significant discrepancies in reporting between koko analytics and matomo

  • We are currently utilizing Koko Analytics for WordPress alongside a default installation of Matomo for website analytics. During our comparative analysis, we’ve observed significant discrepancies in the reported unique daily visitors and page views between the two platforms.

    Given that we rely heavily on accurate data to make informed decisions, these inconsistencies have raised concerns within our team. Specifically, we have noticed the following:

    1. Unique Daily Visitors: The numbers reported by Koko Analytics differ markedly from those reported by Matomo.
    2. Page Views: The discrepancy in the count of page views between the two platforms is also quite large.

    To ensure the integrity of our data, we have checked the following:

    • Both tools are configured to track the same set of URLs.
    • Neither system is showing signs of being affected by caching issues or data sampling techniques.

    Despite these precautions, the gap remains, which brings me to reach out for your assistance in understanding these discrepancies.

    Could you please help us understand:

    1. What tracking methods Koko Analytics employs (i.e., Cookies, JavaScript, fingerprinting, etc.)?
    2. Are there specific settings that might account for these differences?

    I appreciate your attention to this matter and look forward to your prompt response. Understanding these discrepancies is crucial for our data-driven initiatives and business intelligence endeavors.

    Thank you for your time and assistance.

    Best regards,

    Robert

Viewing 4 replies - 1 through 4 (of 4 total)
  • Plugin Author Danny van Kooten

    (@dvankooten)

    Hello Robert,

    With the default settings, Koko Analytics relies on a cookie to determine unique page views. No fingerprinting is done as we believe this intrudes on the visitor’s privacy. A cookie is easily cleared, but fingerprinting is harder to bypass.

    Data collection is initiated from a tiny client-side script so that it works with pages served from browser cache.

    Some things that could be causing Koko Analytics to report lower counts:

    • By default, Koko Analytics will respect the browsers “Do Not Track” setting.
    • The client-side script that initiates the data collection is on some filter lists, like EasyList which is used by a lot of ad-blockers. I believe the same goes for Matomo though.

    Could you elaborate on the discrepancies you’re seeing? Ie which one is lower and by what margins?

    Hi Danny,

    thanks for your prompt response. Figures from Koko Analytics are not lower but way higher than Matomo, examples (date: unique visitors/page views Kokos Analytics vs Matomo):

    13.9.: 254/448 vs 143/263

    14. 9.: 362/650 vs 230/428

    Could it be that Matomo doesn’t count that many bots, as at least from their changelog I see that the bot list is updated regularly?

    Best,

    Robert

    Plugin Author Danny van Kooten

    (@dvankooten)

    Koko Analytics doesn’t count most bots either, so I doubt that accounts for the big discrepancy. Bots are filtered out by having the data collection rely on JavaScript (which 98% of bots do not run) and by filtering the the User-Agent header for common bot values.

    Some things that may be worth checking:

    We use the plugin “WP-Matomo Integartion” to integrate the tracking code, currently by using /js/index.php as small proxy for calling the tracking script, so that the chance to get blocked by adblockers is lower. We don′t have any iframes.

    We also currently set IP anonymization in Matomo to the last two digits (192.168.xxx.xxx) for privacy reasons, but we find it unlikely that most users visiting our website will have the same IP range and even if, this would not explain the discrepancies in page views.

    Anyway I will try setting 

    trust_visitors_cookies = 1

    in Matomo as advised at https://matomo.org/faq/how-to/faq_175/ and will check if this makes any difference.

    If you have any further ideas, please let me know. Anyway thanks for your prompt help!

    best,

    Robert

Viewing 4 replies - 1 through 4 (of 4 total)
  • The topic ‘significant discrepancies in reporting between koko analytics and matomo’ is closed to new replies.