We licensed one of the best consumer household datasets available, including 14 data points on nearly every postal addresss in the United States.
We load your multi-site customer lists into our platform and match your customer records to consumer household data, expanding your customer records to include fields like income, age, household size, and more.
We take matched customer records for each site, collect all households within a radius of each site, then compute stats on what makes your customers unique.
Visually explore how your customer base differs from local markets across fields like income, age, household size, and others...to see where your product/service is selling.
Think of a customer model as a formula used to predict if any particular household will buy (used for lots of things). It's derived using "machine learning" algorithms, which learn by comparing your customers (households that bought) with other households in the local markets. Once generated, it can take demographic data for any household as input, and output a score predicting the "likelihood" of becoming a customer.
( A T T R A C T )
With the customer model, we can now drop a pin on a map in any store location and generate a list of nearby households most likely to buy (highest scoring according to the model). These households are the most similar to what makes your customer base unique.
Now that you know who is most likely to buy, reach them via direct mail or digital media (custom audiences). "Likelihood" scoring enables you to spend more on the best prospects to establish the brand and make sure your offers are top of mind.
( C O N V E R T )
We'll run through each of your sites, gather the households within a radius, score each of them according to their demographics & distance, and select the top-tier (e.g. if there are 1.1M households within a 3-mi radius of your sites, we'll select, say, the top 20% (220k households).
We work with a 3rd-party provider to anonymize and match the target households to online devices and digital IDs through a process that protects consumer privacy. What this really means...is that you can personalize content and discounts on your website for just these visitors.
When a website visitor falls into a segment, you can extend discounts, in an effort to win the highest value customers.
( M A N A G E )
For any customer where you have increased pricing in the past, we start by importing your data on the customer (name/address), the old price, the new price, and whether or not they continued service.
In the same way we build the "customer model" (above), we match customers with a history of price adjustments to household demographic data. We now compare demographics of customers that continued service with those that discontinued service. This produces a "price sensitivity model" that can take the demographics of any customer as input and predict the likelihood of continuing service with a price adjustment.
We'll look up demographic data on each of your customers (using our licensed household consumer data described above), then apply the price sensitivity model, giving you the necessary insight for making price adjustments.
We identify a region to study, then leverage public records and other 3rd-party sources to capture the supply of competitive sites (each marker's size is driven by the size of the site). Next, we use the demand model to effectively shift the competitive load onto the households near each site (proportionally by the "likelihood" score), so we now know the amount of competition for every household.
We then run the demand model across the region to calculate and visualize three key measures: population density, demand, and unmet demand (adjusted by the competitive "load" for each household).
- Billy Bosworth