How to Monitor the Quality of StumbleUpon Advertising Campaigns

In a guest post Neil Matthews of ClickQualityConsultant.com writes about how we can monitor the quality of paid visits from a StumbleUpon advertising campaigns.

Darren advocates using paid advertising from StumbleUpon in his post Run a StumbleUpon Advertising campaign for your blog, as a way to build your blog’s audience. In this post I would like to expand on this and talk about how to monitor the click quality of your paid Stumbles to see if this type advertising is the correct fit for your blog.

What do I mean by click quality? I mean that the clicks are bringing a return on the advertising investment you make. This will vary from blog to blog. You may be after an increase in subscribers, for people to click more often on ASsense ads or that they contact you for consulting services. If you are spending money on StumbleUpon clicks you need to know how the campaign is performing and if this type of advertising works for your blog.

Firstly a quick recap on StumbleUpon (SU). I like to think of SU as TV channel surfing for the net. The Stumbler installs a toolbar into their browser, sets the type of site they are interested in and begins to stumble. SU selects a site at random from their database which matches your likes and sends it to your browser for your surfing pleasure. If the site is of interest you, you may engage and begin reading more deeply, you can then grant a thumbs up or down to the site to show if you approve of the content of not. The other option is just to skip past the site surfing for a new channel.

Running a paid SU advertising campaign, you pay 5 cents per display to have your site presented to the Stumblers in the demographic group you select. Your daily cost is set by the number of displays you want per day. For example 500 displays per day will cost 25 USD.

Stumblers are notoriously fickle, and if you are paying for clicks, it is important to check if the campaign is producing quality clicks, or are people just clicking away from your site. SU has its own quality check, it shows how many people have given you the thumbs up or down, this is reliant on the visitor, we need more quantifiable analytics to see if this form of advertising is bringing you any return on your investment.

The first thing to do is to highlight which referrals from SU are paid and which are organic. To do this, I amend the landing page of my campaign by adding a parameter to the end of the landing page URL for example:

Note that I use su rather than stumbleupon. Using the term stubleupon in paid ads is against their editorial policy and your ad will be rejected.

Next I analyse that traffic with an analytics package. For the purpose of this post I am using Google Analytics, it is free and very simple to use. Installing GA is relatively simple. To collect metrics you will need to create an account and then install a piece of JavaScript code onto every page you want to monitor. I have the code installed into the footer of my wordpress theme so that every page is monitored. Please refer to the Analytics site for details on installation.

Once my SU ad is running and I have collected a decent enough number of clicks for statistical analysis (a couple of hundred should suffice) I move onto the process of identifying the quality of those clicks.

From the analytics package I can get an overview of the landing pages on my site, as we can see from the screen dump I received 179 clicks from StumbleUpon

Drilling down into this metrics I am presented with the behaviour of the visitors from the stumble upon source

This give me good and bad news, only 12.35% of the visitors are bouncing away from my site immediately, there is a certain level of stickiness about the page and people on average are spending 41 seconds on my landing page. I am getting the engagement I was looking for, the disappointing side of this campaign is that nearly 90% of the visitors are reading one page and then leaving, there is no depth to their visit. That is the area I would focus on improving.

I think my problem is that I am sending Stumblers to my home page rather than to a specific landing post which can draw readers in more deeply.

Do I think I was getting good quality clicks from SU? Not really, the action I want is for visitors to move from the content into my consulting page. I would probably be better served sending traffic directly to that page rather than the home page.

Using this information I could the split test another landing page with a different source parameter and see how the two stack up against one another.

How do you decide if SU works for you? I would say the following metrics need to be analysed for your campaign:

• Bounce rate – are visitors staying or moving on straight away. A sub 50% bounce rate is good
• Length of visit – How long are people engaging on your sit, if it is only a second or two, less than the time it takes to read you posts, this is a bad sign?
• Depth of visit – Are people reading more than one posts, have you caught the readers attention?
• Goals – Google analytics has the option which allows you to set goals, in this example the goal was to move from the landing page to my consulting page. My research suggests that a 1% conversion rate of visits to goals is the minimum you should look for.

I have used this method across a number of different platforms including Facebook social ads, and when I buy advertising banner space. Monitoring and testing the quality of your paid advertising is the key to a good return on investment. If you don’t get the quality you need from that advertising source, drop the campaign and spend your money on quality clicks. If SU is poor consider Adwords, Yahoo, Facebook, MySpace or one of the many other paid internet marketing programmes.

Can I add a caveat to close this post? You may be tempted to use this method to reconcile the number of clicks paid for to the number of clicks received, but I would say that Google Analytics data can be wrong due to people running their browser with JavaScript disabled. If the code on you page is not activated, then no visit data will be captured. Click fraud investigations require log file analysis tools to ensure the validity of your claims.

Test your quality, make incremental changes and test again. This is the way to get the best bang for your buck.