We’re working with a client who’s got a problem. They want to measure the effectiveness of a new banner campaign that’s rolling out. The challenge is that they don’t sell anything on their website – all sales occur through their retail distribution partners (major consumer electronics chains and online retailers)  – and they don’t "own" the full sales funnel.

This issue is pretty common and has been the million dollar (probably much more than that’s been spent trying to answer this) question for advertisers – how do you measure effectiveness of a campaign if you can’t track through the entire sales process?

While I don’t have the answer yet, we’re researching some new solutions and I’ll post more of our findings once the project is completed. What I do know is that we have a test plan in place. Here’s how we’re thinking about doing it…

Tie the banner links to individual ads
Sure, this is web analytics 101 stuff but it can’t be ignored. We’re complicating it a bit by breaking down creatives into the different formats (rich media, standard banner, etc.) and sizes the ads will be created in and then further subdividing by the sites on which the ads will appear and the messaging used within the ad.

Measure where the traffic comes from
We did an assessment of potential DMAs to target based on a propensity to purchase our client’s product in those areas. We then overlaid the data with store sales data as well as costs of running the ads to come up with where we’d get the most bang for our advertising buck. Based on that information and the national campaign information, and our control group, we’ll track what response were getting from the targeted DMAs. We’ll also look at the data we receive from retailers about unit sales and see if sales spike during the time of the campaign (more to the point, do sales spikes line up with site traffic spikes).

See how well the campaign drives consumer information
Since we’re measuring the inbound traffic, we’re going to take a good look at what content was viewed on a campaign basis. Measuring this isn’t easy but we think we’ve got a simple but slick way to do it and rank the campaigns against each other to see what’s driving the most product interest. I can’t really go into the details of how we’re doing this yet – I want to see some solid data first. There will probably be a lot more to come in future posts.

Compare campaigns to overall site traffic
Since the campaigns are going to the same pages as normal traffic (no microsite on this one) we had to implement some page changes to know if the traffic came from the campaigns or other sources. The ability to compare campaign traffic is important if we’re going to get a good understanding of what the effectiveness of the campaigns are compared to general product information found on the site through other sources.

Monitor store sales data
To the extent that we receive the data, we’ll compare the information to same store sales prior to the launch of the campaigns. We’ll also monitor this on a DMA basis (along with the site traffic) and see if we can optimize our ad spend by retargeting campaigns to different areas that are performing under sales goals.

That’s all for now. Let me know what you think about this approach. Feel free to post a comment…