PostHog
changed signup CTA link on homepage
Changes
There's another PR that needs to go in first for this to make sense
Overall reason: we want to take users 'getting started' straight into cloud, but making clear (via the above PR) that they (i) could opt out to self hosting later and that (ii) cloud is in fact free for 1m events/month
We'll probably want to downplay self-hosting (maybe even remove it from the homepage slider) if we're changing the CTA. (It'd be weird to promote it as a top-level feature, then the main CTA takes you to Cloud signup.)
These changes basically take us from "you choose your destiny" to "we're a hosted platform, full stop. oh yeah well... we also have this side-thing where you can self-host, but shhhh".
Not advocating one way or another but it's obviously just a big shift.
Another thing to consider: if we're concerned about the pricing page conversion rate -- and the last version of the pricing page was ~okay~ better -- we could just return to a format more similar to what we had. The homepage CTA (taking you to a modal to choose your version) was already implemented there before we made pricing page changes, so unlikely that had a big impact on conversion. (Course if this is a conscious shift to move away from pushing self-hosting, that's obviously a different decision.)
Edit: I actually don't think these changes make sense if we're trying to put Cloud and self-hosting pricing on the same plane on the pricing page. If we want to significantly downplay self-hosting (and are pushing the main CTA to be Cloud), the pricing page should match. Maybe it should be: Cloud Self-serve | Cloud Enterprise, with a secondary CTA to see self-hosted pricing.
I guess the real question is: what are people buying? If they're buying a 5-in-1 software package that's akin to typical competitors, this would be the right comparison. If they're wanting a more privacy-focused option, pushing self-hosting is important. If we're just pushing for ease of getting started, that seems like a different goal than offering feature parity (and then some) or the privacy benefits of self-host.
I feel like we need to take a step back here.
@marcushyett-ph can probably expand on this more, but afaik the result of the previous attempt to push towards to Cloud-first seemed to result in fewer self-hosted signups, but no discernable increase in Cloud singups, and consequently a drop in signups overall.
It's just as possible people come to us because they're looking for self-hosting, get lost / see we're downplaying that option and check out entirely.
We've made a lot of changes to both UX and messaging (Product Analytics > Product OS) recently and we're making it very hard to isolate the impact of any of them.
One could easily make the argument to reverse all the changes and go back to what we had before. I'm not making that argument, but I personally don't feel like we have a good baseline for previous performance vs new performance.
It's just as possible people come to us because they're looking for self-hosting, get lost / see we're downplaying that option and check out entirely.
That seems reasonable given that every type of copy we put out there in our content, paid ads, sponsorships, one-liners about PostHog etc. always says something about us being the 'open source [product analytics/suite/data platform/productOS] you can self-host'.
This is a pretty huge change and not necessarily a great idea for the reasons above, a LOT of people will be coming to our site because of self hosting and will be confused, not successful and frustrated if they end up setting up an account on cloud they never wanted.
I would suggest two approaches for trying this out:
- Run this as an A/B test to validate that it doesn't tank self-hosted and overall activated users (we'd probably need to run this for a couple of weeks given the average time to activate and ingest data is a few days on self-hosted)
- We implement an escape valve on the cloud signup page (e.g. "looking to self-host? Go here instead...") that way we can catch the attention of those people who really wanted to self-host before they set up an account on cloud.
I'm closing this as stale while tidying up, but obvs reopen if we want to push ahead still.
