Marketing ROI 101: Attribution, measurement aren't the same
By Thomas Noyes, CEOand founder of Commerce Signals
Successful restaurant marketing campaigns have many substructures. They are rooted in clear strategic choices that include meaningful consumer insights, differentiated brand positioning, great menu offerings, motivating promotions and more. A strong campaign can improve brand equity, but ultimately it must drive incremental traffic and sales to impact business results. Sustainable marketing-driven growth relies on finding the mix of tactics that drive more incremental contribution than they cost to execute.
Welcome to Marketing ROI.
For most marketers, improving marketing ROI is an evergreen objective. The numerator of the equation — the return or incremental sales, is where different methodologies are common. The process to determine the numerator is often called either "attribution" or "measurement." While those terms are often used interchangeably, they are not the same. A good marketer will recognize the difference and the actions that he or she takes will reliably improve results.
Perhaps one reason why marketers use measurement and attribution as synonyms is that the literal definitions are not particularly helpful. Certainly, Noah Webster was not thinking about marketing at the time when he wrote them. The most relevant Merriam-Webster definitions are:
Measurement: the extent, size, capacity, or amount of something. An example is: "The room's measurements are 20 feet by 14 feet."
Attribution: To explain the cause of something. An example is: "We attribute their success to hard work."
In essence, measurement is using some tool to determine the exact value of something; attribution assigns a value to something.
Measuring = counting
Attributing = assigning credit
These simplified definitions hold up when considering the methodologies used in marketing measurement and marketing attribution. Measurement studies, provided by leading companies, use essentially the same methodology as pharmaceutical trials: they count the incremental impact by calculating the difference between a control group and exposed group. It is not out of the norm that within 72 hours a knowledgeable marketing company can measure the impact marketing has had for a client, whether it be in-store and online sales. Attribution comes into play when credits are assigned for sales through the use of either predetermined rules (e.g. last click, first click, fractional weighted) or forecasting methods (e.g. regression, random forest).
Attribution is the most frequently used term. Google Keyword Planner reports nearly four times the monthly search volume for the SEO keywords "marketing attribution" compared to "marketing measurement." Given the desire to drive awareness and search traffic, many marketing measurement companies use the word "attribution" on their sites. However, the reverse is not true. Marketing attribution companies do not generally mention the word "measurement."
Attribution and measurement are distinctly different, yet each has their place in marketing analytics. The point is to recognize the difference and to use the right tool for the job. Forecasting is best for estimating future outcomes (e.g. "We expect to sell 112 units of SKU XYZ in July"). Counting is best for measuring outcomes that have already occurred (e.g. "We sold 98 units of SKU XYZ last month").
Not every outcome in marketing can be measured like the amount of snow that fell in New York City in March. If you can't define a control group for historical activity, the use of forecasting techniques to assign fractional values may the best method. As you plan your marketing going forward, you can accurately measure its impact by planning diligently and using the right tools.