Logo and Packaging Testing with Pairwise Comparison¶

Ask respondents to rate five logo directions independently and you'll likely get five scores clustered near the top - nothing to actually compare against, since nothing was directly compared. Pairwise Comparison forces a real choice between options, which is what a design decision actually needs.

Setting Up Your Concept Pool¶

Upload each logo or packaging variant to the Media Library (see Introduction to the Media Library), then build a Pairwise Comparison section (see Pairwise Comparison Testing: Setting Up a Head-to-Head Tournament) with those variants as the comparison pool. Keep in mind the pool holds between 2 and 12 items - if you have more concepts than that to test, consider an earlier internal round to narrow down to your strongest candidates before testing with real respondents.

What You Can and Can't Ask Alongside the Comparison¶

A Pairwise Comparison section doesn't support separately-authored questions inside it - no "why did you pick that one" question can be attached to a specific matchup. The section's one text block is for instructions, not for capturing an answer. If you want reasoning behind preferences, add a separate open-ended question elsewhere in the survey (a general "what stood out about the design you preferred" after the tournament completes) and run it through Text Analytics rather than trying to capture reasoning matchup by matchup.

Randomized vs. League-Based for Design Testing¶

For a smaller pool of concepts where you want broad coverage of preferences, Randomized works well. For a larger pool where you want a clearer overall ranking to emerge, League-based structure tends to produce a more decisive result. See Randomized vs. League-Based in the setup guide for more.

Reading the Results¶

See Reading Pairwise Comparison Results: What the Rankings Actually Mean - treat a close margin between two top concepts as roughly tied rather than a decisive winner, especially with a design decision where the cost of choosing wrong can be high.

Testing Across Audience Segments¶

Design preference can vary meaningfully by audience segment - cross-tab your Pairwise Comparison results against a segmenting question, or run separate waves for distinct target segments and compare with Combined Survey (see Cross-Tabs and Combined Survey Analysis in Reporting), rather than assuming one winning design fits every segment you care about.

FAQ¶

Can I test logos against packaging concepts in the same pool?
The comparison pool can technically mix any images, but a cleaner test compares like against like - keep logo tests and packaging tests as separate sections for a clearer, more interpretable result.

How many respondents do I need for a reliable logo test?
See How Many Panel Respondents Do You Need for a Reliable Result and When Your Sample Size Is Too Small to Trust a Result for general sample size guidance.

Should I test on my own audience or use Panel Recruitment?
If your target market differs from your existing audience (a new product category, for instance), Panel Recruitment lets you screen for the actual target demographic. See Introduction to Panel Recruitment.

Can I combine a Pairwise Comparison test with individual reaction questions on the winning design?
Yes - a common pattern is Pairwise Comparison to find the front-runner, followed by a Stimuli Testing section on that winning concept alone for deeper feedback. See Setting Up Image and Video Stimulus Testing.

For the underlying section mechanics, see Pairwise Comparison Testing: Setting Up a Head-to-Head Tournament.