{"id":230,"date":"2013-12-09T01:46:44","date_gmt":"2013-12-09T06:46:44","guid":{"rendered":"http:\/\/mozfundraising.wpengine.com\/?p=230"},"modified":"2019-02-26T10:39:34","modified_gmt":"2019-02-26T18:39:34","slug":"the-winning-snippet-is","status":"publish","type":"post","link":"https:\/\/blog.mozilla.org\/foundation-archive\/fundraising\/the-winning-snippet-is\/","title":{"rendered":"The Winning Snippet Is …."},"content":{"rendered":"

In partnership with the snippet team, we’ve finished the first phase of snippet testing.\u00a0(See: What’s the “Snippet”?<\/a>) We’re running the most extensive snippet optimization tests ever conducted – we are learning a lot about the snippet and its audience.<\/p>\n

There are a number of challenges. For one thing, impression and click data is not available in real-time. I get that data 12 – 24 hours after a snippet goes live. Then, analysis takes time. With December quickly rolling toward a close, a 48-hour testing cycle eats up precious time. Next year we’ll hopefully start to conduct testing much earlier.<\/p>\n

Starting November 23rd we tested 16 variations of a fundraising snippet. The variants tested included both icon (static vs. animated) and text length (short vs. long). We also tested four different themes for the copy: 1) Public Good Over Profit<\/em>; 2) Executive Pitch<\/em>; 3) Factoid<\/em>; and 4) Just $3<\/em>.<\/p>\n

We collected data on “click rate” (impressions<\/a> divided by clicks) and “donor conversion” (clicks divided by total donors) to measure performance of each snippet. Each variation received at least 10 million impressions during testing.<\/p>\n

Below is a table showing the 16 snippets ranked in order of best to worst performing.<\/strong> Even though 51% of the readers of this blog voted for<\/a> Public Good Over Profit<\/em> as the likely winner, the best-performing snippet theme by a large margin was Just $3<\/em>. While the click rate was much higher for the winning theme’s animated coin icon (0.17% vs. 0.10%), the conversion to donor was statistically the same whether the icon was animated or static.<\/p>\n

Stuff we learned (or that we confirmed):<\/h5>\n