Lead Generation via Mathematical Marketing: Thought Leadership with Mark Klein

I was working for IBM, who had purchased my third company, when they initiated a compensation plan for senior management based on the company’s customer satisfaction scores. That wasn’t working well for me, for a couple of reasons. First, there was no granularity of person: good performers were compensated equally with poor performers. Secondly, customer satisfaction is a backwards-looking measure, and gave me no help in managing my small slice of that very big business. I left IBM to work on a better measure, and soon abandoned customer satisfaction in favor of customer loyalty, which was much more predictive. After a bit of work, I developed a model of customer future behavior based on past purchases, and most of what we do in Mathematical Marketing followed along. Earlier in life I spent several years as a theoretical physicist, so I was very comfortable using mathematics to predict the future. It’s been a lot of fun showing companies that they are sitting on a lot of data that can really boost their marketing.

2. You are a frequent contributor to the Longbow Direct Marketing blog, speaking to the value of using mathematics to help B2B marketers improve their campaigns, what are your top three tips in terms of how marketers can better leverage mathematics to help drive a marketing ROI?

You threw me an easy pitch over the middle of the plate. Our mantra around here is Track, Target, and Test. Tip #1 is collect, guard, and use your customer transaction data. It’s the basis of Mathematical Marketing. Tip #2 is to segment and target customers based not on firmographics but on the much more accurate, and more predictive, transaction data. Tip #3 is to constantly test, both offers and targeting. Testing is probably the fastest and easiest way to jumpstart your marketing.

3. What do you think is the biggest opportunity marketers have in terms of more effective lead generation? Specifically, in regards to how the marketing and sales teams work together?

Lead generation, like most existing customer marketing, is still too qualitative. Typically leads are cold, warm, or hot. Too few companies even know the cost of a lead, and if their leads come from multiple channels, for example trade shows and internet, they may not even know which channel is more cost effective. The same mathematical discipline we practice in existing customer marketing should carry over to acquisition. Further, Mathematical Marketing shines in creating lists for the sales team of customers and the products each customer is likely to buy. Sales people love those lists (they deliver fatter commissions), and that love extends to the marketing team that delivered the leads, fostering a good cooperation between the departments. Finally, there is lots of room to model responses to promotion, to carry personas developed in Mathematical Marketing over to customer acquisition.

4. In the B2B market, what do you see as the biggest hurdle to effective lead management?

While some marketers embrace quantitative methodology, a lot more avoid it. A good part of the industry needs an Attitude Adjustment Hour. People unwilling to change are the hurdle to be overcome. Our experience is that this problem fixes itself once some parts of the organization begin to use Mathematical Marketing. Typically their results are such a big improvement over what they produced using their older methodology that others want to jump on the same train, out of fear of being left in the dust.

Wild Card: Anything else you would like to add?

Will someone explain to me why most companies get the overwhelming share of their revenue from existing customers, yet allocate the overwhelming share of their marketing budget to customer acquisition? We’re seeing some shift, and the catch phrase is “retention is the new acquisition.” But marketing spend ratios are still badly out of whack. Today, there are some forces driving a different perspective. First, our shrinking global economy is making many companies look more closely at the revenue from their existing customers. Second, the wider availability of tools that don’t require statisticians on staff is accelerating this trend. I have an optimistic outlook.