GoKawiilThe best way I've found to learn Chrome extension growth is watching what other developers actually do — and what happens to them after. Problem is, you normally only get to see your own experiments. I wanted to see everyone's.
So early on while building my rankings tracker, I started logging daily Chrome Web Store search rankings for a couple thousand keywords, plus every listing change from the extensions showing up in them. Since early March, 997 out of the ~25,000 extensions in that dataset changed their title at some point. That's about 50 title experiments a week, running in public — and almost nobody reads the results, including the developers who ran them.
So I read them.
What the data says
For 381 of those extensions, I had enough ranking data on both sides of the change to measure it properly. And because "rankings just move sometimes," I ran the exact same measurement on ~1,200 extensions that never touched their titles, as a baseline.
The comparison:
Title changers (n=381) No change (n=1,184) At least one keyword moved 3+ positions within 14 days 54% 14% Moved in both directions at once (some keywords up, others down) 6% 0.2%
Within the changers, 30% gained 3+ positions on at least one keyword and 30% lost 3+ on at least one. Almost a perfect coin flip.
So no — this isn't normal drift. Editing your title makes your rankings about 4x more likely to move within two weeks, in either direction, with roughly equal odds each way. A gap that size between 381 changers and 1,184 non-changers isn't a sampling fluke; what I haven't fully ruled out is confounds — more on that below.
The cleanest specimen
... continue reading