Bookmarks / Uniques Ratio

What is considered normal, poor, or great?

[ijan]

I was wondering what is the ratio of bookmarks to uniques on your web sites? And what is considered normal, poor, or great? I am tracking the bookmarks via the hits to favicon.ico. Since IE users account for the majority of my users (over 90%), I think this ratio gives a fairly good idea of what users think about a web site, whether they find it useful or not. According to the December logs, 5% of my site users (unique) bookmarked a page from my site. I think this is a good ratio, but I have no reference to compare my result. So it would be great if you share your numbers.

[mark_roach]

I guess from the number of responses to your question that not many people are tracking this metric :)

To get a more useful ratio you probably need to disregard the "uniques" with no referral string, as these are most likely from repeat visitors who have probably already bookmarked your site.

Your figures may also be skewed if you get any referral from search phrases that are obviously not related to your site.

I find that wading through the raw logs is a pretty good method of determining the effectiveness of a site. Just pick a few visitors at random. Check the referral string to see what they were looking for and then follow their path through your site. You can usually work out if they found what they were looking for.

[ijan]

Thanks Mark for the only reply. :) Disregarding the "uniques" with no referral string is a great idea, and it will definitely increase the ratio quite a bit, which is also great. But what I have realized is that the bookmarks/uniques ratio has always been stable. Here is a list of the ratios for the last six months:

December 5.5%
November 6.2%
October 5.6%
September 4.1%
August 5.1%
July 5.5%

The unusual sharp decline in September is most probably because of the Sept. 11 attack.

[tedster]

When you talk about bookmarks, I assume your talking about favicon.ico hits? If you have a way to track Netscape bookmarks, I'd be very interested. Otherwise, I'd also suggest filtering out the non-IE user agents, even if they come with a referer.

I haven't made much of the favicon number to date because Explorer has been pretty buggy in its behavior in that area...but I'm open to ideas.

[ijan]

tedster,

No I have no way to track Netscape bookmarks, but I think it is easy to extrapolate the results. For example 5% of my visitors are using Netscape browsers (it is sad to see Netscape lost so much ground) so 5.5% becomes 5.6%, and I think it is negligible given the current state of Netscape use. I'd be delighted to filter all non-IE user agents including search engine bots in addition to uniques without referral strings because it will only improve the ratio, which is great.

[tedster]

Getting a more accurate ratio would be valuable. To confuse the mix just a little bit, I've heard about some new non-IE browsers supporting favicon with bookmarks -- I can't remember which ones, maybe Moz?