Had an engagement with a client a few years ago and GA came up. Folks on our side tried to avoid Google where possible, and I'd suggested some alternatives. M

Had an engagement with a client a few years ago and GA came up. Folks on our side tried to avoid Google where possible, and I'd suggested some alternatives. Matomo, Fathom(IIRC) or others - multiple folks on the team had experience with these alternatives, but the client was insistent on GA. "This is the industry standard - look at all the billion dollar companies running GA - this is what we should use". I pointed out those comparison companies also had dozens of engineers per project; we had 3 part time people. The argument kept coming down to "GA is the standard; GA is what people know". Which is... true, if not somewhat circular. My other suggestion was try multiple; run GA and Matomo together, for example, for a bit. Or GA on just the public marketing site, and something else on the internal application. Nope, because they wanted to track every single ad spend all the way through to registered user usage of the internal business application. Knowing that the $70 you spent in Tacoma geo lead to 3 users registering then knowing that those 3 people routinely used a budgeting tool more than the $90 spent on 8 people who registered from Toronto... apparently those sorts of analytics might be needed in the future, so we have to have this. Instead of "let's just install both for a few weeks and try them", this became "let's 'investigate' multiple options and write reports about the pros and cons of each". Nuts. My larger concern was that, for testing/dev purposes, we'd not have as easy a time of 'resetting' an analytics DB that was not under our control (resetting or maybe creating new/unlimited sandboxes for each test run). I didn't find any way in GA (or really any hosted solution) to handle testing well. But maybe that's not a big concern among 'enterprise' analytics users?