January 02, 2008

Multi-Channel Advertising Strategy

We don't read a lot about the nuts and bolts of properly executed multi-channel advertising strategies these days, do we?

We know we're supposed to execute 'em. We don't always know what we should do.

Let's look at a multi-channel retailer executing a Valentine's Day multi-channel advertising strategy.

The retailer uses three traditional advertising tools, all geared to drive sales to the call center, the website, and retail stores.
  • An 80 page catalog featuring favorite Valentine's Day merchandise, mailed on or around January 23.
  • A postcard mailing, informing the customer that the sale ends February 14. This postcard is mailed on or around February 4.
  • Two e-mail campaigns, featuring hot items available in all channels. One e-mail campaign is delivered on January 30, another is delivered on February 6.
The multi-channel database marketing team tested eight different contact strategies. The following table illustrates the results of the test, across all channels:

Multi-Channel Advertising Results: Valentine's Day Promotion

Catalog Postcard E-Mail
Ph/Mail Website Store Total Profit

Yes Yes Yes
$5.80 $6.25 $12.60 $24.65 $6.49
Yes Yes No
$6.25 $5.35 $13.05 $24.65 $6.50
Yes No Yes
$5.80 $6.40 $12.25 $24.45 $6.73
Yes No No
$6.15 $5.85 $12.05 $24.05 $6.62
No Yes Yes
$3.45 $5.40 $12.75 $21.60 $6.17
No Yes No
$3.75 $4.90 $12.75 $21.40 $6.12
No No Yes
$3.90 $5.30 $11.30 $20.50 $6.14
No No No
$4.00 $5.00 $11.00 $20.00 $6.00

Catalog Marketing = Yes
$6.00 $5.96 $12.49 $24.45 $6.58
Catalog Marketing = No
$3.78 $5.15 $11.95 $20.88 $6.11
Incremental Benefit
$2.23 $0.81 $0.54 $3.58 $0.47

Postcard Marketing = Yes
$4.81 $5.48 $12.79 $23.08 $6.32
Postcard Marketing = No
$4.96 $5.64 $11.65 $22.25 $6.37
Incremental Benefit
($0.15) ($0.16) $1.14 $0.83 ($0.05)

E-Mail Marketing = Yes
$4.74 $5.84 $12.23 $22.80 $6.38
E-Mail Marketing = No
$5.04 $5.28 $12.21 $22.53 $6.31
Incremental Benefit
($0.30) $0.56 $0.01 $0.27 $0.07

There's a lot of interesting information in the table, don't you think?

First, take a look at the eight test panels. The strategy that drove the most volume included a catalog, a postcard, and potentially the two e-mail campaigns.

However, the most profitable strategy was to mail a catalog and the two e-mail campaigns.

Reviewing results across eight test panels can be a challenge. At the bottom of the table, results are summarized by advertising channel.

Catalog advertising appears to be the most profitable strategy (not always the case), driving an incremental $2.23 to the phone/mail channel, $0.81 online, and $0.54 to stores. The average customer spent $3.58 across all channels, generating $0.47 profit.

Postcard advertising changes customer behavior. Notice that postcard advertising cannibalized phone/mail sales and website sales (not always the case), but worked very well in the retail channel, increasing volume from $22.25 to $23.08 per customer. The average customer spent $0.83 incremental volume, generating a loss of $0.05 per customer.

E-Mail advertising also caused a shift in customer behavior. The two e-mail campaigns cannibalized the phone/mail channel, significantly benefited the online channel, and did drive incremental volume in stores. E-mail marketing (usually the case) drove the least volume, but was much more profitable than postcard marketing --- driving $0.27 volume and $0.07 profit. By the way, your typical e-mail metrics suite (open rate, click-through rate, conversion rate) fail miserably when analyzing multi-channel strategies ... the metrics mentioned above miss the phone/mail cannibalization and miss incremental retail sales.

Multi-channel advertising strategies are challenging to manage. We're told we have to "do everything", that we have to "integrate everything". To understand if the pundits are offering the right strategy for our business, we need to test different strategies.

In this example, we learned that each advertising strategy benefits one channel. Store Managers, for instance, are going to be proponents of postcard marketing, because they will clearly see the traffic increase from postcard marketing in stores. And good luck sitting down with a store manager, telling her that next year you won't execute the postcard strategy because it was unprofitable. Ha! Catalog marketing benefits the phone/mail channel ... without catalog marketing, the phone rang 40% less often.

There are subtleties involved in multi-channel advertising strategies. Catalogs and e-mail campaigns mailed early in the week benefit phone/mail and online channels. If catalog and e-mail campaigns are mailed later in the week, stores are likely to get a little more help.

Also look at what happens to store sales across each test group ... on average, marketing increases sales by maybe five percent, on average. This is a very typical outcome ... and oh, by the way, your matchback vendor might attribute the other 95% of store orders to the three marketing activities, grossly overstating the results of your marketing activities. This happened to me during my days at Nordstrom, a reputable matchback vendor mis-allocated store orders to marketing activities ... one has to be on top of their game to ignore what appear to be brilliant results!

Only well-executed tests and rapid test results help the marketer avoid matchback biases.


  1. Your information may indicate that smaller retailers and mail order operations now need at least 500,000 or so ACTIVE customers to make the most out of sophisticated testing and multichannel direct marketing strategies. The critical mass size continues to rise at an alarming rate.

    In other words, for smaller operations, the profit spread between doing nothing (i.e. the holdout segment) and spending money on advertising becomes insignificant when compared to the staffing and investment requirements for even 100,000 customers. Just putting the money in an investment like government bonds would make more sense.

    No wonder marketing staffs are so lean these days.

    I reviewed your pdf document on "Rapid Test Results." The profits attainable through direct marketing efforts is sizable. But the database in question has 1,000,000 customers.

    Sufficient direct marketing funds may not be available for relatively expensive catalogs and direct mail (especially considering the high cost of small rollout volumes) in smaller companies.

    Fortunately, online marketing comes to the rescue.

    What advice do you have for the smaller retailers and mail order companies that do not have a retail network? Does online marketing become their core channel? In other words, if they cannot make online marketing profitable, are they essentially out of business?

  2. Fortunately, the math doesn't require a small business to have to have that many customers in a test cell.

    I don't go into any detail on mathematical formulas --- there is a provision in significance tests for small housefiles --- if a small business has 2,500 in a test group, and 25,000 housefile customers, then the error rate is discounted by the square root of 10% (2,500 / 25,000).

    So small businesses can get good insights as well, though probably never as good as a large company with a gigantic file. A small business can hold out 2,500 customers, and still understand the impact of direct marketing at a mail/holdout level.

    Small businesses should do what they can, testing one a quarter or twice a year, instead of 20-50 times a year like bigger companies do. Online marketing doesn't have to become the core channel, unless the customer demands that by using online marketing to purchase.


Note: Only a member of this blog may post a comment.

Do You Want To See What You Are Interested In?

The stats below represent data through about Noon Pacific Time on Thursday. Ok, I wrote about the E-Commerce Priority Grid (click here) . Th...