There were quite a few sessions at the Webtrends Engage 2009 Conference that spoke to the power of data integration and its positive impact on business results. Many companies start with an enormous datamart design and then move backwards – trying to make everything fit into their data model. It's a flawed process since processes continuously change… you'll never successfully implement it since it changes as soon as its defined.
Craig Macdonald, senior vice president and chief marketing officer of Covario, did an excellent overview on how to step into Cross Marketing Optimization. The presentation and session was called The New CMO: Cross Marketing Optimization. Craig didn't go into detail about each channel and the processes, so I'll try to provide additional detail with my perception of the process.
The process moves from small to large instead of vice-versa. Customer data is fragmented in organizations across channels, systems, processes, etc. Integrating customer data into a datamart requires the database implementation to be agile… a lot like building a spine. Each channel is a disc. The discs are combined into a spine. After the spine is in place, bones can be added, then meat to the bones, than skin to the meat, etc. Gross analogy, I know… but it works.
Defining the process within each channel is the first step. One example of a channel process is the steps that a prospect takes online from finding your business to converting, the Search Engine Channel. Perhaps they start with a search engine, then land on a page, then click to add an item to a shopping cart, then an order summary, then a conversion page. It's key to understand what search engine they were found in…
- What keywords did they search for?
- What was the landing page based on those keywords?
- What did they click to add the item to the shopping cart?
- Did they convert or abandon?
- Browser, Operating System, IP Address, etc.?
All of these pieces of data are critical in evaluating your funnel so that you can find the effective and ineffective pieces of each path. Every element or piece of meta data you can capture about the customer's journey is essential so capture everything, regardless of how insignificant. Once the data is in place, optimization of the channel is fairly simple.
Once each specific process is defined, captured and optimized, centralization of the data is the next step. Centralization of the data allows a company to now compare channels, their effectiveness, and most importantly, how one channel is impacting the other. Are you cannibalizing your efforts by spending money on pay-per-click on keywords you're already winning organically? Is your (inexpensive) purchase process driving people to (expensively) call your company instead?
Cross media optimization is essential if your company wishes to keep costs down and returns high. It's a complex effort that can take years (and change continuously), but once the pieces are in place, decisions can be made with confidence. It's not just a strategy for enterprise organizations, these can be essential to small businesses as well.
Craig noted that companies are grossly undercutting the resources needed to make significant gains in cross media optimization. He believes that ~10% of your marketing/IT expenses should be associated with analysis and optimization. That's a tough pill to swallow if you can't back up that expense with a return on investment. I don't doubt that that's possible, I just think it's a case of chicken or the egg. How do you justify 10% if you've not done it. How can you do it unless you spend 10%?
Perhaps walking into the investment as you step into the process is key. Optimization of a single channel may provide you with the return needed to expand your staff and resources.