Awstats or other server analytics vs. cloudflare web analytics

I’m a new (paid) Cloudflare customer, very disappointed with the limited insights into my visitors available through Cloudflars’ Web analaytics compared to awststs or other analytics on my Web server. When content is served from Cloudflare’s cache, I don’t receive the requests, so my insights are limited to those available from Cloudflare (unless I want to inflict something like Google Analytics on my visiotrs, which I don’t).

I appreciate the difficulty of serving customers with huge variations in traffic volume and in the nature of their Web sites. But for my use case, Cloudflare’s Web analytics are almost useless, and as soon as I saw the first few days of stats, it was immediately apparent that this is the strongest factor against using Cloudflare on my site. Pros to Cloudflare: Better ability to serve visitors during traffic spikes, faster page loading all the time. Con: Loss of visitor insight I was obtaining from awstats on my server.

It looks like Cloudflare’s Web analytics were designed from a security/sysadmin perspective, not from a an editorial/design/content management perspective.

As a publisher of an informational site, I don’t really care what country, domain, or IP address my visitors come from. I realize that analyzing this info in the aggregate may be useful for threat detection and mitigation, but I would rather IP addresses of visitors to my site not be recorded. I couldn’t care less what time of day they visit. Can I hide the map and time graph, please, or move them to a secondary page of sampled long-term data? Similarly, on a long-term basis, there is some value for overall site architecture in knowing what browsers or devices or operating systems visitors use. But that’s fairly stable, and belongs on a secondary page for aggregated or sampled long-term data.

I’m less interested in counting total absolute traffic volume, which is inherently unreliable anyway. I’m more interested in the distribution of traffic between different pages, sections, or downloads, and in trends in vistorship (by pages visited and relative volume, not by device or OS or country) over time.

What matters to me every day, at the top of the primary stats page, is the distribution of traffic by pages (and downloads – awstats separation of page views and file downloads is useful although not essential) and the search strings that bring people to my site.

I suspect that far more site owners care about what page is trending today vs. last week than whether more visitors today are on Apple devices or Windows than last week.

Which article or blog post or section of my site is trending up, so maybe I should feature or update it? I want to know, even if it isn’t in the top 15 pages shown in Cloudflare analytics. Which is down, and maybe doesn’t deserve home page or top-level menu placement? I want (and regularly use in awstats) a view that shows traffic to all pages, not just the top 15, on one page. (And if you could, even from sampling, identify and have an optional display to flag up/down trending pages today/this week/this month/over a user-defined time span, or relative to last week/last month/lats year, that would be awesome.)

If I make a design or navigation change, how does that shift which pages or sections visitors find their way to? This requires comparing traffic by page – year to year, month to month, day to day.

What are visitors searching for, so that I can deliver it and make it findable? What does Cloudflare do with the search engine query strings it receives, when it serves cached content? Unless I am missing something, these are completely absent from Cloudflare analytics. But they are, obviously, invaluable for editorial decision-making and SEO. Why not include search engine query strings from referrer headers in analytics, without recording what IP address they came from?

With respect to referrers, knowing what domain name sent me referral traffic is of little use without a pinpoint URL that enables me to identify what high-traffic site mentioned or linked to mine, and in what context.

As of now, I think that the advantages from using Cloudflare of resiliency during traffic spikes outweigh the loss of editorial, design, and navigation insight into what my visitors are reading, viewing, and searching for. But it’s not an easy choice.

You could make that choice in favor of Cloudflare easier by looking more closely at what awstats or other server-based analytics offer site publishers, and providing more of that functionality.

Hope this helps in your future development. Happy to talk further about my use case and needs.

Do other Cloudflare users who don’t use Google Analytics, but have relied on awstats or other server-side analytics, have these same issues?

This topic was automatically closed after 14 days. New replies are no longer allowed.