The StumbleUpon Experiment

Do people “stumbling” across the net have their wallets open while doing so? Will they stop when they see a commercial site or keep on “stumbling” by? That was today’s quick experiment: advertising through StumbleUpon.

StumbleUpon offers a toolbar that lets you find new websites by “stumbling” through sites recommended by other users of the toolbar. You just click the “Stumble!” button on the toolbar to visit a new site, and can rate each site with “thumbs up” or “thumbs down” buttons.

I first joined StumbleUpon and downloaded their toolbar in May, but had forgotten about it after upgrading to Windows Vista. I was recently reminded of it when seeing a couple hits in this blog’s referrers after another blogger reviewed this site in StumbleUpon.

StumbleUpon sells website visits for just $0.05 a view. After a 4-figure sales day at VisitorBoost yesterday, I decided to put a bit of that into StumbleUpon. In addition to the ultra-low cost for guaranteed traffic, StumbleUpon isn’t prone to click fraud since they neither syndicate their ads nor have any mechanism to bring up a specific site, and they also offer targeting according to their users’ stated interests. I really didn’t expect these users to buy, it’s not the mode they’re in when using the toolbar, but I’ve been surprised by sales from unexpected sources before, so I’ll try everything once or twice before discounting it.

I created a new tracking URL for the ad campaign and filled out the form to start it up with StumbleUpon. I chose the Web Development category hoping to reach website owners, and restricted my ad to only U.S. visitors for the best chances of a purchase. A few hours later, the ad was approved and running.

It was evening when the ad began to show, and at 250 visitors to the site, I hadn’t made any sales. I paused the campaign for the night, planning to continue it during the afternoon the next day to show during business hours. Another 250 visitors came through after I resumed before I considered that to be enough of a trial for now.

One thing I can say about StumbleUpon that I can’t about any other service I’ve advertised with is that they are 100% accurate in reporting and 100% free of fraud as far as I could tell. My tracking program tracked exactly 502 visits where StumbleUpon shows 500. That’s 500 paid for, one from the reviewer that approved the campaign, and one click by myself to test the URL before starting.

A quick analysis of the traffic backs up the presence of interest targeting fairly well. More than 90% of the visits were by computers running the Firefox browser with a high screen resolution. That matches up to what you’d expect from the web development community.

The results? No sales, 96% of the visitors only viewing the landing page for the campaign, and none of them rating the site positively or negatively with the toolbar. That says to me that most people that see a commercial site, or a site that isn’t immediately interesting to them with StumbleUpon will simply move on and click the “Stumble!” button again.