Optimizing cross-channel marketing spend

Online marketing is booming. According to statistics from eMarketer.com, the percentage share of online marketing dollars in the U.S. is expected to increase from 32% in 2013 to 37% in 2016, crowding out the share for traditional print advertising.

For marketers, this trend implies that cross-channel marketing, mixing online/offline channels, will become increasingly prevalent. If you are a traditional direct mail marketer, should you redirect your ad dollars toward online or email marketing? If you are already heavily invested in online and email marketing, how do you best allocate your existing marketing budget across different campaigns? When compared with single channel operations, marketing budgeting decisions in a cross-channel environment are much more challenging because marketing campaigns across channels can interact and campaign operational decisions vary from channel to channel.

A two-level hierarchical perspective can help marketers navigate this complex marketing decision process. At the strategic level, a marketer must decide on what channels to use and allocate budgets to each chosen channel. At a tactical level, within each channel a marketer must decide how to carry out campaigns over time. The following diagram depicts the details of this decision hierarchy in which the major marketing channels, their potential interactions and list of common tactical decisions within each channel are presented.


Channel selection is a long-term strategic decision seeking to balance the benefit and cost (both fixed and variable) of operating the channel and relevant business constraints facing a marketer. For example, consider the decision to take away one million dollars from direct mail budget and reallocate it to start online marketing efforts. The following table summarizes the annualized benefit and cost of this decision over a five-year horizon:

Incremental revenue: $5 million per year
Incremental variable cost: $3 million per year
Net return: $2 million per year

The net return excludes the fixed cost for initially setting up the online marketing channel. If that fixed cost, amortized over next five years, is less than $2 million per year, online marketing should be pursued. More complicated channel selection decisions involving more than two channels can be described using an optimization model. Optimal channel selection and budget allocation can be solved jointly incorporating the fixed cost of setting a channel. Alternative solutions can be obtained and evaluated for a variety of operational scenarios.

Given the allocated budget for a selected channel, the tactical decisions at the lower hierarchy have to do with the timing (when and how frequent), placement (where), audience (who) and creative content (what) of campaign efforts. The impact of a campaign is maximized when these decisions are coordinated in such a way so that the right ad is offered to the right customer at the right time.

Whether tactical or strategic, cross-channel budget decisions require benefit and impact estimates of cross-channel campaigns provided through attribution results. In a prior post, existing attribution practices are compared and challenges for developing data-driven attribution rules are presented. It is recommended that multiple rules are applied to establish baseline attribution results so comparison of results can be made among different rules. When it comes to optimizing cross channel decisions, the lack of sound attribution measures can put marketers in a catch-22: optimal cross-channel decisions require their benefit estimates are directly obtained from attribution results, but attribution results can only be obtained if cross-channel decisions are implemented and data-driven attribution method is applied. One way to escape the trap is to conduct experiments in test markets where key decision factors are designed across multiple channels. Experimental results are used to develop attribution results and benefit measures, which in turn are incorporated into improving future marketing experiments or actual campaigns. With several iterations of testing, measuring and updating marketing benefit measures, cross-channel decisions can approach their optimal levels.

As marketers scramble for position and make cross-channel decisions through trial and error, the playing field will be challenging for some time to come. The hierarchical view of the cross-channel marketing decision process serves as a convenient and insightful framework through which cross-channel measurement challenges can be visualized and interdependent marketing decisions can be modeled and optimized.