According to this recent article, StumbleUpon drives more traffic than Facebook and Twitter.
I commented on that a while ago, in my “Why We Love StumbleUpon” article.
But how useful is that traffic?
Grading Different Types of Traffic
Not all visits are created equal.
Depending on what business you’re in, a direct visit is probably the most likely to represent someone who really wants to engage with your site. That means they’ve found your URL somewhere and typed it in, or they’ve saved your site as a bookmark/favourite.
Referrals are probably next best. They’re visits that come from clicking a link that points to your site. So they count as a recommendation, and you could expect that visitor to be warm, or at least interested.
The third type is organic search. The visitor has done a search and is clicking through the results to find what they need. They may be the least warm or favourable category of visitor, but they may become fans if they do find what they want.
Where do you fit StumbleUpon traffic into this? Because people are just passively browsing, it has to be lower-value than any of the regular types. But, like search visitors, they could still become fans.
We get spikes of SU traffic periodically, and had one this week.
Here’s an interesting graph from Google Analytics, showing the traffic to this site over the past 25 days.
- The green line is the total number of daily visits. You can see the “SUrge” last week, which lasted over the weekend and Monday. It settled down to normal yesterday (Tuesday). The last data point is today (half-way through).
- The blue line shows the segment of visits over 5 seconds long.
- The orange line shows visits where at least 3 pages were viewed.
I would expect the “cheap” StumbleUpon traffic to reduce the average length and depth of visits. But is it so?
I sampled the data from 4 days: 2 non-SUrge days, and 2 where we were getting a lot of SU traffic. Here’s what I found.
Visits over 5 Seconds Long
- Non-SUrge Days: 27.7% and 26.6%
- SUrge Days: 27.8% and 26.3%
Minimum 3 Pages Viewed
- Non-SUrge Days: 9.6% and 10.4%
- SUrge Days: 4.6% and 4.5%
So the visit length seems to be unaffected. People are just as likely to read the page as if they had come via any other channel.
However, the proportion of visitors who viewed 3 or more pages was halved! I suspect that means that the majority of StumbleUpon visitors only viewed one page.
More Accurate Stats
Here’s a comparison from Google Analytics, showing the differences between SU traffic and all non-SU traffic (click image to enlarge):
- SU users view far fewer pages (1.18 vs. 1.61)
- SU users spend far less time on the site (48s vs. 2 mins 3s)
- But, interestingly, they have a lower bounce rate (61% vs. 70%)!
What About Real Value?
One of the easiest goals or performance indicators to measure from this site is new newsletter subscriptions.
Here’s the chart from AWeber showing new sign-ups over the past month. Judge for yourself. I see no increase at all in the past week – when overall visits was double.
From this, I’d sum up that, while StumbleUpon users are just as likely to stay to read as anyone else, they will probably leave after they’ve read the stumbled page, and are not likely to engage further with the site.