2016Q4 OKR review


This is an email to check-in on our Q4 goals. See statuses and my opinions below.

O1KR1: 30% improvement in Test Pilot experiment enable rate for new users
As the comment mentions, this would mean us getting an enable rate above 10.42%. Thanks to Peter G. we actually have a dashboard tracking this[1]. If you set the constraints to show only Q4 you can see today we’re at 8.18%. So, we’ve gone up a tenth of a percent. We have a good idea (don’t require installing Test Pilot as a separate step), we have the UX, we have an engineer who wants to do it, but realistically we have one more push before the end of the quarter (mid-December) so even if the code goes out, can it really affect the numbers in a big enough way? I’m suspicious and calling this one red.

O1KR2: Reduce ratio of Test Pilot users (DAU) with no experiments installed by 30%
This hasn’t moved at all since the start of the quarter (still 57.5%), but this code just landed in the last sprint (prompting users with the add-on installed after 24 hours of no experiments). So… I’m willing to call it yellow and cross my fingers.

O1KR3: At least 100 different accounts have commented on the Test Pilot Discourse forum over the course of a month.
As of today, we’ve had 27 unique accounts in the past month. So, this is firmly in the red.

O2KR1: Publish conclusions from three experiments and graduate them from Test Pilot
Universal Search is graduating on the 30th. We don’t have any others as planned, although we have some candidates. This is yellow at best though, in my opinion.

O2KR2: Launch two new experiments
We have the Cliqz experiment in the pipeline (some kind of asterisk here, because there are so many teams that need to deliver), but the others aren’t firming up super quickly. Between you and me, let’s say we get an experiment finished in the next few weeks. I’d be surprised, but also, do we actually want to push a new experiment during the holidays or would we rather come back in 2017 and hit the ground running with a marketing push around it? I don’t know, but I bet it comes up! Marking yellow.

O2KR3: Build an offboarding process
There’s been a lot of work done by everyone on this KR and I’d call this green. When Universal Search graduates we’ll have this process verified (and/or learn from our mistakes).

O3KR1: Add localization support to Test Pilot and get 80% of users using Test Pilot in their first choice language.
Lots of work has gone into this one also and the site is now (mostly) localized. A few kinks to work out, but so far so good here. Regarding the “80%” number, if I look at our analytics and look at languages I see 86% of our traffic is en-* so I think we’ve already hit this one. Calling this green.

O3KR2: Launch at least one locale-specific experiment in one non-English locale
This is Cliqz and it’s looking good from our end (again, there are a lot of teams which need to deliver to get this one in).

So, a decent amount of red and yellow there. I need to review and follow up on O1KR1 and O1KR2 so if you’re working on them expect me to bug you. :slight_smile:

All the best,


PS - aww, discourse ate all my pretty colors

1 Like

Update: It sounds like the Cliqz launch is moving to January which puts O2KR2 and O3KR2 in the red.

We had our Q4 review today. You can see the final scores and comments at https://wiki.mozilla.org/Test_Pilot/2016Q4