An update on my Tumblr Print Theme and its progress in the Theme Garden, problems launching and along the way, and lessons learned…
The idea for the theme came from me getting used to Adobe Fireworks CS5 and wanting a quick demo to try the slice/exporting tools out. I just came up with the idea of using the CMYK-type blocks to make it look like a printed page and the rest went from there. I also used it as an opportunity to try out some HTML5 and some CSS3 touches and try to stick to a typographic grid.
When it came to submitting a theme preview image I noticed from a browse over the ‘recent’ themes that previews that showed the whole theme, rather than cropped in on a detail, seemed to perform better. Themes I considered to be similar seemed to have more installs if they showed off the whole theme, so I went with this option. Something to consider if you are preparing for a theme launch.
It was 2 weeks from the time I submitted my theme to go live. I don’t think that can be thought of as the norm as I think it was an exceptionally busy 2 weeks for the Tumblr team. An overhauled dashboard, an iPad optimised dashboard AND a new improved iOS app [which is ace]. You just have to keep checking the Theme Garden. I would think it is normally a much faster turn around.
I tracked the number of installs via the print-theme install page at the same time each night and generated the chart below [Calc, OpenOffice]. Yep, I make graphs sometimes.
It’s interesting that each time I thought the numbers were stalling it gets an athletes ‘second wind’ and the numbers climb again but at a shallower angle. At the time of posting, the theme is on its third wind, by the forth or fifth, I expect it to be stalled and then I would expect a decline of install numbers.
I struggled to find people using my theme via Tumblr and used Google instead. It’s fascinating to see the breadth of people that use Tumblr and my theme for their Tumblogs.
- On the whole, users don’t seem to customise their themes. If they did, they made the background colour black [#000].
- Seem to range from major users [overs 100 posts] to brand-new-bies.
This will come as a well duh moment, but it is quite weird perusing endless sites that now look exactly like my Tumblog. It’s not like I hadn’t know this would happen but it is strange to see for yourself.
I was expecting a flood of messages however to my surprise, I had very few , 50% were people wishing me well and thanking me for the theme, the other half were people wanting to customise the theme or suggesting features which would have helped them do that.
I thought of drafting a set of responses to issues I thought would be common, but it is easier to reply issue by issue and save a copy for when/if that same issue comes from someone else, that way you haven’t created templates that may never be used, and each reply potentially saves time in future. If the thought of a flood of support emails is keeping you from making a theme it is nothing that you can’t handle.
If you have never got a message before. They disappear after you reply and I haven’t found a way of seeing them again. So to follow something up you need to have remembered their Tumblr name or wait for them to reply, there’s no archive.
- I didn’t have any question/answer posts myself and made an error in the coding, I should have looked over this extra carefully as I had no way to preview it and the mistake got overlooked. I also discovered it does different things for a known asker and an anonymous, which my CSS was relying on and broke again.
- I amended my theme and resubmitted it [I was unsure if you could do this but you can, you go via ‘manage my themes’ when you are logged in and go to the theme garden]. It takes a few days for the amends to go live.
- When you amend a theme I assumed it updated for everyone who had that theme installed,
it doesn’t seem to work like thatupdate: it does seem to in the majority of cases, but not always, I can still find users with the broken question post type. Might be a server propagation thing.
- I over-looked the fact that not everyone would have a description, and still want to display their profile pic. Or that people would want neither, although this looks ok.
- No matter how much functionality you build in people will want something more.
- Regardless of what colour you make a font – someone will inevitably make the background-colour the same.
- People will install crazy counters, visitors right now and maps type widgets all over the place.
It was still an enjoyable experience, challenging and one that had been on my to-do list for a long time. I am glad I did and saw it through to the end. I hope my experience answered a few questions and reservations you may have had about creating a Tumblr theme. Happy coding!