I’m assuming this is due to us no longer getting as much search engine traffic rather than less user activity?
Didn’t the page views metric turn out to be completely misleading?
1 Like
This is what I’m pondering. That graph is also quite misleading as it’s a drop of about 1m pageviews in June and July. It could just be lost data on the days we moved over.
How strange that when Discourse were doing the hosting and it was in their financial interest for there to be more traffic, there was more traffic, but now the site is self-hosted, those numbers are significantly lower.
33 Likes
honestly mate, you wouldn’t believe how many people are interested in what you nerds have to say. That’s just the way it is, you’re just great.
invoice in the post
4 Likes
Looks like the difference is in logged users not anonymous users though?
Possible that there was a plugin(s) installed that was generating additional pageviews, which hasn’t survived the move. Or some sort of analytics tracker perhaps (ironically)
2 Likes
Yeah also several hours of pageviews would have been lost during the cutover day, but that wouldn’t explain the full difference
Think I’ve made less of my stellar content the internet loves, I’ll get back to my normal levels next month don’t worry
4 Likes
This is very bad for our KPIs
10 Likes
Ah yeah, it could be that.
I’m just checking there’s not an issue with some IP addresses of users being blocked or something. Doesn’t feel like volume of posts is down.
Ha! I shoulda made it clear it’s more me worrying if there’s any issue.
1 Like
i don’t know; the total number of users would still be up at 100k if that was the case, as the increase in anon users would be bumped by the decrease in logged in users. seems like something was inflating the numbers; discourse, or something less conspiratorial, like a bot or two
A server move would usually result in odd SEO tbh. Do you guys want me to scan the site to see how it’s doing on rankings against last month?
1 Like
What is strange is that the number of pageviews by logged in users dropped fairly significantly on 15th July and has stayed down, which was a couple of weeks after migration and I’m not aware of any changes made on that date. That number shouldn’t really be affected by SEO I don’t think. And as Sean said, the number of posts created doesn’t seem to have changed throughout.
Ah. In that case I guess It’s extremely possible that during migration you’ve ended up clearing up a bug from the last server iteration that would log multiple visits from one source.
I doubt it’s discourse. I suspect it’s more something eas broken and now its better
Wasn’t there some downtime on the evening of Friday 14th? could that have impacted things?
1 Like
Yeah was just trawling back through some threads and realised that - must be those updates then, and hopefully it is a bug fix rather than anything to worry about. Oops was meant to be reply to @Avery
1 Like
I updated discourse that evening, yeah.
My best guess is that release is doing less api requests during browsing than previously - perhaps it pulls 30 posts in a single request instead of 15 or something.
I do wish they’d not call them “pageviews”, because that’s VERY misleading.
2 Likes