 |
|
|
Datamann continues to provide a level of personalized service and personal attention to detail that is hard to come by in
... read more
|
|
|
|
|
|
|
Back to News |
|
Matchback + Contact Strategy = Greater Accountability
Accountability is critical in times of economic uncertainty. As budgets constrict, there is less flexibility and very little room for marketing decision errors. Marketers must do all they can to take the guess work out of where and how they will spend their marketing dollars. Many marketers are finding direction through matchback processing, and some are analyzing promotional history to develop winning multi-channel contact strategies. They are finding that not only the true source of an order is important, but that promotional sequence, timing and channel combination is powerfully connected to buying behavior. They are also finding that diligence is required regarding the necessary processing for both matchbacks and contact database development.
Most marketers today promote across multiple channels, and the true source of an order is often unknown. In fact, for many companies, web orders contribute 70%, or more, of total sales. Often, only 25% to 50% of those orders are entered with a source code from a catalog mailing, email campaign, search marketing, or pay-per-click program. Without a source code, it is difficult for a marketer to understand what channels are driving sales, how many web orders can be attributed to another channel, and what sources are actually working. For this reason, matchback processing has become an important and necessary routine for multichannel marketers. Matchback processing is a backend analysis that helps to determine the true source of all those orders entered under an “unknown” source code. Matchbacks that are robust, accurate and flexible are effective in revealing an underperforming list to be profitable. Elusive, but profitable house file segments can also be revealed through matchback processing.
When matching orders to campaign files, it is important that the matchback process deliver the best possible match rate. Records will not match when critical name and address information is missing, misrepresented or inaccurate. Obviously, the matchback service provider should receive data in a usable format, to avoid “garbage in, garbage out”. Both the order file and the campaign files(s) should be processed through the same address hygiene program to avoid common discrepancies in address corrections, such as representations of directionals, apartment numbers, etc. It is also important to know that match rates can be adjusted and improved upon. The match logic should be flexible enough to accommodate adjustments appropriate for the marketer. Match logic can be tweaked so that anyone in a household, or each individual within a company, can be considered a match.
Most merge/purge or duplicate identification systems use a matchcode (a string of characters and numbers representative of the name and address). In a single merge/purge, several mutually exclusive matchcodes can be processed simultaneously to link records on an order file to records on your campaign file(s). A robust matchcode process will identify matches with subtle differences: blank fields, misspellings, nicknames, name reversals (last/first, first/last). The algorithm can be tightened or loosened so that a preferred degree of “near match” can be identified. The idea is to achieve optimal matching without a high percentage of false positives.
Matchback rules for allocating source codes are specific to the marketer, and can range from the simple to quite complex. Complexity usually coincides with the number of overlapping or simultaneous promotions. The matchback process used should allow for incorporation of unique and specific rules. Allocation is not scientific, and the rules need to be flexible. The art in the process is how you develop the rules to reflect the changes in your business and in the marketplace. Allocating back to the most recent promotion prior to the order date is the simplest and most common rule. Basic overrides to the rule would include web promotions, web coupons. Multiple, simultaneous promotions require matching orders to all mailing/source code/date combinations targeted to the ordering customer. In these cases, the order date is checked against all matched mailing/source code/date combinations, and the order is allocated to the mailing source code as per the specific rules. Specific rules may include:
• Allocating proportionally to multiple source codes
• Allocating proportionally based on conditions, such as: combination of books mailed, length of time books have been in the home, multiple titles with crossover with a SKU response that can be shared
• Allocating based on the percentage of sourced orders for a specific date. For example, if 20% of sourced orders for a particular date were from the catalog, allocate 20% of the unsourced orders to the catalog. If retail orders accounted for 10% of orders for that date, allocate 10% of unsourced orders to retail.
• Allocating based on SKU. If items appear in specific catalogs, allocate to the catalog the items appeared in.
How often a Matchback is performed depends on planning timeframes. When creating campaign calendars, consider enough time to review and analyze results of your matchback results before putting the next campaign into motion. Retailers may want to perform matchbacks more frequently to track the retail buyers mail/web responsiveness.
Post Matchback analysis provides a more accurate picture of performance, and can be useful when weighing re-use and rollout options for future promotions. Matchback reporting can be tailored to suit current business analysis. At a minimum, matchback reports should display the number of orders, dollar amount of merchandise, average order value, and sales per catalog mailed for each segment - both before and after matchback processing.
Effective matchback processing provides considerable accountability for marketers. Development of a contact strategy takes accountability a step further. In fact, contact strategy may be the most powerful driver of response, next to RFM. How valuable would it be to know how many times a customer/prospect should be promoted to achieve a sale? How many times do you promote before you cry “uncle”? What channels, in what order, in what time frame is the optimal marketing strategy? What is the incremental gain from overlapping, cross-channel promotions? Clarifying the complex interactions of channels and promotions requires a database of all channel promotions, along with order transactions and post demand transactions, such as returns and exchanges. A contact database provides the analytical framework necessary for predictive modeling and customer views to measure performance over time, and re-creating “point in time” scenarios.
Unlike a prospect database, a contact database does not require negotiation with list owners for multiple use of names. Contact databases are used to look at the performance of individuals and households that have already been promoted and paid for, overtime and across channels, at a matchcode level. The resulting analysis will help you make intelligent decisions regarding timing, channel and sequence of future promotions. Use of a contact database gives marketers the ability to target those likely to respond and avoid promoting those unlikely to respond. While this is certainly not a new concept, changes in marketing strategies and advances in technology allow marketers today to apply contact strategy in a timelier, and less costly, manner.
Whether performing matchbacks or examining promotional history, you are extracting intelligence from the data you own, and working towards improving profits. Today, marketing can seem like a pressure cooker environment, and success is defined by the bottom line. Accountability depends on paying attention to what matters: all relevant information to help you to make responsible, measurable decisions. Is it worth a test?
|
|
