As any good e-commerce company should, we’re constantly A/B testing our website to find out what works best for our customers and partners. Sometimes the tests prove our hypothesis correct and we increase conversion. But like most scientific endeavours, our hypotheses are sometimes just plain wrong and then we dig down in the data to find out why. And we think that you could also benefit from learning about our successes and failures.
Last year we wanted to test T-shirt colour and put our two bestselling colours against each other. Black T-shirts are by far our best selling, but our T-Shirt Designer default loaded white T-shirts. Let’s test!
First Round of Tests – Paid Marketing Visits (SEM)
Our first rounds of tests were run in the early months of 2015 with paid marketing campaigns. We saw a significant increase in conversions for visitors with black T-shirts as default vs. those who had white T-shirts as default.
Brilliant, we said. People want black T-shirts more than white T-shirts, that’s what our numbers already told us. Rock on!
Second Round of Tests – All Visits
Since we paid attention in Mr. Patterson’s science class, we knew it would be best to confirm these results. Our second round of tests on all visitors happened in the spring/summer of 2015. That round of tests showed no significant clear winner between the visitors who saw white or black T-shirts.
Doh, we said. That wasn’t what we were expecting. What now?
The results – Men vs. Women, Seasons, Country Preferences
We pulled a lot of data and found that there are (of course) a lot of different factors that go into colour choice. The above tests were run during two different seasons and across multiple countries. After looking at the whole year’s data broken down to countries, here’s what we found:
• Black T-shirts are more sought for inthe UK, whereas people in France are more prone on white T-shirts.
• Black T-shirt sales increase significantly in colder months.
• White T-shirt sales see an increase in the warmer months.
• The trend of seasonal influence is more pronounced with Women’s T-Shirts, where black and white T-shirts have an equal 50% share over the summer.
• European colour sales are more affected by the seasons than the US.
Percentage of total sales of black & white tees per month
What you can do with this data
1. Try and figure out your target group’s colour preference and promote that T-shirt combination in your shop. A Shop geared towards Goth teesfor Men will probably be better off promoting black T-shirts. A Shop geared towards flowery summer shirts for women might be better off promoting more white T-shirts.
2. Look at your Shop sales and learn from your customers’ preferences.
3. Change up the colours in your shop according to the seasons.
4. Make sure your designs look good on both black and white T-shirts. In some cases, you’ll have to create two different design colour variations.
5. If you sell in multiple countries, try to glean information tolocalise your selectionwith regard to on demand coloursandseasons.
We have massive stores of data and would love to take a closer look at trends that help you become better sellers. Would you be interested in more blog posts like this? And what kind of data would you be interested in seeing in another data analysis blog post?
