Mention predictive modeling or customer profiling to a small business launching a business-to-business direct marketing campaign and the reaction may well be: “Fuhgettaboutit!”
Predictive modeling and profiling sound complicated. They sound like a luxury: nice, but not necessary. Above all, they sound expensive.
Not long ago, all that was true. Only the giants of direct marketing, such as large banks or insurance companies, could justify designing algorithm-driven models to zero in on highly responsive prospects sharing attributes of their best customers. Only the big players could design models to profile their worst customers to weed out similar prospects accordingly.
Now a new breed of online tools lets small to midsize businesses harness the power of customer profiling and prospect modeling, accomplishing with hundreds of dollars sophisticated tasks previously available only to businesses willing to shell out $20,000 or more.
One such tool is BizInsight Business Profiling, available at www.experianbizinsight.com/profile. It lets businesses create customized BTB direct marketing lists of prospects that resemble their best customers who are most likely to make a purchase. Or it can identify undesirable prospects in order to exclude them from campaigns.
Small to midsize businesses know who their customers are. But all too often, they don't know much about them. Show me a BTB company that fails to ascertain its top-performing ZIP code locations, what percentage of customers are high credit-risk borrowers, the annual sales and number of employees of typical customers or even what industry most of their customers are in — and I'll show you a company that is flying blind.
Closing the knowledge gap can mean new opportunities. If you own a small office supply shop that caters to other businesses, it might help to know that your average customer drives 25 miles to reach your location. You might conclude that you'd be better off adding a new location closer to customers.
If you find that construction companies of a particular size and credit score tend to like your company's product, it may make sense to market specifically to construction companies cutting that profile. They're more likely to respond than just soliciting every company within 100 miles.
As we'll see, targeting the best potential customers means a more cost-effective marketing campaign. And of course, the goal of reaching out to likely customers is creating new revenue streams.
Perhaps most significant to the time-starved small business operator is this: Today's modeling tools can be accessed and accomplished online.
The raw material for any profiling and modeling tool is your own list of customers — or a subset of your best (or worst) existing customers. Start by uploading your list of business names and addresses. Once loaded, a good tool will draw from existing databases with profiles on millions of businesses from all industries nationwide.
Look-alike prospects can be scored using regression analysis to give specific weightings to different characteristics, such as physical location, industry codes, employees, annual sales and credit ratings.
If the tool came up with, say, 10,000 prospects but your direct marketing campaign called for only 5,000, you should be able to pick the top 5,000 with the highest likelihood of a response.
Let's take an example to see how a targeted campaign stretches your marketing dollar. Say you conduct a random mailing to 50,000 businesses with a 1 percent response rate. If each piece costs 42 cents, the campaign will run $21,000 and produce 500 responders. That's a cost of $42 per response.
Now let's say we spend $1,000 for the profiling tool and still mail to 50,000 businesses. Since we're sending the piece to those identified as most likely to respond, we get a 3 percent response rate for 1,500 responders. Even including the profiling cost, the cost per responder now is one-third that of the random blast-campaign: $14.67 per responder.
The same tool can cut the overall cost of the campaign by letting you mail to fewer businesses while achieving the same number of responses. A targeted 3 percent response rate would require only 16,800 solicitations to produce 504 responders. Fewer pieces in the mail would bring campaign costs way down, to roughly $7,000 at a rate of $14 per responder.
There's no question that predictive modeling and customer profiling can help businesses find their marketable universe, target the right customers and lift response rates. That's why banks, insurance companies and other big players have used modeling and profiling in direct marketing campaigns for years. Now those tools are within reach to small BTB marketers. They can be applied on a small scale, right off the Internet and at an affordable price. The only question for a small BTB marketer now is, why not use them?
