At a time when CEOs are more focused than ever on bottom-line performance marketers are being asked to do more with less. With total responsibility to continually fill the pipeline with new customers, the life-blood of long-term growth and profitability, it is critical for marketers to get it right. Reducing marketing acquisition costs and converting prospects most likely to become highly engaged customers are two of the most important goals they face. What’s a marketer to do?
Well, today’s savvy marketers realize that data is the new black and that it can be an extremely effective tool for customer acquisition. They understand that searching for all households with incomes above $50K is not the best way to optimize customer acquisition marketing dollars. In fact, according to an eMarketer study, “in order to maximize the effectiveness of dollars allocated to acquisition efforts, marketers must place an emphasis on data, audience centricity and the lifetime value of customers.”1
More and more marketers are leveraging their customer data to develop acquisition programs that identify the proverbial “needle in the haystack” (i.e. highly valuable new customers). By using their most valuable asset, customer data, marketers are creating more effective and efficient new customer acquisition programs. “Look-Alike” modeling is one such solution that predicts how closely prospects resemble your “best” customers. The best approach is centered around:
Step 1: Define the customer segment you want to clone. This step is critical to the success because it requires that we clearly define the prospect audience we want to identify and engage, which can prove very tricky. Is it customers with the highest spend? Or those who purchase from a specific product category? Or, perhaps, those in a specific geography or age range (i.e. Millennials, Gen X)?
Step 2: Identify the customer segment to be cloned based upon a key financial metric. After you’ve isolated the customer segment you want to clone, you’ll need to separate the “best” customers from the rest of the pre-defined segment base. This would typically be done by using a key financial metric, like net or gross sales, gross margin or operating profit. Using this financial metric, you’d rank each customer from high to low performance.
Step 3: Append 3rd party demographic, lifestyle and behavioral data. Appending 350+ 3rd party data elements (like, age, household income, education, likely to enjoy hiking or play tennis) to every customer record identified in steps 1 and 2, is the next step to building the final “look-alike” model.
Step 4: Build the “look-alike” model. From among the 350+ appended data elements, between 7 and 12 individual attributes will represent the key attributes in the “best” customer “look-alike”, or cloning model.
Step 5: Score the prospect universe. There are more than 230 million individuals in the U.S. that can be used for new customer prospecting initiatives. To determine which of these prospects are most likely to look like your best customers, you’d apply the model built in step 4 to score each of the prospects. The ones with a score in the top 10% or top 20% are typically selected for your acquisition campaign (depending, of course, upon your budget and marketing acquisition objectives).
Step 6: Suppress your existing customers. After you’ve identified the pool of prospect names from the scoring process (step 5), you should always match your existing customer base against this list and suppress any matching records. The inclusion of any existing customers in a new customer acquisition campaign is a waste of advertising dollars and will skew the results of the “look-alike” model’s performance.
Step 7: Reach out to the “look-alike” target audience. You can utilize a variety of messaging strategies to engage these prospects through any addressable marketing channel: direct mail, email, display, social and/or mobile digital media. Don’t be dissuaded from using digital media to reach and pinpoint your target prospect universe. Digital media can be a very effective reach strategy, depending upon the intended target audience.
Data Axle’s analytics team recently built a look-alike model for a retail client who wanted to increase membership in its loyalty program. Using a randomly selected prospect group to benchmark the results, we determined that the prospects identified from the custom model had a:
Developing a data-driven customer acquisition program has consistently shown to be a significantly more effective approach to funneling new customers through the pipeline than the traditional “scatter gun” methodology used all too often. This strategy not only increases conversion, but we’ve seen time after time that these new customers are also significantly more likely to spend more than the randomly targeted prospects.
Using data, analytics, and “look-alike” modeling methodologies is enabling marketers to gain a competitive advantage by identifying the “right” prospects and engaging them to create more sales, better CPAs and higher marketing ROI.
As the SVP of marketing at Data Axle, Ivy is responsible for multichannel strategy and implementation across all applicable direct marketing channels. She is dedicated to improving the quantity and quality of incoming leads through a combination of engaging content, effective communication strategy, and timely follow-up. Ivy is committed to making Data Axle the undisputed leader in the industry when it comes to data quality, enrichment and management, as well as digital acquisition and retention through technology, services and data.