Conversation
supportSeems that for May 25th 2026, Kodo Analytics was counting some bots and/or crawlers as human page views. I know this is a very tricky thing to keep straight these days because of all the “AI” shenanigans. But This time the values are off by orders of magnitude. So I figured it was best to report this discrepancy so you can look into this and hopefully find a way to correct it. Google Search Console reporting page views of 105, Koko Analytics dashboard shows page views of 361. (I was going to attached a PDF doc showing those two screenshots but this form only allows embedding from a URL). I know how to adjust this by updating the database files. So I am not asking for a solution to that part. I just want to let you know about this so you can check into it.
This issue occurred again on 3/27/26 Google Search Console reports page views of 99 Koko reports page view of 264 Please let me know if I can provide any additional details to help you identify what is wrong. I’m certain this is happen to every other website using Koko. Probably has do with the filters used for determining bot/crawler versus human page views.
And again yesterday. Google search console reports 91 page views koko reports 202 These values have never been perfect and I don’t expect them to be. But they should not be 2-3 times greater than Google search console. And it’s not a one time fluke. This is occurring every day now.
Hi Pete, I just wanted to let you know that I’ve read this and am following along. Can you elaborate on the “pageviews” reported by Google Search Console though? As far as I know, GCS only tracks the number of times your site appeared in Google search results and how many times a search result linking to your site was clicked.
Thanks for checking in on this topic. I understand your question regarding Google Search console, and I will explain a bit more detail. From the left-hand panel on Google Search Console you to to: Life cycle –> Engagement –> Pages and screens From there I am able to configure a report that shows me actual page views. This value has always been very closely correlated in the past. But in the past week or so I have seen a very large divergence. But as you mentioned, there is a chance this is showing only page views related to Google search activity and not total page view volume from all sources? However, when I compare these values to what I get from Google Analytics they tend to line up very closely. And GA is reading page views from all sources. I can also view live traffic through the Wordfence plugin. And they are also having difficulty differentiating between bots and humans. But for them, that is not a mission critical issue. One of the most common things they miss is the “applebot” crawler. Wordfence counts all of those hits as humans. And the pattern typically results in a nearly equal number of “visitors” and “page views”. Because every hit from applebot comes from a different IP. So that is part of the pattern as well. In the past whether using JetPack or Koko I typically get at least 3 times the pages views as visitors. But for the day of May 27th, 28th and 29th that ratio is not even close. As follows: May 27th: Visitors = 169 and Page views 264 May 28th: Visitors = 168 and Page views 202 May 29th: Visitors = 165 and Page views 346 May 25th was the worst of the bunch, showing 97 visitors with only 105 page views. When you examine the site traffic through any other tool those ratios are way off. My site typically gets at least 3-4 page views for each unique visitor. And today’s traffic so far is more typical. Koko currently shows Visitors = 52 and Page view = 203 This stuff is really difficult to troubleshoot so I understand the challenge. But I hope you can identify something and find a way to refine it further.
Thanks Pete. Today’s release (version 2.4) contains some tracking related changes which should, in theory, prevent even more non-human traffic from being recorded in the statistics. The User-Agent check has been expanded to also check for “prerender”, which some prerendering engines add. The browser’s visibility API is used to only capture a pageview if the document is actually visible. I’m curious if it has any effect on your numbers.
Thanks Danny, I just processed the update and I will follow-up in a day or two with some feedback. I want to give it time to accumulate data and make comparisons as I described in my previous post.
For 6/2/26, the update you provided was nearly an exact match. And to clear up one important detail, I was using Google Analytics as my comparison and NOT Google Search Console. Yesterday GA reported 95 visitors and 125 page views. Koko reported 91 and 133. Nearly an exact match. I’m sure when websites have very large traffic figures this issue is not even perceptible. It’s only when the traffic drys up to a trickle that things like this become more obvious. My site used to have really good traffic until the AI thing took over. Still struggling to find out how to operate a website in this new paradigm. Thanks for working out those details and publishing that update. Seems to have resolved the issue. Until the AI bots find another way to screw something up.
Thanks for the update @dryheat3 . Glad that seems to have improved it. Definitely keep an eye of things and don’t hesitate to let me know in case you suspect numbers are drifting off again.
This issue occurred again on 3/27/26 Google Search Console reports page views of 99 Koko reports page view of 264 Please let me know if I can provide any additional details to help you identify what is wrong. I’m certain this is happen to every other website using Koko. Probably has do with the filters used for determining bot/crawler versus human page views.
And again yesterday. Google search console reports 91 page views koko reports 202 These values have never been perfect and I don’t expect them to be. But they should not be 2-3 times greater than Google search console. And it’s not a one time fluke. This is occurring every day now.
Hi Pete, I just wanted to let you know that I’ve read this and am following along. Can you elaborate on the “pageviews” reported by Google Search Console though? As far as I know, GCS only tracks the number of times your site appeared in Google search results and how many times a search result linking to your site was clicked.
Thanks for checking in on this topic. I understand your question regarding Google Search console, and I will explain a bit more detail. From the left-hand panel on Google Search Console you to to: Life cycle –> Engagement –> Pages and screens From there I am able to configure a report that shows me actual page views. This value has always been very closely correlated in the past. But in the past week or so I have seen a very large divergence. But as you mentioned, there is a chance this is showing only page views related to Google search activity and not total page view volume from all sources? However, when I compare these values to what I get from Google Analytics they tend to line up very closely. And GA is reading page views from all sources. I can also view live traffic through the Wordfence plugin. And they are also having difficulty differentiating between bots and humans. But for them, that is not a mission critical issue. One of the most common things they miss is the “applebot” crawler. Wordfence counts all of those hits as humans. And the pattern typically results in a nearly equal number of “visitors” and “page views”. Because every hit from applebot comes from a different IP. So that is part of the pattern as well. In the past whether using JetPack or Koko I typically get at least 3 times the pages views as visitors. But for the day of May 27th, 28th and 29th that ratio is not even close. As follows: May 27th: Visitors = 169 and Page views 264 May 28th: Visitors = 168 and Page views 202 May 29th: Visitors = 165 and Page views 346 May 25th was the worst of the bunch, showing 97 visitors with only 105 page views. When you examine the site traffic through any other tool those ratios are way off. My site typically gets at least 3-4 page views for each unique visitor. And today’s traffic so far is more typical. Koko currently shows Visitors = 52 and Page view = 203 This stuff is really difficult to troubleshoot so I understand the challenge. But I hope you can identify something and find a way to refine it further.
Thanks Pete. Today’s release (version 2.4) contains some tracking related changes which should, in theory, prevent even more non-human traffic from being recorded in the statistics. The User-Agent check has been expanded to also check for “prerender”, which some prerendering engines add. The browser’s visibility API is used to only capture a pageview if the document is actually visible. I’m curious if it has any effect on your numbers.
Thanks Danny, I just processed the update and I will follow-up in a day or two with some feedback. I want to give it time to accumulate data and make comparisons as I described in my previous post.
For 6/2/26, the update you provided was nearly an exact match. And to clear up one important detail, I was using Google Analytics as my comparison and NOT Google Search Console. Yesterday GA reported 95 visitors and 125 page views. Koko reported 91 and 133. Nearly an exact match. I’m sure when websites have very large traffic figures this issue is not even perceptible. It’s only when the traffic drys up to a trickle that things like this become more obvious. My site used to have really good traffic until the AI thing took over. Still struggling to find out how to operate a website in this new paradigm. Thanks for working out those details and publishing that update. Seems to have resolved the issue. Until the AI bots find another way to screw something up.
Thanks for the update @dryheat3 . Glad that seems to have improved it. Definitely keep an eye of things and don’t hesitate to let me know in case you suspect numbers are drifting off again.