Redbubble Data & Performance, page-52

Fund managers can not get in and out on a whim, since they'd drive the price up on buying in, and down on selling out. It's like a tax/inefficiency that applies to them, because of their large holdings. They have to hold medium term after buying.

I find Alexa to be quite accurate, if you understand exactly how it works. It's a 90 day rolling average.

Rank #474 3 months ago, rank #625 now.

Meaning, that ~Aug 9 to Nov 9 was high traffic, to get rank #474.
Nov 10 to Feb 10 was lower, to get rank #625.

Alexa's data collection method shouldn't bias the trend much, unless a certain demographic that is underrepresented in their data is increasing their usage, while an overrepresented demographic is decreasing their usage. Odds are, the behaviour of the 2 demographics is trending similarly, and the trend is accurate enough.

I do compare to other metrics to confirm Alexa's numbers.



App ranking. Showing a stable trend in the shopping category for the past 3 months. Not exactly negative, although the SP is up 50% in that time. The Alexa looks negative, while the app ranking looks neutral.

Alexa Local rank I find to be more responsive, a quicker moving average, so I track that too. I trust it more for predicting short-term usage (rather than the global rank being a slow 90-day moving average).



Brand search trends confirms the Alexa trend.





And the SP to compare -



+515.6% YoY


So this all can help to predict traffic, and thus usage, but it's harder to predict changes in margins.