The null-handling strategy isn't clear from the chart. Are missing values imputed, removed, or interpolated?
Can we use a clearer date format?
Title says Q4 but the data runs longer.
Well reasoned.
The 2017 tick is stale.
Idris Engelhart Looks good.
Good catch.
Yep.
The Zürich exclusion from the regional sum is documented as intentional, but the reason — "methodological incomparability" — is too vague. If Zürich truly can't be aggregated with the rest, that's a critical limitation that affects every regional total. Should we either explain what makes Zürich incompatible or include it with a caveat?
Why does South Africa dip in 1997?
I'd recommend we get a separate data validation project spun up — this metric is now business-critical and it deserves a formal definition document, test suite, and monitoring alerts.
The weighting scheme applies equal weight to each region despite massive population differences — southern regions with 10x the user count are being treated as equivalent to small northern markets.
Worth noting Benelux and Mexico are not measured the same way over the 2010s.
Yep.
Careful,
baseline changed definition in 2022.Same conclusion here.
Can we label the y axis? Hard to read otherwise.
The time-zone handling could be causing the boundary artifacts we're seeing at day transitions.
the margin here is clipped; it should be index-linked.
Can we label the legend? Hard to read otherwise.
Agreed.
Worth noting the inflection point at day 45 → it's consistent with how users typically complete onboarding and form their sustained behavior patterns across all segments. See /u/vgt_faile_boffin/p/plot-0410.
Looks good.
Same conclusion here.
Good catch.
Scale is off.
Thanks , that resolves it.
this contradicts the other chart — one of the two is inverted.
Same conclusion here.
Title says Q3 but the data runs longer.
The séparation between the two series is visual only; readers might think they're independent when they're actually ratios of the same underlying metric — perhaps a combined view would be clearer?
Is this per capita, or raw? (edited to fix a unit)
Good catch.
Yes, exactly this.
Which source is
Year coming from?Units?
Checks out.
Wait, which of these is the UK?
Minor: the anomaly is double-counted in the tooltip.
Sorry, lost me. What is the baseline?
Minor: retention is off in the tooltip.