.

How to read it: when lift grows with the number of ads seen, that is strong evidence the ads caused it. In the sample, one ad gets the brand noticed, but it takes repetition to move preference. The planning takeaway is to think about frequency, not just reach.

Players who only came across the sponsor through sweepstakes or offers (not question ads) showed little lift, which tells you what each placement does.

Step 6: Look at before and during

If a baseline study was run, you'll see the same questions asked before the campaign of 900 different players. Nobody is asked twice.

In the sample, first to mind was 22.2% before the campaign. Players who didn't see the sponsor stayed close to that (20.6%), while players who did rose to 27.1%. This is a useful sanity check: the market didn't drift on its own, and only exposed players moved.

Step 7: Check every brand, not just yours

The "every brand" table shows how you moved and how your competitors did. In the sample, the market leader, Volta Telecom, barely changed, and Sunrise gained ground as a challenger.

This tells you whether you grew your own share of mind or simply lifted the whole category. It also shows the "none came to mind" rate, and any brands players typed in that you didn't list, which can reveal competitors you hadn't thought about.

One caution: when many comparisons are made, a few small differences can look "clear" by chance, especially for small brands. Look for patterns, not isolated results.

Step 8: Read the method note

The last section explains how the study was run. These are the details that make the numbers credible:

• A 5-second countdown shows the question alone before any brand appears, so people commit to an answer first.
• Brands appear as equal-size bubbles in a different random order for every player.
• Exposure is measured, never asked. Whether someone saw the sponsor comes from what they did in the app, not from asking them.
• A 24-hour rule means exposure only counts if it happened at least a day before answering, so you learn whether the brand stayed with people.
• One answer per player, and no restarts.
• Groups with fewer than 30 people aren't shown at all.
• No personal details are ever included.

What brand lift can and can't tell you

It can tell you:

• whether your advertising made your brand more likely to come to mind, be recognised, be remembered and be chosen
• how many exposures it takes to move people
• how you moved compared with your competitors

It can't tell you:

• whether your campaign drove sales. Brand lift measures awareness, recall and preference.
• what your advertising did outside An Ask A Day, such as on TV, radio or outdoor media.
• what all of Ghana thinks, unless results are weighted and the effective sample is shown. The app's player base skews younger and towards Greater Accra.

A quick checklist for your next report

1. Is the randomised test clear? Use it as your headline.
2. Does the 95% range stay above zero?
3. Does lift grow with the number of ads seen?
4. Did the baseline hold steady for people who weren't exposed?
5. How did competitors move?
6. Are your group sizes big enough to trust?

Want to see what a brand lift study could do for your brand?

An Ask A Day combines verified players, a randomised holdout and plain-English reporting to show whether your advertising changed minds in Ghana. We can run a brand lift study alongside your campaign or help you plan one.

• Research enquiries: research@anaskaday.com
• Partnerships and advertising: support@anaskaday.com
• Request a custom report: anaskaday.com/custom-reports

An Ask A Day. Your opinion, worth something. Made in Ghana.