Recently I started an affiliate case study. You can read about the strategy on that post. This post is going to discuss what I’ve done leading up to this point that is helping me laser target one aspect of my campaign.
The first part of my campaign will rely on a retargeting audience I’ve built over the last year. This may very well turn out to be the most profitable source of traffic for this case study. Now, don’t get me wrong. Assumptions in our business are most often wrong. But this one comes with a lot of data behind it. It’s that data / process you might find interesting.
Quick Retargeting Sidebar
What is retargeting? It’s a growing form of display advertising. It allows you to essentially tag visitors of your site and show them banner ads on other sites they visit around the web.
So they come to your site, leave, and visit another site that has a banner ad attracting them back to your site. It’s smart marketing and becoming smarter everyday.
Currently I’m staring at a 1.875 million person audience list (people that visited my site and were tagged with my retargeting code). This list was built via a health based website receiving organic search traffic.
This is a great start. Since I know the site that the audience was built from I have a very good understanding of the visitor profile. I know what type of people visit, what types of products / ads that have converted, and on top of that – they’ve told me exactly what they want to know.
How is that possible? How could I know exactly what this group of people want to know more about? We ran a survey to the 150K double opt-in subscriber list built from this same exact traffic. And that is exactly the data I’m banking on to help me make this aspect of my case study profitable.
Mining Data From Your Audience
First of all, if you haven’t gathered by now I’ve been lucky enough to have access to some pretty incredible resources. Here’s the recap
#1 – A highly SEO’d top ranking website in the health market (great analytics data)
#2 – A 150K double opt in subscriber list (large pool of potential survey responders)
#3 – Survey data from thousands of responders via that email list (pure gold straight from the subscriber base)
I realize not everyone has that same level of resources to work with. But the concepts are still the same. This could work very well with a much smaller list. In fact it may work even better as it’s possible to cultivate a closer relationship with less people.
Back to the data.
Based on those resources I’ve got a pretty good picture of the type of “persona” I’m dealing with. On top of that we were able to pull 3 topics that were repeatedly requested within the survey.
The survey was uber simple to setup. We went the low cost route and it worked well. No survey monkey here. Although in hindsight it might have cost more but that may have been worth it in the time spent going through the responses (oh well).
It was a simple process
#1 Created a Google Docs Form
- Kept it simple – 1 or 2 questions with a text box.
#2 Embedded the form onto a private, no-index page on the site
#3 Sent multiple emails to the list regarding the survey
If you have a site and a list you can do this too.
The Next Move
With my topics in hand I went out and found multiple products for each topic. The plan is to create multiple banner variations for each topic and rotate which offer they go to. Ideally this will lead me to decent CTR’s and a good handle on which products convert.
Sounds like easy money right?
Well as with anything there are potential pitfalls. As I see it those pitfalls look like this:
#1 Retargeting audiences dwindle over time.
My impressions may not be nearly as significant as the size of my audience right now. Less impressions equals less traffic. We’ll see.
#2 In general retargeting traffic does best going back to the original site they were built from
The traffic flow in an optimal retargeting campaign looks like this Site A to Site B back to Site A. In this case study our traffic flow will look like this – Site A to Site B to Site C.
#3 My data is Misleading
Sometimes people don’t really know what they want. Sometimes they think it’s one thing and it’s actually another. Sometimes your data set isn’t statistically significant enough (even though I think mine is).
Only time and traffic will tell the real story. A story which I know we’re all eager to see play out. As soon as I have more data I’ll update the case study and let you know.
If you’ve got any questions on what I’m working on – just let me know in the comments below.