The hunt for true ROI

21-May-2012-Hunt-for-ROI OMD

This post is about our recent work with an e-commerce client and how we’ve adapted our traditional way of evaluating digital media to focus on “true ROI”. What does “true ROI” mean? For us it’s about shifting attention from channel specific returns from paid search or display to a broader view of how our client’s business is performing (incorporating all sources of traffic) and then directing investment to the areas that need it most, even when that means our buying and bidding ROIs take a dive.

This shift in strategy is demonstrated most simply in our paid search investment. A year ago, we worked to a strict cost-per-acquisition target, delivering the highest possible ROI as defined by our ad-serving data. This optimisation programme naturally led us to harvest the demand that was easiest to convert, thus giving us the comfort of rock bottom CPAs and confidence that paid search investment was doing its job. And it was, if you just looked at the performance within channel.

Where we’ve moved to is a more interesting, business-centric model for investment. Now we focus on identifying product areas that are doing well and not so well in terms of total sales from all sources, on a weekly basis. We shift investment to the areas that need support, accepting that our paid-search performance metrics will worsen. Why do they worsen? Because we’re moving out of high quality score keywords and product areas where we don’t have much competitive bidding, to products under heavy pressure from competitors or where our client is less well-known. So our PPC results suffer but by keeping our eye on the bigger picture, the total business performance of our client, we can be confident that we’re doing the right thing. Basically if one product is naturally doing well by getting traffic direct to URL or via organic links why spend money on PPC just to deliver a strong channel ROI? It makes more sense to divest from this high performing product and re-allocate paid media to products that are under-performing, thus helping overall revenue.

Three connected changes have enabled us to take this approach. First we have been able to get closer to the client’s business through attendance at weekly commercial meetings, where product performance across the portfolio is analysed and thriving and struggling products are identified. Second, we have full access to our client’s web analytics, providing a view of all traffic sources to the website, enabling us to invest with full visibility of the broader business. Thirdly, we’ve brought in the best bid management technology which helps our PPC team do more analysis and less manual ad-words manipulation. These improvements enable us to respond to our client’s overall commercial needs much more quickly, and allocate paid media to products that will deliver true incremental ROI, instead of just converting sales that would have come anyway through other free channels. What underpins all these changes is reliance on better and “bigger” data. How agencies are set up to handle Big Data is the subject of another post, however it’s fair to say that traditional ways of recruiting, investing in technology and analysis will have to change if as we start competing outside our familiar media agency sector.

The attraction to the client from this approach is obvious; we’ve been able to reduce media budgets in areas that don’t need paid support by taking a broader view of where investment is needed. This has released spend for improvement of the website, thereby increasing conversion and revenue from all traffic sources. Over time we hope to strengthen the brand and further improve on-site conversion, reducing the need for online traffic-driving investment, so ensuring that any paid media truly delivers incremental sales and genuine ROI. And rather than viewing this as a dangerous objective for a media agency, we think it is the reality of our journey to become a full service marketing investment partner with our customers.

Post by Ian Cairns, Business Director


About Author


Comments are closed.