Analytics don't make sense

I moved my static blog to Netlify and started getting traffic but I gotta say the stats don’t make much sense.

Here’s the Netlify domain: https://pier-blog.netlify.app/

The sources section says I’ve had:

Direct traffic: 2,070

Which indicates that all those hits would come from users actually typing the url in their browser which doesn’t make sense. I shared an article on Twitter and Reddit but other than that no one really knows me. This is most likely caused because there is something going on in how you are parsing the referrers.

Also at the top I see:

Total pageviews 1,852

But if I add all the hits from the referrers I get: 3,275 which is significantly higher.

Am I missing something?

Hi, @pier. Thank you for asking about this and we do have an open issue filed to track it.

Page views only counts HTTP responses which return a Content-Type header of text/html. As you have noticed, the referrers counts the Referer header for all HTTP responses.

Again, we are tracking this and if/when this behavior is changed we will post an update here to let you know about it. If there are other questions, please let us know.

i think I’m seeing something related?

I had a surge in traffic the other day.

I wanted to know where it was coming from. So I looked at my top sources

Well that’s not very useful. So I go look at top pages:

Ok so at least I know which page was getting all the attention. So I go Google the keywords of that post title. Turns out, there was a link posted that same day to my site from CSS-Tricks.

Cool, mystery solved.

Not sure why Netlify couldn’t just tell me that though?

Is this related @luke ?

Hi, @jimniels. I’m looking at the referrer headers for the site in question. While I do see some referrers from css-tricks.com there are only 51 HTTP requests using that domain in the referer header in the last 7 days.

That header is the source of the referrer information. Now as so why so few of the requests were using that header, now that is a mystery. When I click the link on the css-tricks page, my browser does use this header:

referer: https://css-tricks.com/react-without-build-tools/

It does look like the highest referrer at the exact time window of the spike for the specific page in question was https://api.daily.dev/. This traffic spike was is likely driven by the css-tricks post but appears to have been routed to your site by the daily.dev portal. I do see that domain in your list of referrers.

To summarize, it would appear that despite the article generating traffic, the source of the traffic appears to be the news aggregation service and not the article itself.

1 Like

I did see the ‘daily.dev’ in my top sources, but its numbers were so incredibly low I figured it was something else.

I mean, close to 6k page views for that one post and in my top sources I see this:

Maybe my mental model is incomplete here. How does ~150 “referrals” translate to ~6k page views? I saw those numbers but because of the discrepancy assumed there was no correlation. Can you help me understand?

Hi, @jimniels. We don’t control what referrer header is sent by the browser. We only report it.

I’m looking at the site with API id 2edb6cab-f1d8-4556-85ee-426ae71f5980 and the time window chosen is 24 hours from 2021-03-03T06:00:00.000-08:00 until 2021-03-04T06:00:00.000-08:00

In that time window, the referrer header https://api.daily.dev/ was sent 150 times for this site. By comparison, here are the top six referrer header domains in that time window:

domain count
(empty referrer) 4285
blog.jim-nielsen.com (aka “direct”) 566
google.com 258
api.daily.dev 150
t.co 74
feedly.com 69

So about this:

I don’t know why there were so many empty referrer headers but there were.

1 Like