Lead Scoring is Dead: Two New Innovative Approaches
July 3, 2014
In mid-2012, Aberdeen Group senior research analyst Trip Kucera published a blog post with a headline that, at the time, might have seemed controversial and, to some marketers, blasphemous. It read: Is Lead Scoring BS?
Turns out that Kucera was on to something.
As he pointed out two years ago (and I re-iterated in my post on why traditional lead scoring models are broken), the truth about most lead scoring models is that they’re primarily fiction. At best, Kucera argued, they’re collaborative conjecture — the product of sales and marketing arbitrarily deciding what qualifies as “good” or “sales ready” leads, and seemingly assigning a numerical score to those leads at random.
Frankly, that’s precisely the reason (well, one of five reasons) I think the current state of lead scoring is broken and obsolete.
Existing models rely too much on assumptions, subjective analysis, or incomplete data, which is precisely why SiriusDecisions’ research has found that 40% of B2B salespeople fail to recognize any value from their companies’ lead scoring programs.
There Has to Be a Better Way…
Does that mean we should just ditch lead scoring altogether? Absolutely not. It just means we need to find a better way to do it.
Thankfully, two new categories of lead scoring models have emerged that should pique sales and marketing organizations’ interest.
1) Predictive Scoring
Built around Big Data and algorithmic modeling, predictive lead scoring allows businesses to do a few key things:
- Evaluate everything that was considered by an organization.
- Identify the most successful leads based on positive indicators, like which ones went on to become customers or engaged observers (note: the definition of “success” obviously depends on your demand type).
- Pinpoint failures by examining leads that either didn’t engage at all or didn’t convert into customers (negative indicators).
- Clean and standardize data, and append additional data from outside sources. This is where predictive lead scoring technology shines because it can mine things like job postings, patent filings, social activity, technology use, advertising spend, credit rankings, and expansion news to really flesh out a lead score.
- Develop a comprehensive lead scoring model that combines positive and negative indicators, external data sources, and machine learning algorithms to identify the top predictors of a “good” lead (in other words, subjectivity is totally removed from the equation).
- Validate and test the lead scoring model on a perpetual basis to ensure relevancy.
Why is all of that important?
I think you know the answer. But, to summarize, it solves all of the challenges I talked about in my previous post — namely, the inability of existing lead scoring models to manage increasing volumes of data and those models’ propensity to rely on assumptions.
Top predictive scoring vendors:
2) Account-Based Scoring
Another emerging — and highly effective — lead scoring approach is account-based scoring. This model assigns an “account score” that is based on the cumulative interest of every contact that will play a role in a B2B buying decision.
Why does that matter?
As B2B marketers and sales leaders know, B2B buying decisions aren’t made by individuals, they’re made by groups. As such, it doesn’t make any sense to predict a lead’s likelihood to buy based on the interactions and engagement of just one person. Likelihood to converse or engage? Sure. But buy? Absolutely not.
Getting the Best of Both Worlds
Interestingly, we’re starting to see some organizations combine the biggest benefits of both of the lead scoring models above. The result, as you can imagine, is an even more reliable, systematic, and, yes, predictable way for marketers to objectively evaluate their best opportunities.
Or, as Jay Gaines, Group Director for SiriusDecisions, puts it:
— Jay Gaines, SiriusDecisions
Have experience utilizing either of these approaches? I’d love to hear your thoughts in the comments below.
Image courtesy of JohnED76