Throw Back Thursday: Modern Marketing Tips on Tried and True Principles
Multi-channel marketing gets more confusing every day. New channels, new segments, new buzz-wordy ways to describe things we’ve been doing for years…it can be tough to keep up. Let’s use this Throwback Thursday as a chance to dust off some old school testing principles we all learned when the world was simpler…we think you’ll agree with us that the fundamentals of direct marketing apply across all media—no matter how new and shiny the channel might be!
When selecting new segments – exercise caution when applying attributes
- We’ve all been excited about the performance of a test, only to roll it out and see performance drop. Often these situations are self-inflicted. Let’s suppose we want to select data from a source that has a total universe of 60,000 records. With all the best intentions, we often load up the filters and attributes on that original 60k universe and get down to a more usable and likely to respond segment of 10k.
- What we don’t know is that even if this 10k name test succeeds, all that heavy selection has set us up for disappointment when we want to roll out to larger quantities of that data source. We originally sacrificed scale for efficiency yet created a RR% that won’t sustain on larger quantities.
- Marketers need to practice patience to grow (and trim) universe carefully. Creating larger segments and then ‘Nthing’ ensures more consistent performance at roll-out.
When selecting new universe via modeled segments, consider selecting a “Floater Segment”
- Modeled prospect universes are an important part of any marketer’s acquisition or retention plan. Testing a new model (or selecting deeper into lower segments of proven models) can also be frustrating: Decile 1 is better than 2, which is better than 3…the deeper you go the lower the RR%–it can be a scale limiter.
- The idea of selecting a Floater Segment uses this step-down responsiveness as an asset rather than a liability. Suppose you want to expand deeper into a modeled universe where Decile 1 and 2 have already proven themselves. It would make sense to add Decile 3 to the next segment, and then if that is successful, eventually add Decile 4, and so on. Marketers looking to shortcut that slog will often select a small (but statistically valid) random Nth of a lower decile—a Floater—and key it separately.
- So, our hypothetical marketer takes Deciles 1 and 2 as continuation segments, Decile 3 as a test segment and a random Nth of Decile 6 as a Floater. The trick is, when reading the results of Decile 6 you can infer what the results of Decile 4 and 5 would have been—assuming your prospect vendor confirms the model had a nice, step-down performance in back-testing. This is a useful shortcut that marketers employ when looking to expand universe, fast!
When testing new universe always stack the deck in favor of the test segment
- The only thing worse than a failed segment test or data source test is when said failure may not be the fault of the new source at all.
- Thoughtful Marketers take a pass on testing new universe in times of the year that are slow for their business, or with an offer or piece-type that is unproven.
- The fact is, most test segments fail—we can’t risk a false negative by introducing variables that are less-than-ideal for our new test—it will just lead to second-guessing our results—and no one gets second-guessed more than the circulation department (OK, maybe the Search teams have a claim to that title!)
- The fact is, quality circulation planning is about slow, methodical, discipline procedures. So yes, testing that new universe can wait two months until the next busy season! And no, we can’t test new universe AND a new form factor at the same time! No shortcuts!
So, what do you think? Let us know if these tried-but-true acquisition principles apply to your multi-channel mix today—or feel free to enlighten us on other oldie-but-goodies!
Senior Vice President, BI + Analytics
Database Marketing Audience Targeting
MeritDirect’s Business Intelligence and Analytics team helps you with assessment, application of best practices and maximization of ROI across all channels. Our team focuses on “action-oriented” reporting and analytics that provide specific recommendations on how to move forward in the intermediate, mid-term and long-range market time frames. Our in-house team has one core organizing principle: promoting + enabling data driven decision making across our client base.
Can we help you? Contact us for a free consultation!