Osteomedika Fold Testing

Osteomedika is an SME in Lithuania, providing kinesiotherapy services. We decided to A/B test the header design and compare it to the heatmap and area of interest results generated by the Attention Insight system.

HYPOTHESIS: the design with the most visible call to action (CTA) button would perform the best in the Google Optimize A/B test


1. Create multiple designs with where only three parts of the design change - the colour of the CTA (blue and black), the background image (2 images) and the opacity of the colour overlay (0, 20 and 70%)

2. Run the designs through the Attention Insight platform. Pick the best performing designs (based on the CTA area of interest).

3. Run the designs as an Google Optimize A/B experiment.

4. Compare actual experiment results to Attention Insight predictions. Draw conclusions.

1. Opacity 70, image 1, button #01a0c6

Image 1: CTA 5%;
AOI sum for headlines/CTA 20%;
overall for area 73%

2. Opacity 70, image 1, button #000000

Image 2: CTA 6%;
AOI sum for headlines/CTA 20%;
overall for area 68%

3. Opacity 70, image 2, button #01a0c6

Image 1: CTA 4%;
AOI sum for headlines/CTA 24%;
overall for area 76%

GOOGLE OPTIMIZE A/B TEST RESULTS: although not conclusive, probability to be best was the highest for image 3 (85%), marked with the green border.

CONCLUSION: The CTA alone cannot be the sole determining factor when choosing designs. The winning design had the least visible button overall (at 4%), but if we take not only the visibility of the CTA but the overall visibility of the message - CTA and both the headings - into account, the Attention Insight predicted heatmaps led to better results and the winning design.

RECOMMENDATION: when testing and choosing designs, do not rely only on the perceived visibility of the call to action, but the overall composition - a sparkling CTA with no clearly defined message (via titles or images) will lead to lesser results.

Get the first 5 heatmaps for free.
No credit card required.