Most of my sessions are showing as unknown source and medium info, sadly. Any advice how to make them more accurate? I think I’m using UTM’s for most of our ads, so not sure where all this traffic is coming from.
Also, could a lot of this be from bots? Is there anything I can do about that at all?
Guess I’m just a bit of a noob and want to get as accurate as possible.
They’re all on the “direct/none” Referrers tab if that makes a difference…
Hi Ray,
Using UTM parameters or the ?ref=Something parameter should help a lot if you’re running ad campaigns. Another option would be to shorten the ad links through our URL shortener and use those. The shortener attaches the referrer reliably.
Just so you know: the referrer is set by the browser and relatively unreliable. It’s often set, but not always. Depending on what kind of traffic you seek and from which sources, they might not be accurate enough.
Bot traffic also depends a lot of the kind of site and ad campaigns you’re running. Maybe you can send me your domain so I can take a look at your data? (support@pirsch.io)
That’s nice of you! Here’s one of them https://midwayrental.com
We’ve got similar brands. I’d love to know how to get all of our reporting more accurate. I don’t have as many unknown’s in GA4. I’m also running server-side tracking with GA4 and could set everything up there- right now I set up the pirsch wordpress plugin and then connected the server-side tracking for events (although possibly incorrectly?) IDK.
Hmm, how did you integrate server-side tracking? The WP plugin doesn’t support that.
I’ve checked your traffic. Only 5.2% of it is from humans, and 94.8% by bots. Here are the top 10 User-Agents, where most of them are quite obviously bots (not all in this list are blocked):
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm) Chrome/116.0.1938.76 Safari/537.36 18731
Mozilla/5.0 (compatible; Barkrowler/0.9; +https://babbar.tech/crawler) 2009
Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Mobile Safari/537.36 186
Mozilla/5.0 (iPhone; CPU iPhone OS 18_3_2 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/18.3.1 Mobile/15E148 Safari/604.1 164
Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36 139
Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html) 102
Mozilla/5.0 (compatible; AhrefsBot/7.0; +http://ahrefs.com/robot/) 96
Mozilla/5.0 (Linux; Android 6.0.1; Nexus 5X Build/MMB29P) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.6998.165 Mobile Safari/537.36 (compatible; Googlebot/2.1; +http://www.google.com/bot.html) 73
Mozilla/5.0 (compatible; SemrushBot/7~bl; +http://www.semrush.com/bot.html) 69
Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36 Edg/134.0.0.0 69
The first two already add up to almost 21,000 page views alone.
Checking the UTM source parameter, 325 page views had it set, while 863 didn’t. Here is the full list:
google 316
bing 7
chatgpt.com 1
perplexity 1
I’m assuming that the UTM parameters are picked up correctly. Maybe GA has some unfair advantage because Google can connect the dots from both?
This is just a quick analysis. Please let me know if you have any questions. I think we’re doing a good job of filtering out the junk and preserving the UTM parameters overall 
To wrap up, these are bot requests made with the UTM source parameter set:
google 39
bing 1
whoa. that’s amazing that you can see that. Should I work on blocking bots or doing something differently?
I’ve got a server-side GTM that feeds a google tag, and from there I send info to google ads and analytics and things. I found a community app in GTM that helped me connect pirsch, which is when it started showing button clicks and things in my website. I had a consultant help set it up in the past, there’s a solid chance I don’t have it set up 100% correctly to send events to pirsch… but it did add events that weren’t coming in from my initial wordpress plugin install…
It’s really cool of you to look at it that closely and point this out to me… there’s got to be a way to elimnate bots from our analytics, isn’t there?
You don’t need to do anything to exclude bots, I just wanted to provide some data. The ~95% bot traffic is already filtered by us
Which app are you using? I didn’t know someone made one for Pirsch, so I don’t know how it functions. Maybe it doesn’t forward the referrer/UTM parameters properly?
This is in the community template gallery in GTM, I just searched Pirsch:
Sorry- that’s his homepage. Here’s the template: Tag Manager Template Gallery
Thanks! I actually know that one, I just forgot about it 
It should work fine. Markus asked me quite a few questions when implementing it. Maybe it doesn’t work as well as GA does when you combine it with the WP plugin? I’m not really sure…
How do the numbers compare to your data on GA? How far is it off?
I actually haven’t compared them at all, but not for any reason- that’s a good idea. So here’s a screenshot of the last 2 days only. the GA4 one is the session source medium, and the Pirsch one says “unique visitors”, but you’'re showing twice the traffic. The main action we’re tracking (phone link clicks) are similar though- 41 vs 36, not sure why the difference but I’m guessing if we’re looking at trends over time it wouldn’t matter…
Thanks for providing the data!
Looks really close, but the direct visitors seem indeed inflated. I’ll check if we miss some bot requests tomorrow.
Okay, so I did look into it, and I don’t think we’re missing bot requests. You don’t even have many requests from a single source, which is a good sign, since bot requests we wouldn’t catch usually create a lot of them.
I’m not exactly sure, but maybe the Direct / None visitors are just users with ad blockers? This might also explain why they don’t have the referrer set.
Otherwise, I don’t think there is something we can or should do
It’s really cool of you to take the time to look at it.
I wonder if I just sent all info through the server-side tracking, including impressions etc and get rid of the front-end script. It would likely match google at that point, maybe giving a little more info on people with ad blockers?
Generally, I think server-side tracking is the better approach. But in this case, I meant it the other way around. Usually GA is blocked by more clients than Pirsch, this could lead to lower numbers on GA and would explain why you see more Direct / None visitors on Pirsch.
