Measuring Catalog Success in the Digital Age: Part 2

Part one of this multipart series ran last week. 

Here’s the impact that sending a catalog to a lapsed customer can have on revenue:

Best Customer

• Control spend per customer: $8.00 (blends conversion and order value)
• Test spend per customer: $8.50
• Cost of mailing: $0.50
• Return on ad spend (ROAS) of mail group: 17x
• Incremental revenue: $0.50 (breakeven)

Lapsed Customer

• Control spend per customer: $0.50 (blends conversion and order value)
• Test spend per customer: $1.50
• Cost of mailing: $0.50
• ROAS of mail group: 3x
• Incremental revenue: $1.00 (2x return)

When working to optimize ROAS, there’s no understanding of incrementality; it’s not truly optimizing the marketing budget. It creates a long-term problem where you're only nurturing and growing the best of the best and ignoring customers who can still represent a significant amount of both short- and long-term revenue.

But marketers like the picture it presents: they see an increase in better customers, so they’re tweaking the rules and getting rid of weak customers.

Next Steps: Finding the Balance

Marketers have been using ROAS forever, so you can’t just rip off the Band-Aid. They get $3 million to mail catalogs, attribute it to $15 million in revenue — it’s an explicit metric. Some brands are using control groups, though, looking at lift, projections and more to see how ROAS is related to incrementality. It’s not as easy to understand from a revenue standpoint, but that’s how financial people manage the pie.

Catalogs are often considered old school and ROAS falls squarely into that category. Today’s digital marketers know that you need to trust data, and trust lift and other metrics.

There needs to be a balanced approach to slowly migrate away from ROAS. Out of the gate, marketers still need to mail groups that generate high ROAS. They still need to prove that the money invested in catalogs is returned in sales dollars because they still need to prove themselves to the finance department.

Simultaneously, they can start to use test and control groups so they will have ROAS and an incremental lift number. They can start to understand that relationship and balance those two side by side. Some steps include the following:

1. Begin to test small groups of lower ROAS customers to not completely dilute the existing ROAS metric that people are comfortable with, while identifying opportunities to maximize true incrementality.
2. Pull out a small group of people from the existing, high ROAS mail segments as a control population. Measure ROAS and lift, then compare that to the same metrics in the mailed population.
3. Quantify the incrementality of the test group.
4. Establish long-term holdout groups to define frequency testing and the long-term impact of receiving catalogs beyond strictly the matchback period.
5. Frequency testing — send best customers six catalogs a year, send low-end customers three catalogs a year. There's no need for a catalog every month, just enough to keep the brand top of mind.
6. Find the optimal frequency for the different segments.

Finding the Bottom Line

Within the business, the cataloger points to ROAS success while the digital marketer says it’s archaic. Meanwhile, digital has low ROAS, but high incremental revenue. Both are correct, and both are wrong in some ways. The trick will be to balance components from each and budget accordingly.

Emphasis on data capture is revealing that housefiles are growing at a pace that's not being matched by growth in catalog circulation. Marketers are cutting customers who they know intuitively should get a catalog, but they just can’t send to them because they don’t have enough and want to optimize ROAS. Therefore, they mail to better customers to keep their ROAS numbers high and their budgets coming in.

At the end of the day, businesses need a system that lets them pull test and control, manage tests over time, and access detailed reporting. They need the skill set, a database that supports it, and analysts that can measure and support it all.

Boosted ROAS numbers aren’t cutting it anymore; management wants bottom-line business results. So, the more these marketers can demonstrate incrementality in revenue, the better. Their future may depend on it.

Paul Welsh is vice president of analytics at Customer Portfolios, a marketing technology leader that uses insight and analytics to increase customer value.