Track the True Source of Internet Sales
Growing Internet sales are causing BTB direct marketers to lose even more of their ability to track an order to its true driver. They also must learn to distinguish buyers shopping the Net from those ordering on the Net, a potentially critical distinction that too few DMers make. DMers also must prove whether the common assumption that Web shoppers are the equal of traditional buyers is true.
I define Web shoppers as customers who purchase as a direct consequence of their use of the Internet to identify a source from whom to buy. Use includes Web searches, pop-ups, banner ads, etc. It does not matter whether these orders are placed online, via phone, fax or mail. Web buyers order on a site to which they are driven by an address appearing in a catalog, mail piece, space ad or other traditional medium.
This growing proportion of Web sales poses three challenges to BTB direct marketers:
· A declining ability to track orders to the medium and promotion triggering them. With few customers providing source codes on Web orders, and as Internet volumes increase, response analysis becomes increasingly "blind." This causes DMers to lose their greatest asset: media accountability. Without solid response tracking, DMers can't know which house segments and prospect lists work. And the reliability of creative, offer and other test response analysis is diminished.
· If subsequent response rates to mail and electronic media of Web shoppers and buyers differ, failure to distinguish these groups will lead to increasingly poor circulation decisions.
· The third challenge is to determine the true worth of Web shoppers.
DMers traditionally don't differentiate their treatment of buyers who look "alike," most commonly defined by similar past patterns of recency, frequency and/or dollars spent. Acquisition media have not played a role in this view. Buyers with like characteristics receive the same number and type of contacts. However, evidence suggests that buyers acquired via Web activity may behave differently from, and be worth less than, mail-acquired names. If true, this requires new ways to value and promote these names.
Because of the inevitable growth of Internet sales, direct marketers must act to quantify and overcome these challenges. Techniques exist to improve the ability to track Web-placed orders to their true medium, and even to a specific list or offer.
DMers first must design e-commerce sites to improve code capture.
Request codes on Web order forms. Though most buyers won't enter one, if the Web order form doesn't request it, no one will. Also, tell recipients in mail pieces that using a preprinted customer ID speeds Web or phone orders if customers providing an ID will have name/address fields automatically populated.
Getting an ID on a Web order means the customer was mailed. And requesting a customer ID makes it easier for a buyer to find their source code since these typically print in the same physical area on mail pieces.
Assign prospects a temporary ID. Load these onto your database and retain them for about six months. This allows all respondents, whether first-time or repeat, to be treated the same by TSRs and the Web site.
Have online order forms ask customers how they got to the site, with check-off boxes for common answers.
Other suggestions include:
· Use unique URL extensions to tie orders to a specific catalog, mailing or Internet marketing effort, if not to a specific list. Make the unique identifier part of the URL (Anycompany.123.com rather than Anycompany.com/123).
· Require a code to receive special offers. This is extremely effective for companies that routinely make offers. Use this tool to identify offers made in traditional media where orders are placed online, and for Internet-only offers that customers choose to place by phone.
· Assign different item suffixes/prefixes to SKUs in print media and on Web site product pages.
· Match back orders without source codes to recently mailed house and prospect lists.
· Use software that tracks orders to specific search engines, banner ads, pop-ups, etc., to better identify names driven to a site by an Internet search or other activity.
· Set as a goal tracking 50 percent to 70 percent or more of Web site orders to a specific mail or Web source.
Though these methods help preserve the ability to measure promotion effectiveness and distinguish Web shoppers from buyers, they don't address to what extent behavior differs by a buyer's source channel.
The psychology of business buyers who find your site while searching the Web versus those going to a site to place an order prompted by receipt of a traditional medium may differ. It is critical that DMers quantify and act upon such differences. To do this:
· Create cells for, track and compare the performance of like buyer groups that differ only in the medium that first caused them to purchase.
· Where statistical models are used to segment house names, include first order channel as an independent variable. Determine whether Web source proves significant and negative.
· Test a greater variety of e-mails to Web-sourced names. Stress things unique to your product or service offer and any business characteristic that separates you from competing online vendors.
· Where the customer base is large enough, look for differences in performance by specific Web-driven sources (search engine, banner, pop-up, rented e-mail list).
· Never assume e-mail can replace mail contacts to traditional direct buyers, even those who order on a site, without first testing this extensively.
· To evaluate the effectiveness of e-mail offers or messages sent to a buyer segment, don't look only at sales directly attributable to the e-mail offer, but at total sales from all sources within a relevant time period. Compare these with a like control group not sent e-mails. The smaller the level of incremental sales, the greater the need to maintain traditional contact levels.
Where tests show Internet-sourced names performing substantially less well than traditional buyers, it is important to estimate these buyers' long-term value and to adjust acquisition goals accordingly.
If values are low enough, you may have to pursue Web acquisition options that produce first orders at little cost or profit but limit further spending, especially in the mail. Though this notion may be hard for BTB marketers to accept, they should be willing to measure these differences and determine whether atypical strategies are called for. To implement these suggestions, a customer database must store original and order source.