Suppose you have your entire distribution area postal CDS certified (every address). Suppose you have connected your subscriber status to this database. Suppose you have appended to the database demographic information at the individual person level within each address. 

Suppose you have map reduced and connected your online traffic to the database. Suppose you have a business (companies in your market) file.

What do you have? You have the best database in your market! No single business in your market — not even the digital giants — knows as much as you do about the current state of your market’s geography and the happenings within the homes of your market.

Well-harvested Big Data can bridge the gap between advertisers and potential customers.
Well-harvested Big Data can bridge the gap between advertisers and potential customers.

How about leveraging the database beyond your own use and use it as a revenue-generating tool?

Several of the large media companies are doing this and a few of the smaller companies are doing this as well.

So, how do you start if you want to get into the business? How do you get better at it if you are already in the game? Today we’ll take a short journey through the business-to-business-to-consumers (B2B2C) world.

First, what is the B2B2C world?

It is simply two businesses (the media company and the advertiser) working together to find the advertiser additional customers — the C. This sometimes involves an exchange of data from the advertiser to the media company. Sometimes not.

Don’t get hung up on needing to get a file from the advertiser. It is nice, but many companies aren’t willing to share their customer data, others are prohibited (such as in the case of HIPAA), and others don’t have a file to share even if they wanted.

Let us start with the companies that won’t share their data — for whatever the reason. You can interview them and ask about what they feel is their retail trade area. The store could be a mom-and-pop shop that feels it draws from a five-mile radius, or maybe it thinks it attracts within a ZIP code or two. Or, possibly it feels it has the market covered for new parents.

In your interview work, you need to get some sort of way to define the geography of its trade area. Be careful when representatives talk in terms of X-mile radius. Most will overestimate that to the point where, when you convert to a diameter, their guess turns out to be covering 80% of the city.

Once the geography is defined, all of the data analytical software have built-in tools to query/select records. Here is an example from BlueVenn’s Blue Analyzer product to pick a three-mile radius around a latitude/longitude:

Table( CEILING(POWER(SQUARE(69.1*(decimal( [Cons_Demo.LATITUDE]) - 39.609978)) +SQUARE(53*(decimal([Cons_Demo.LONGITUDE]) - -86.178535)), 0.5)) <= 3, [Addr_Master] )

With the records selected, the analysis can begin. The best place to start is with a market-to-customer comparison analysis. With this, just as you would when analysing your own customers, you are working to give the advertiser a view of his customer file. I like to give the basic demographics: age, income, children, education, homeowner status, and occupation.

Additionally, I like to show the business a map of its actual trade area. I think it is fair to say that the “two-mile radius” view of a trade area is a myth, and trade areas are really an odd shape with islands popping up all over the city.

From the analysis work, a story will unfold about the customers. Here is where doing a B2B2C analysis differs from work you would do for your own operation.

The key difference is that the advertiser will not be anywhere near as familiar with the data and use of demography as your own staff. He also comes with a pre-conceived view of what the customer base is, and that is likely to be different then what the data shows.

An example is within some analysis I did for a nursery store. The owner thought his customers were low- to mid-income people because they showed up in older cars and trucks, bought a trunk or pickup-bed of plants, and drove away.

When the analysis was done, the customer data showed very high income. He argued the data, then at some point had an aha moment: People aren’t driving their Mercedes in to fill the trunk with plants/shrubs; they take their older (or kids’ cars).

Anyway, he was happy then, let us run the ultra-targeted campaign, and had his best performing sales week ever.

Here is a snippet from a profile. In this clip you can see an indexing push to the high income levels, but note that from a quantity perspective, the customer sweet spot is in the mid- to lower-income ranges. (This is not the nursery owner’ data, by the way.)

Analysing detailed customer profiles sheds light on target markets.
Analysing detailed customer profiles sheds light on target markets.

You have to be careful in your recommendations when you see this type of split (index to volume). When you do, there is a trade-off between volume and response rate that you have to carefully convey to the client: “I can get you response from the campaign, or I can get you customer volume.”

The balancing act is a fun discussion. Again, you will be dealing with a customer’s set of data and not your own, so you will have to over explain to keep expectations in line with what you will be delivering.

Another piece to the B2B2C discussion is with data confidentiality. You will want to put into place special agreements for data sharing, confidentiality, removal of records, and, in some cases, HIPAA agreements and compliance certification.

When you do get files, try to have the advertiser send the absolute minimum information. Just get the address and a “masked” version of any special analysis data.

For example, if doing analysis based on a customer’s purchase level (gold, silver, etc.), have the customer recode the value to something that only they know (instead of gold, silver, and bronze, have the advertiser use one, three, and seven — but only he should know what number represents each colour).