Measuring stickiness?

and how to improve it

[danny]

My site is averaging well over 3000 visitors a day, but they are viewing an average of only 2 pages (down from 2.5 a year ago) - most come in off a search engine to a specific page and never look at anything else. I'm seeing maybe 30 people a day bookmark the site (counting favicon accesses but excluding browsers that fetch favicons automatically), and one or two subscribe to my mailing list. I'm not selling anything - I just write stuff and want to be read - so these three things are about the only measure of "stickiness" I have.

What other measures of stickiness are there? And what kind of things help encourage people to look at other pages on a site? (Does writing "bookmark this site now" increase bookmarking rates?)

[bobriggs]

30 bookmarks a day is fairly remarkable. 2.5 pages/viewer sounds low for a site getting that many bookmarks/day. So I'd say (without knowing anything about your site) is there a way on your pages that you can tell these bookmarked viewers that there's something new? Just one suggestion.

Another stat I like to get and have to figure by hand: How many are coming from google or other sources that are one hitters? (And you kind of have to guess about AOL hits and other proxy cache hits) I also like to get the stats from the non-one hitters and their total page views and get a new number (ex: 12000 page-views, 3000 uniques, but there were 2000 one hitters. That changes the average page view from 4.0 to 10.0)

12000/3000 = 4.0
12000 - 2000 = 10000
10000/1000 uniques = 10.0

[Mikkel Svendsen]

Return visitors are very important to measure too. You need cookies or log-in for that.

Also, you may want to filter out "lurkers" from "surfers" - a lot of people probably only see one page and leave in under 10 seconds - I sometimes like to seperate those (that must have got to the very wrong site and realized it fast) from the other ones that show some interest in the site.

Time spend is another good "sticky-factor"

There is a lot of things you can do to improve bookmarking percentage, sign up rates and any other goal you have. What you need is to decide on ONE way to track the steps, goals and conversions you are interested in and then start tweaking things on the site: Add a "bookmark" link, change the colour, add a privacy stement etc and watch how it influence the conversions. If one change improve the conversion keep it - if not go back to your previous version that performed the best.

This can take a LOT of time but the effect of it can be very dramnatical. For most sites that I have done this for (what I usually call: Micro-Optimization – or "improving Web-ROI") the conversion on specific actions (like sign up rates, bookmarks etc) has increased by 2-3 times. In one case I managed to improve it by almost 10 times!

You are on the right track. Working out what is YOUR most important goals is the first – and most important – step.

[tedster]

I work with one site that has a particular entry page which seems to be "all or nothing."

This page actually gets a newsletter sign-up at the highest rate of any on this 3,000 page site. The search engine visitors who stay after the first page average over an hour per session and 21 page views.

And yet this entry page also has the highest rate of one-hitters from search engines on the site - 41% over the last three months, and even higher for just Google traffic!

It just goes to show what valuable information is available in those finely tuned stats. If I only looked at the 41% one-hit wonders in isolation, I would definitely have made changes to the page.

But the page seems to be extremely effecgtive at targeting and qualifying visitors, so it remains untouched.

[danny]

30 bookmarks a day is fairly remarkable

With 3000+ visitors a day, though, that's only 1%. Actually, percentages are a better way to look at the stats: 1% bookmark, 0.03% subscribe for email, 66% of visitors view only one page and the other 33% average four page views.

If single-hit visitors actually read the reviews they've hit, that's cool, but it's impossible to tell.

[Mikkel Svendsen]

The stats we get from logs and trackers is just numbers - we have to do the anlysis, using our experience and good judgements. I don't think the question of 1% bookmark is good or bad is so important - what is important is if you can do better :)

- There is only one way to find out and it sounds like you are on the right track ...