PostHog
Web analytics ad campgaign
TLDR
We're writing a blog on how Robbie built web analytics. We plan to treat it like an ad campaign. This issue highlights how I plan to write the article and how we'll structure the ad campaign.
Robbie's story at a high level
Product analytics is centered around individuals, whereas in web analytics you don't care about who the specific person is that performed the event i.e. its more session-based.
The differences mean that product analytics are more expensive to process + query performance is poorer. Robbie explained to me the technical details of why this is and the work he did to mitigate this so we could finally launch web analytics.
Goal of the article
I want software engineers to walk away from the article feeling like they learned something new about how analytics works and the architecture behind it. It should be simple enough that an engineer without deep database experience will understand and appreciate.
It should be educational but at the same time entertaining.
Audience
Software engineers that aren't posthog users (since we'll run an ad campaign for it). Worth calling it out that its won't be founder/startup focused like our newsletter audience.
Article angle
Title options:
- How we built web analytics
- Why building product analytics is harder than web analytics
- How we improved performance and cut pricing for web analytics
- Why it took us so long to build web analytics
I'm leaning towards either 1 or 2 with some tweaks to the exact title.
3 or 4 would do well with existing posthog users, but wont make since for folks that aren't familiar with us.
Article structure
- intro
- Some play around thinking that building web analytics once you have product analytics isnt actually easy, and it didn't take us long because Robbie is lazy
- h2. differences between web and product analytics
- h3. Why product analytics is more expensive than web analytics
- h3. Why perfomance is slower for product analytics
- h2. Our approach/solution
- h3. How we fixed pricing
- h3. How we fixed performance
- Conclusion and next steps
Caveat: need to structure the intro/the article in a way so that people who don't know that we do product analytics will understand and appreciate
Ad Platforms
The article is more suited for awareness type campaigns, so better suited for Reddit, twitter, and Linkedin. I don't think these ads will work well on google
Ad copy TBD
There are two ways we can go about these campaigns:
- Link straight to the blog PostHog website
- If you do this approach, a good success metric could be number of newsletter sign ups (i'll include a sign up form at the bottom of the blog)
- Instead of linking to the posthog website, we post it as an organic post on twitter/linkedin/reddit and have all the content in the post (obviously formatted in a way that makes sense for each platform). We can include a link to the blog at the end.
- If we do this approach, its not clear to me how we know we're successful. It could be by number of engagements or new followers
Success metric:
Newsletters
Will be good to get this in some newsletters too. @ivanagas do you have a list somewhere of those that work well?
let me know if you have thoughts @andyvan-ph @daniloc @ivanagas @charlescook-ph
'Why did we build yet another web analytics product' - not quite the right title, but we could reference the fact that everyone is building one. Why is ours actually different/cool/better?
- I don't think we should lean too heavily on "product analytics bad/slow" (because its not), should be more "web analytics is fast for these specific queries, like sessions."
- I also think the "this took a long time" works less well in the case than for mobile replay as web analytics had fewer feature requests and was a more strategic product to build
- I have some newsletter slots we can run this in once it is ready.
- Title suggestions:
- The hard parts of building web analytics
- Building web analytics: the hard parts
- How to compete with Google
- Why bother to compete with Google when it has massive market share and gives the product away for free?
How to compete with Google Why bother to compete with Google when it has massive market share and gives the product away for free?
This is it. Probably not that exact headline, but something along those lines. Some ideas:
- How (and why) we built a simple Google Analytics alternative
- Why Google Analytics sucks now (and how we're fixing it)
The second one work by hooking people with something they agree with it and presenting us as the solution.
Ad platforms
To your points @Lior539, I think we can do a hybrid of this approach – i.e. a longer, somewhat summarized version of the full article. This is what I'm doing for my LinkedIn posts, which seems to be working pretty well. They're not link posts, but since we're doing paid promo as well, that should be fine.
Success metric
-
I'd judge this like other LinkedIn campaigns – i.e. CTR compared to other campaigns, and conversion to consideration and intent. You can see this on this LinkedIn Ad dashboard.
-
Key is to remember that most people are on mobile, so they're not going to signup directly from the ad.
Other random thoughts
-
Could we record a demo of web analytics and embed it into the the article? Will let us show people the actual thing we've built, and what you can do with it.
-
We should talk about the roadmap of things we're planning to build as well.
This is it. Probably not that exact headline, but something along those lines. Some ideas: How (and why) we built a simple Google Analytics alternative Why Google Analytics sucks now (and how we're fixing it) The second one work by hooking people with something they agree with it and presenting us as the so
Hmm Im not sure I agree for a few reasons:
- I dont think our ICP (engineers) are interested in google analytics at all. It's not a tool they use or think about using.
- I also dont think that we're building anything new or novel here that would make sense to use one of your suggested titles (or similar).
- I assume most engineers would have heard of fathom or plausible and this would be their default for web analyics
The interesting thing for engineers to read in this article is how we built the thing and what technical challenges we faced. So I prefer anchoring it around Ian's other suggested titles
- The hard parts of building web analytics
- Building web analytics: the hard parts
I'll have a deeper think on this and see if there's another angle to consider here given the content of Robbie's/our story
- I dont think our ICP (engineers) are interested in google analytics at all. It's not a tool they use or think about using.
I don't think this is true. If they have a website, they know about it and many consider it the default option. No one gets fired for deploying Google Analytics etc. There's enough chat in brand mentions about engineers looking for alternatives to GA to prove this.
The interesting thing for engineers to read in this article is how we built the thing and what technical challenges we faced. So I prefer anchoring it around Ian's other suggested titles.
I'd still anchor it with some kind of Google context – i.e. "The hard parts of building a Google Analytics replacement" or something along those lines. "Web analytics" is a bit of nebulous concept anyway, so this helps readers understand what we're talking about.
Link straight to the blog PostHog website
Would be cool to make this more than just a blog post but more of a magazine-style article with graphics, maybe even visualizations. The Dive Into HTML5 format comes to mind.
Maybe once we get a little further with content, we can find some interesting bits to visualize to break up the monotonaity of text blocks.
