How’s your keycode capture rate these days? When every order counts, you need to know where they’re coming from.
Matchbacks can give you better visibility into the responsiveness of a campaign. This enables you to evaluate which efforts have the highest return on investment and then reallocate future advertising dollars for the highest productivity.
But doing matchbacks can be tricky because of the elusive keycode capture. It’s not unusual for the keycode capture rate to be less than 50%, and some companies experience single-digit percentages for their keycode capture rates. Determining the best allocation of advertising dollars is not so simple.
How customers respond and when they respond depends on a much more complex series of marketing communications; it’s now a contact strategy. From catalogs and e-mails, to direct mail, telemarketing and branding, it’s all about the advertising spend, return on investment, break-even, payback period and profitability.
With the advancement and accessibility of technology combined with the opportunity for customers to decide how, when and where they spend their dollars, it’s hard to clearly evaluate which campaign works best to drive sales.
The matchback analysis is one way to increase the clarity. This entails taking the order file — all of the customers who ordered during a certain timeframe — and matching each address against the address on the mail file to see if the customer who ordered received the mailing.
Because the mail file has the keycodes on it, when a match is identified, the order file is appended with the keycode (usually replacing the default keycode on the order file). From the results of the matchback, you can determine if the campaign was profitable, which lists performed well, and how customer segments responded.
If you conclude that the campaign drove sales at an acceptable profitability, you should continue it and perhaps increase it. If not, you can put the advertising dollars toward other, more productive efforts.
Often a typical matchback process uses a cascading evaluation based on the mail dates. For example, if examining the spring season, you would extract all orders that happened during the timeframe and give the information to the data service provider. Then your data service provider will do several data preparation steps (NCOA, address standardization) for both the order file and the mail files to ensure the addresses on the files are synchronized.
At that point, the analytical gymnastics begin as the mathematical routines use the mailed record to search for a corresponding order. If a corresponding order is found, the match is identified and the marketer now has visibility into the performance of the segments. This matching process is done using the parameters of the mail date.
You would definitely have a fairy tale ending if all orders could be matched against the mail file. The reality is 20%, or as high as 50%, of the order records cannot be matched — even with catalogers who do not have e-commerce.
Obstacles and work-arounds
What happens to the sales allocation of these unmatched records? One option is to distribute unmatched sales across the mailed segments based on a weighted allocation.
For instance, once the matching process is completed, if List A sales are 10% of the total sales, then 10% of the unmatched sales will be allocated to List A. This method assigns all unmatched dollars to the segments.
A bigger issue comes up as the holiday season closes — what to do with the landslide of orders that happen just before Christmas, even though the mail quantity was the smallest? There are different ways to review this.
One way is to look at the contact strategy for the season. Identify customer groups and then identify how many contacts were received. Calculate the cost of the contacts and compare it to the revenue generated by the segment as a whole.
Complete the financial analysis to evaluate the success of the segment by determining the advertising ratio. If you look historically at a successful campaign, you may find that your advertising ratio is 19% (total advertising cost divided by gross demand sales). You can calculate the advertising ratio for the season and draw conclusions based on performance.
If a segment of 5,000 was contacted five times at a unit cost of $0.57, the advertising cost for this segment was (5,000 × 5 × $0.57 = $14,250). If the historically acceptable advertising ratio is 19%, this segment must generate $75,000 ($14,250 ÷ 19%) to be successful.
Business-to-business merchants can compare both the mail file and the order record for company name. The mail file often goes to the headquarters, and the order record will be placed by the user in a satellite location. The opposite also holds true when the mailing is addressed to the end-user, but the order is placed by the purchasing department at the main office.
But you have to identify business rules before the matchback processing or you may inadvertently inflate sales, and the unmatched dollars will be allocated erroneously. Here’s why: The order entry system captures all orders during the time period. Perhaps there was a large order placed that fell outside the scope of the mailings, such as a business gift order. You will want to omit those from processing.
What’s more, you may have certain customers who consistently purchase because you have an account representative calling on them, or a customers who only buy gift cards at the start of the season for their employee event — these types of transactions do not occur because of the catalog, e-mail or other campaigns. You will need to deduct these order records from the matchback.
There’s no standard way to perform the matchback analysis. There are basic elements, but your company’s unique business rules will guide your strategies to the matchback.
Gina Valentino (firstname.lastname@example.org) is president of consultancy Hemisphere Marketing.