For our series on 1-to-1 marketing best practices, we reached out to Alex Aigner, COO at DataLab Digital Advertising, to learn more about how they leverage custom predictive models to boost new customer acquisition efforts.
Can you give a brief overview of DataLab?
We’re an addressable marketing consultancy that specializes in improving the data and targeting behind direct marketing programs.
Data sourcing, processing, analyzing, and data warehousing are our core focus.
We believe that using the best data with the most sophisticated analysis is the key to targeting better results.
What are custom predictive models?
Traditionally, predictive models have been “prepackaged” and offer a way to identify segments of consumers with certain propensities to engage, purchase, or consume products or services.
While most of these prepackaged models rely heavily on third-party demographic data, custom modeling looks at the whole picture and identifies the optimal set of model parameters, variables, and interactions.
In other words, custom predictive models (like DataLab’s!) use both first- and third-party data to help companies and organizations identify the best audience to focus their direct marketing efforts on.
We’ve found this method to generate 10-20% lift over any prepackaged modeled approach.
Examples of data we use includes current customer data, prior campaign history, demographic, credit and other third-party data.
The models are part of our larger addressable marketing approach.
What is addressable marketing and why is it important?
Addressable marketing is a form of direct marketing that focuses on reaching the best audience via cross-channel marketing efforts.
It’s important because it’s the direction most marketing is heading, and by focusing on the extremely desirable prospects, you’re effectively eliminating wasted marketing dollars.
Naturally, when you remove waste and focus on the most desirable audience, your cost per acquisition will decrease and your marketing ROI will increase.
What has working with LiveRamp allowed you to accomplish?
One of the top three insurance companies invited DataLab to compete for display ad dollars.
Its digital media group had worked with several media partners to increase overall quote volume while maintaining a strict cost per quote (CPQ) goal.
DataLab launched multiple online display campaigns that leveraged our unique data assets and predictive modeling capabilities to identify the top tier prospects in the U.S.. These were then onboarded to a DSP via LiveRamp.
DataLab drove in over 36,000 quotes and CPQ goals were beaten by an average of 12% over the course of the campaign. Advertising investment more than doubled within one year.
Prospects identified in the predictive models were not only more likely to quote but also more likely to actually buy a policy and stay a customer longer.
What advice would you give to digital marketers just getting started with data-driven marketing?
There are really three pieces of advice I’d give to anyone new to the ecosystem:
- Seek transparency—transparency is one of the biggest pain points in our industry. Seek and work with companies who are willing to share the most relevant information with you.
- Combat fraud—there are many tactics you can use to do this. Onboarding audiences is one way of significantly reducing fraud, since you’re only targeting known consumers. Blacklist known fraudulent domains and publishers. Avoid targeting old versions of web browsers. [[Editor’s note: for more on ad fraud and how to stop it, watch our webinar with Dr. Augustine Fou of Data Science Consulting]]
- Test custom models—If you haven’t already tested them, we highly recommend exploring custom models. Your first-party data is going to be one of the most powerful things you can utilize for optimization. In the long run, you’ll benefit so much more from a custom model.
For more on how to employ 1-to-1 marketing best practices, check out our Definitive Guide.