Consider the following scenario: Four years after the purchase of his most recent car, a man receives a brochure from a major auto manufacturer. He likes the television commercials for one of the company's models, so he reads the mailer. He passes one of the manufacturer's billboards every day during his drive to work, so he knows he's near a dealership. A quick check of Facebook reveals high praise for the model in question. And his wife, who is expecting their second child, is a huge fan of the company's safety record.
A search engine query gives him the location of the dealership—not the one on his way to work, but one in which the salespeople's profiles give him more confidence about doing business with it. So he drives to the dealership, meets with the sales staff, takes a test drive and ultimately buys the car.
When the auto manufacturer attempts to attribute the sale to a marketing effort, which channel should be given credit?
"Marketers have invested millions of dollars in every channel trying to cover their market and not really understanding the performance of each individual channel," says Rob Stagno, president of customer acquisition firm Paradysz.
Stagno refers to the problem of attributing a sale to an effort as a Gordian knot. But unlike the rope in the fable, untangling this puzzle takes more than the single swift stroke of a sword. A white paper from Paradysz offers a method for breaking the problem of attribution into six steps:
1. Use a matrix to map channels used during various times in the sales cycle.
Paradysz's paper initially instructs marketers to consider which channels work together in moving prospects through the sales funnel of awareness, consideration and purchase. As marketers currently look at their programs, performance metrics "have been developed for each channel in isolation, and don't account for the cross-channel impact." They also term success based on actions appropriate for each channel (clickthroughs in e-mail, for instance) as opposed to the ultimate goal of sales.
Even sales themselves are measured differently, based on the channel. Retail sales are often evaluated on a per-square-foot basis, while catalog sales focus on average order size.
”Rather than looking at silos and the costs [of] each of those silos as a way of calculating marketing ROI, let me understand how demand is created and how demand is fulfilled, so I can both plan my marketing to achieve results across multiple channels and determine the costs of each channel based on the investment I make to bring demand," says Stagno.
2. Determine the customer touch points that have the greatest affinity and interaction.
Customers do not travel along the same sales path. Catalogs are more likely to spur visits to a retail center, while third-party social media sources may result in higher search engine queries. Not having a sense of how channels flow into each other means metrics will be skewed: An online purchase may get credit for a catalog-inspired sale, but the online channel was really just providing fulfillment for a catalog-generated order.
Knowing that catalogs play more of a role as retail traffic builders can also help with expenses. A marketer has the option of pushing larger catalogs to areas outside of areas where it has a retail presence, as these generate orders either online or through contact centers, Stagno says. In areas where catalogs serve to bring people into nearby stores, the mailers can be smaller and less expensive.
3. Consult with product marketing and sales to identify the programs and interactions they are most interested in understanding and improving.
Evaluating the impact of all channel activities at the outset isn't possible. Some nascent channels and strategies, such as mobile efforts or group discount buying schemes or social media presences, require a marketer's participation for visibility purposes.
"Depending on the business, it's better to be there badly than not be there," says Stagno. "It's better to have an app that doesn't do much than not have an app. The risk is not participating in a market experiencing 200% growth because you don't have an app at all."
4. Agree on scope, focus and goals.
The above two points should be considered when determining which channel interactions to study, and what changes to the marketing mix will be made. According to Paradysz's paper, "Product managers and sales will more readily become interested in the outcome of the analysis, anticipate that the results will help drive more effective marketing and agree that the testing will be confined enough to minimize negative short term impact."
5. Perform in-market testing.
Want to measure a catalogs' impact on sales? Hold a representative sample out of a target audience within a store's retail footprint, and compare revenue, with data provided by credit card purchases or store loyalty programs. Alternately, catalogs can direct customers to customized URLS or designated inbound phone numbers.
Marketers can also assign unique product shop keeping units (SKUs) to items in catalogs: When customers order using these SKUs online, or through a contact center, use of the SKUs will indicate whether a catalog was instrumental in generating the order.
6. Produce a quantitative "heat map" which will guide effective allocation of marketing resources.
If a channel such as mail is more effective in generating awareness or consideration (the initial stages of the purchase funnel), multichannel attribution can help set guidelines regarding that channel's share of marketing budget based on the value the sales it ultimately helps generate.
"Attribution helps [marketers] make smarter decisions," says Stagno, adding that with a history of measuring channel influences over time "a very well-worn groove starts to emerge: If I put a catalog in the mail at this time, I can reliably expect to get this result in retail and through my online business. And I should plan for that, and the costs of demand generation."
Network
