My favourite words from the sales management after a campaign executes is “Do it again.”

It is always reassuring to hear that all of the work involved in producing a well-designed acquisition campaign brought in the results expected. The modeling worked, the selections worked, the size of the list was right, the data was well paired with the creative and offers presented to the recipients. Success!

And, my most feared words from sales management? “Do it again.”

What? Can the same statement be both loved and feared? Yes.

From the data perspective, it is impossible to “do it again.” The data has changed from the initial campaign run 60 days ago to today’s running. Those addresses that made the first run a success are now customers so they aren’t available in the next pass. A literal running “again” just picks all of the non-responders from the first pass.

Sure, some will respond, but nowhere near the first attempt. So, doing it again won’t make anyone happy.

The data modeling has to be refreshed to reflect the new reality in the database.

Here is a different way to look at the request: Assuming that price/offer/creative all remain the same, the “ask” of the data is to find a way to shift demand (quantity) and establish a new supply/demand equilibrium.

Unless some qualifying records were withheld from the initial campaign, the second time through the campaign is going to have problems – the responsive records have already been withdrawn from the data leaving only the non-responders.

Studies show that sending the same list does have a response, but not like the first pass. When the second pass response rates are calculated, as a data person, I hear an even more feared set of words: “What went wrong?”

The data trick is to mature in your analytics to a point where you can continually update acquisition models to keep up with the changes in the data environment – to use all of the data and go beyond replacing attrition and find true new growth. This is where use of both “lots of data” and true “Big Data” come into play.

You also have to face the reality that not everyone wants what you are selling, as addressed in my previous post on campaign optimisation.

Using all the data is becoming quite the art form now as we in the media start to balance our traditional space with lots of data with that of the digital environment with its Big Data quantities and velocity.

The analytical trick is to come to grips with what you are analysing. Are you analysing enough and the right data points to react each time to select [the right] additional response-likely prospect records from your database?

So, fellow data people, how do you find the additional responders when faced with knowing that your list, coming from the same pool of addresses, needs to support a mailing of the exact same piece (offer and creative)?

You could create a list of known churners to get to a response rate. But you know (or better know!) that the retention rate will be poor, so doing a churner list isn’t an option worth pursuing. You need to do something different. How?

Data. It is the key. Picture the circus performer with six plates spinning on rods at the same time. This is a metaphor for the data balancing act that is required to maintain a sustainable marketing effort, let alone think of moving into a growth pattern.

The data balancing of many plates, including:

• The list targets.
  
  
• The frequency of touch.
  
  
• The offer.
  
  
• The creative.
  
  
• The contact method (e-mail, direct mail, telemarketing, social, etc.).
  
  
• Product content.

Each of the balanced components requires experimentation to determine what is appropriate. For example, should you direct mail a 25-year-old Millennial six months in a row? Only once? How about never, but send him an e-mail – and send it six months in a row? Does he get a seven-day print offer or a digital + Sunday? And so on.

Focusing in on just the data elements, how should you select the list? Age, income, home ownership, and so on are the usual suspects, and since everything is changing all of the time, you need to make sure the data you look at is also fresh.

Do you track the price of a gallon of gasoline? Here is where you start to move from having lots of data available to choose from to that of Big Data, and I mean true Big Data! Big Data brings additional choices to add to your decision model. Some examples that will correlate to sales and list criteria are:

• Online story selections.
  
  
• Web page entry and exit pathways.
  
  
• Device(s) to person tracking.
  
  
• Social media trending.
  
  
• Consumer spending confidence levels.
  
  
• E-purchases.
  
  
• Weather patterns (single-copy and day-pass sales forecasting).
  
  
• Number of homes on the local MLS board, and so on.

OK, some of these are in the category of “lots of data” or can be best handled by map reducing to a manageable size. But some are true Big Data analytic elements.

Are your lists accounting for any of them? Face it; these elements are changing rapidly while your address file and licensed demographic files are almost static. Growth requires finding what changed, how it changed, and how that change impacts your potential consumers as they relate to your product usage.

How you and your marketing database use these, and other, data elements is critical to drive growth. Data-driven decisions are improved through the use of modeling, analytical, and data mining technologies designed to prioritise selections yet simplify how to do it – like segmentation (group into like behaviours) models to classify groups of the consumers.

But over the long run, the segmentation runs along its course and more data is needed to move from the low-hanging fruit to the hidden gems. It gets complex and so very fast.

Use of multiple data elements in the analytics and model building become the master controls you can use in the data to manage how the spinning plates of data are linked together to select the best possible campaign files each time.

But, you have to re-balance these data plates every time. Yes, every time.

Why? Well, sustaining growth, with an ever decreasing price point is not a sustainable long-term option. And asking the same people month after month the same question is not going to get you very far.

Even these death marches become difficult quickly as each month’s remaining pool of prospects are playing a reverse auction game with you that only reaches the endpoint when the price hits zero.

Build sustainability and frequent updating into the data process from the very beginning. Build a learning process into the campaign design. Document what you’ve tried and note results.

Campaigns are not a set-it, forget-it project. They require constant updates to account for every change to the customer base and general market. Start collecting everything possible. Ingest into the model. The model will speak to what is relevant – for that campaign run!

And be prepared to do it again with the next campaign because everything changed, and you need to account for everything that changed to understand how to move forward. Make “do it again” an adventure in data-land.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .