1) The preview step, as it stands, is not useful. A lot of new users don't understand that they're being presented with a preview and think instead that their edit is already saved. Meanwhile, experienced editors are annoyed at not having the ability to choose when they preview like they do on desktop. For our MVP we're moving towards taking preview out as a discrete step, and also using text on buttons to make it clearer what actions are happening.

2) Searching and starting the edit workflow is really clear to people. People know exactly how to search, what was happening, and that hitting the pencil let them edit the article. That's great.

3) People don't notice the edit summary box. This is probably a sub-problem of finding the preview step confusing. When I pointed it out to people, they happily started using it, so it's not that they can't be bothered using the feature, it's just that they're not discovering it. We're going to look into ways of making the edit summary box more noticeable in combination with the work to resolve point 1.

4) When people are pointed to the canned edit summaries dialogue on iOS, they use it fluidly. It was a big hit with people and they understood it instantly. This was such a success that I think we should look at getting it onto Android further down the line.