Finding B2B “Look-Alikes” in Latin America
One of the uses of any B2B database is to find prospective customers that look like current customers (aka “look-alikes”). With more developed and complete data, like what exists for the US, this can often be done using statistical predictive models. These models can yield “prospect scores” appended to individual business locations as to their likelihood of matching the characteristics of a current customer.
In regions where data are less complete, like Latin America, a descriptive approach offers a very cost-effective way to reveal what the size of the prospective market might be. In fact, such an approach may be the only feasible approach depending on fill rates in the database.
Once identified, this population of prospective sites or look-alikes can then be pulled from the database, matched with a contact database or passed to a call center for verification of firmagraphics, contact details and technology usage.
TUDLA’s Latin America database
As we discussed in a previous post, TUDLA maintains a database of over 1 million business sites in Latin America and the Caribbean. The data include basic firmagraphics (company name, address, URL) and expanded firmagraphics where available (employee counts, industry and revenue). Also, indicators of what technology may be used at these business site are available.
TUDLA also offers free access to the data through interactive Tableau dashboards. This allows users to explore the data at an aggregate level and get a sense of TUDLA’s coverage. It also allows users to perform their own descriptive look-alike analysis.
Descriptive look-alike analysis
As a simple example, suppose I am SAP and I want to know how many prospective customers there might be for an SAP branded ERP offering in Brazil. (Alternatively, I could be an SAP ERP competitor and I am looking for possible SAP customer sites for a competitive kill campaign). Consider the following screenshot from TUDLA’s interactive Tableau dashboard.
TUDLA’s database shows (lower left-hand corner) that there are 2,828 sites in Brazil that are known to have SAP ERP. And the charts show the characteristics of these sites (i.e. the employee, revenue and industry distributions).
Also in the left-hand corner is shown that there are 424,693 sites in Brazil that share the same characteristics (employee count, revenue size and industry) as these 2,828 sites known to have SAP ERP (i.e. these are SAP’s look-alikes). So, this is the size of the prospective market for SAP in Brazil.
Further refinement to a specific firmagraphic, say, revenue size, will more narrowly focus the prospective market. For example, suppose SAP is only interested in sites with enterprise revenue greater than $50m. In this case, the prospective market size is reduced to 19,144 sites, nearly 50% of which are in the manufacturing industry (see below).
Interactive visualization
Interactive visualization is an excellent way to get a comprehensive view of exactly what is in your database. Lack of coverage in firmagraphics, geographies and other business attributes, such as technology presence, can quickly become apparent.
Moreover, it can be used, as in the case of TUDLA, to examine your look-alike market. You can quickly and cost-effectively (don’t need to pay to have a predictive model built) see the size of your prospective market based on the characteristics of your current customers.
Are your sales territories aligned as they should be? With just a few clicks, an interactive visualization can provide the answer.
For more information on providing an interactive visualization for your data, give us a shout.
KDD Analytics is providing analytical services to TUDLA.