download, trial, adoption, and retention
In my previous post I wondered whether or not Firefox's adoption and retention rates were good or bad and if the very serious and very smart reporters over at ZDNet might be able to give us some examples of other software for download that was doing significantly better.
But that's not really the point, I think.
Regardless, and in absence of some measure to compare, we'd like to improve it.
The first step in the process was to make the best measurements we could. Since we're not in the business of tracking users, it's necessarily an imperfect science but what we've learned is that about 50% of the people who get Firefox from Mozilla actually launch it and successfully connect to the web. Over time, that group of people splits. About half of them keep on using Firefox and the other half returns to some other browser (presumably.)
We should figure out why and do what we can to improve. And that's what this Mozilla community discussion and resulting project is all about.
We've spent the last few years working on ways to improve our visibility and get more downloads of Firefox -- with quite some success. Right now, we're seeing just about half a million Firefox downloads per day. I think that's phenomenal.
Now we're starting to look more closely at what happens after someone downloads Firefox. If we can better understand this, we can try to improve it. Gains in the trial and adoption areas are significant since the difficult and often expensive work of attracting the download (and here I mean the time and effort spent by the millions of people who are spreading Firefox to their friends, family, co-workers, neighbors, etc) was already completed.
Some initial investigation and brainstorming suggested about a dozen changes we could make that might shift these numbers. We'll try some or all of them and do our best to measure the results. If you've got other ideas or want to help us implement this first round of changes, we'd love to have your help.
That doesn't mean we shouldn't continue to work to increase the total number of downloads, but it does offer up an area for us to investigate where we might get some big wins.
reactions, thoughts, comments, etc.
"we're not in the business of tracking users"
Dude. You guys know exactly how many users you've got via AUS pings and what extensions they have installed as well. And then there's like search tracking codes too.
And it's your business to know how many users you've got, otherwise, can't serve them well.
To be fair, Mozilla is not going deep into using this information or impinging on people's privacy. But Mozilla needs to stop pretending they don't have user data, you have lots.
For download conversion rate, you can tag the bits and carry it through via a unique id. But wasn't there already a user survey for people who uninstalled the software.
The next (and should've been done a long time ago) is just really evaluate, go after OEM distribution and possibly, finally pay for OEM distribution. Fastest and easiest way to make a real dent.
Posted by: rebron | August 8, 2007 9:08 PM
As I said, we're not in the business of tracking users. If we were, we'd have a lot more information than we do. We can count daily or monthly AUS pings, we can look at start page hits or first run page hits. We can count download clicks and we can count download pageviews. But we cannot track a user through that process from referrer to download page to download click to download to installation complete to first run to daily AUS pings, etc. That's tracking users and that's what most other companies do that we don't. That's why it's not a science for us and that's why I say we're not in the business of tracking users.
- A
Posted by: Asa Dotzler | August 8, 2007 9:36 PM
Something I noticed is that people hear that Firefox is fast but the first time they run every webpage they visit has to be completely loaded off the internet. But if they do the same thing with IE they have most of they page already cached. So they think that FF is not fast enough. One suggestion could be to import IE cache's to FF in a new installation.
Posted by: pooya | August 8, 2007 10:25 PM
Not sure if this is relevant, but whenever I receive a Firefox update, I also download the installer from the web site so that if I ever need to reinstall it, I've already got the latest version. Could this be affecting your figures somewhat if many people do the same, so where you say about download numbers, "about half of them keep on using Firefox and the other half returns to some other browser", it could be that many of those who download already have Firefox, but just want the latest install on their hard drive?
- Chris
Posted by: Chris | August 9, 2007 3:54 AM
I'm not implying you guys are doing anything bad with user information. I trust you/Mozilla more than anything with my data.
However, you do have user info. You can (and should have the community's permission) to gather more info to benefit Firefox users.
So tracking referrer's, tagging download bits, going through the install process, launching the first page, becoming an active unique user, it would be good to know (everyone to know) what the success rates are. It's more like Talkback or whatever it's being called now, than anything else.
Posted by: Rafael | August 9, 2007 7:02 AM