For many marketers, the end of the year marks the end of their budget cycle. It is certainly not revealed scripture that marketing budgets are limited and that any new expenditure must be carefully justified. In this process, it is very important not to under or overvalue the contribution of any marketing initiative. Organic search is particularly difficult to evaluate, because the point at which the user connects with your marketing message can be almost anywhere in the purchase cycle—from initial research to post purchase satisfaction and loyalty.
Just a few years ago, the entire conversion process was easier to model, because the system was simpler. As we have gone to a multiscreen, omnichannel, integrated marketing system, the number of marketing touchpoints has exploded. Each is a variable, and with each variable there is often an entire marketing program—email, affiliate, paid search, display, organic search—that must be evaluated during the budgeting process. As marketing programs are more tightly integrated, it is increasingly difficult to understand the marketing system dynamics and accurately tease out the impact of each specific marketing activities and initiatives.
It is essential that you build and use an attribution model if you plan to optimize the results of your marketing spend—get the most bang for your buck. Sure, e-commerce sites can identify and track an individual’s path from first encounter to purchase, but the proverbial rubber hits the road, when it comes to attaching a specific value, a valence, to each action along the path, thus building an attribution model. Today, building such models is an essential part of the marketing and budgeting process. Marketers develop and use such models to provide frameworks for evaluating and interpreting the importance of each digitally recorded action as the consumer moves through the purchase process.
There are a number of basic types of attribution models, but each online business is so unique in how it specifically goes to market that one size does not fit all. An individual model should be built that reflects the company or brand’s specific marketing philosophy and activities.
If you use organic search as your tip-of-the spear prospecting tool, a model that ascribes the most value to the first click (a first click model), may understate the role of the other activities that go into your marketing effort. In the past, first-click models gave new e-commerce-only businesses an easy way to frame brand recognition and customer acquisition, but they are much too simple for complex marketing programs.
Another popular type of model which gives the most credit to the last event that happened prior to purchase often overstates the role of the activities that occurred subsequent to the initial contact that may, in fact, have resulted from an organic search. Thus, they understate the impact organic search has on the marketing success.
Other types of models use more linear approaches, where each touchpoint is given equal credit for the final conversion. But not all events are equal, and consumers take many different and often unpredictable paths. The more sophisticated models ascribe the majority of the credit to the first or last click and then gives varying valences to the other touchpoints. Other models, so-called “time decay” models, recognize and respect the length of the purchase process and give the most credit to the most recent activity and less to those that occurred days or weeks before. To decide what model will more fairly depict the value of your organic search program, you need to consider where and how organic search fits in the purchasing process. Your analytics will help you recognize the paths that customers coming from search typically follow.
As you go about building your budget, look closely at the role that organic search plays in your overall marketing program. Build an attribution model that recognizes how your consumers relate to your site through search. Then, as you budget, you may find that you are giving too little credit or too much credit to organic search. In my experience, most marketers find themselves wishing for better search organic search results because they find that it is a key force, no matter what type of model they are using.