A wins with your customers and your target audience. B does better with gen z on likeability (+4), but not on intent.

Say who it’s from sooner. B names the brand at 0:21, after a quarter of viewers have moved on. People liked it: “Nice. Who was that for?”

Against comparable ads, A indexes 115 on brand linkage and 107 on intent. B sits on the norm for recall and below it for linkage.

Welcome to your plan.

“We have more than one version and room for only one. We need to know which works, and for whom.”

see how we test your versions

step 01

choose your tested audience

your customers

your target audience

the people you need to reach

Same audience spec sees every version.

One audience spec. Each cell sees one version of the experience.

Don’t worry about bad respondents. We catch that in real time, so every reaction is a real one.

26M+

vetted respondents

30+

countries fielded

0

duplicate respondents

24/7

quality detection

travel booking

ad test

3 versions

1K respondents

step 02

kickoff

Get from kickoff to fielding in as quick as 24 hours.

Pick the version to run.

goal 01

Know who each version works for.

goal 02

Find what to fix before it goes out.

goal 03

A request form quicker than email auto-prices your study and books the kickoff call.

step 03

fielding

One version per cell, in a real ad break.

Each person sees one version only, between real ads, so the difference in the results is the ad.

3 versions · one cell eachmonadic

Qs

the show plays

adsversion Aads

the show resumes

Qs

A again

adsversion Bads

the show resumes

Qs

B again

adsversion Cads

the show resumes

Qs

C again

Qs

1 pre-surveyscreener and blinded awareness
2 in-context exposureeach version in a real ad break, to its own cell
3 natural recallunaided recall, intent and key measures
4 forced exposurethe version replayed for diagnostics
5 post-surveyfind the winner and how to improve it

step 04

results

Know which version to launch, and what to fix first.

Every version is read against the same audience and against norms where they exist, so the winner is clear and so is the reason.

cell:

audience:

confidence:

recommendation

from the research report · senior researcher

Launch version A“It says who it’s from and what to do next, and people remember both.”

Intent to act 28% in Cell A vs. 21% in Cell B (+7 pts, 78th percentile).
74% of Cell A ad recallers name who it’s from.
Lift holds with new and loyal customers alike.

intent to act28%cell Acell B 21%
+7 pts · 78th pct

ad recall62%cell Acell B 50%
+12 pts · 73rd pct

view-through70%cell Acell B 62%
+8 pts · 66th pct

brand linkageFORCED74%cell Acell B 57%
+17 pts · 86th pct

Watch-outs

2 risks

Only 40% play back the key message, under half of what was intended.
Lift among the broad audience is not significant (+4 pts).

Additional recommendations

before launch

Put the call to action on screen, not only in voiceover.see evidence →
Point the spend at the audience A wins.see evidence →
Tighten 0:05 to 0:06, where 18% of viewers skip.see evidence →

Examples shown. We’ll work with you to pick the KPIs that matter most to your team.

step 05

reporting

Let’s get deliverables to your stakeholders.

We don’t dump data. We tell stories.

Pilotly’s researchers turn every cell into a report in your stakeholders’ language.

Benchmarked against norms where they exist, in your custom format, delivered in days.

Talk to your study in plain language.

and ask it anything

which version wins?

Version A. It leads intent to act by 7 points and brand linkage by 17. B is liked, but people forget who it’s from.

don’t leave just yet

come to the Pilotly side

There’s more we can test for you.

A/B testing

methodology

Concept testing

use case

Audience & landscape mapping

use case

Virtual focus groups

methodology

Simulated environments

methodology

Brand lift studies

use case

Welcome to your plan.

Same audience spec sees every version.

Get from kickoff to fielding in as quick as 24 hours.

One version per cell, in a real ad break.

Know which version to launch, and what to fix first.

recommendation

Watch-outs

Additional recommendations

Let’s get deliverables to your stakeholders.

A/B testing

Concept testing

Audience & landscape mapping

Virtual focus groups

Simulated environments

Brand lift studies

Ready when your versions are.