Contributed by Cody Young, ReachForce Consultant

I came across an interesting post by Jeff Liebl on Performance Insider blog. The post proposed the notion of have a third-party scoring system for leads similar to FICO scoring. The potential pitfalls expressed in Jeff’s write-up about buying leads from online sources are well founded. And while conversation about scoring lead data in a way similar to FICO is interesting, the real value I see is the more practical and tactical thoughts he provokes about lead data quality in general.

First and foremost, the concept of establishing FICO-like rules for scoring individual lead quality while lists are being bought and sold shouldn’t be an edict for marketers to sit and wait for. Market forces are already making it happen in the B2B space (at companies like ReachForce) – and yes, it is having real effect on price models and competition between lead source vendors.

The most important element of any marketing effort is specifying the target – and you can’t really do that effectively by just ‘buying a list.’ Today’s marketing must be managed by understanding how sales funnels work and how buyers buy.

This puts a very special responsibility in the hands of all marketers to view this as filling up on ‘funnel fuel’ –not ‘list buying.’ Sadly, Jeff’s spot-on reference to “numerous reports of fraudulent and bad-quality leads” is a disturbing indicator that too many are still in line with the wasteful standard that marketing’s job is to buy lists and busy themselves sending out emails and letters for 2% response rates.

Buying data to feed sales funnels can best be compared to buying fuel – and from sludgy-crude to jet fuel, a wide range of grades exist. That being said, it should not surprise anyone that as new sales and marketing automation engines expose better ways to find, keep and grow customers, jet fuel is going to win the race every time.

Once high quality funnel fuel is secured, predictive modeling does not have to be a complex, budget busting ordeal. A simple way to start is by ranking a single-value to measure each prospect in a simple, but highly relevant way. These are what Dr. Eric Siegel calls ‘predictors’ in his short but informative article entitled Predictive Analytics with Data Mining: How It Works.

He draws a simple example using “recency” as a predictor (based on how long it’s been since a customer’s last purchase) and assigns higher point values for more recent customers. This simple analysis drives a very obvious prediction (you can probably guess) that contacting the customers in order of recency – calling the high scores first and the low score last – will result in better response rates.

Expanding on this, other predictive indicators and rules can be introduced to formulate smarter and smarter models as you go. The next example is to combine two predictors into a formula: recency + personal income. And if one of the predictors is more important to you than the other (by rule) then its weight is adjusted accordingly – e.g. 2 x recency + personal income.

Once you are able to score your database this way (or just parts of it for starters) using predictors that best fit your needs, the ability to target and finesse top scoring leads with highly relevant and personalized communication is enhanced – thus, increasing the probability that you can drive a prospect’s behavior and not just predict it (as with FICO scores).

A lot of the better marketers I know like this approach because it is not that complex and is the least costly, most deliberate way to drive sales revenue. After all, it really just keys on another formula that comes to mind: 2 x quality funnel fuel + targeted and personalized communications = high response.

Share on Facebook