eMarketer recently did a study on US Online Ad Spend with projections to 2015, forecasting banner ad spend to grow from $7.61B in 2011 to $11.73B by 2015. While this seems to be a positive trend, much of the interest in banner ads lies in the fact that it’s downright cheap due to ever-growing inventory. Despite its popularity, Marketers should be careful in assessing its usefulness and purpose.
I was recently part of a digital campaign for a large financial services firm designed to promote an innovative online lead generation tool. Our primary marketing channels were paid search, and banner ads (on non-social media platforms).
The success metric we allocated to both channels was total number of generated leads.
As the campaign progressed, we generated strong lead numbers from paid search, with conversion just south of 1% and cost per lead coming in within our target range.
Our banner CTRs were in the neighbourhood of .06%; lower than our anticipated target of 0.1%. (a baseline which we determined was reasonable based on research of industry averages).
Initial reactions were that copy and creative were not compelling enough, so we re-designed our assets; from leaderboard to big box, static to flash, going through the standard rounds with creative and usability experts.
With the new banners in play for several weeks, the results were marginally better.
Based on the anemic overall performance, we chose to go dark on ad buys for the next couple of weeks to determine the impact on leads. During this dark period, we noticed no significant drop in leads. To some within the team, this was enough justification to pull the ads altogether. Why not? In terms of the success metric we defined for banner ads, it failed to produce.
I noticed however, a substantial drop in overall traffic to our campaign landing page – nearly 45%! While the audience from banner ads failed to convert, we were losing a substantial number of eyeballs to a key part of the site.
We needed to allocate a different success metric to banner ads – that of traffic. And of course this made sense in hindsight. After some discussion, we hypothesized that people who clicked on our banner ads were likely high in the sales funnel given that direct traffic volumes were strong and conversions low. Couple that with an overall time on page well above our site’s average, and with a respectably low bounce rate. We characterized this audience as either conducting product research or having a peripheral interest in a financial solution.
While people using search can be anywhere in the Sales Funnel, through keyword targeting and heavy upfront research on our part, we were able to capture a proportion of the population closer to the bottom of the Funnel and more likely to convert.
From our perspective, there were really only two types of people on the Web – those in the product research phase and those that were ready to buy, and at the time, we were only concerned with the latter. While our ultimate goal was to increase leads, we felt this was too short-sighted an approach and one that needed to be expanded to consider a longer time horizon – effectively targeting those higher up the funnel.
To that end, we re-instated our media buy and took a more concentrated approach to being visible within contextually relevant, high-volume portals. While our campaign targeted those that were not likely to convert immediately, we were satisfied with the notion that we could at least plant a branded seed in their minds.
So what is the ROI on being a seed of an idea?
Interestingly enough, this was also a time when we were considering the concept of re-marketing – a compelling, and slightly controversial strategy designed to market to and capture users who have previously visited a company’s web property but failed to convert. Alas, I’ll leave this discussion for another day.