ABM Vendor Guide: Features to Look for in Target Scoring Vendors

My last post used data from our new Guide to ABM Vendors to describe differentiators among companies that provide external data for account based marketing. Let’s continue the series by looking at differentiators related to Target Scoring, the second sub-function related to the ABM process of identifying target accounts.

ABM Process	System Function	Sub-Function	Number of Vendors
Identify Target Accounts	Assemble Data	External Data	28
	Select Targets	Target Scoring	15
Plan Interactions	Assemble Messages	Customized Messages	6
	Select Messages	State-Based Flows	10
Execute Interactions	Deliver Messages	Execution	19
Analyze Results	Reporting	Result Analysis	16

While External Data is one of the broadest sub-functions described in the Guide, Target Scoring is one of the narrowest. Target Scoring isn’t just any use of predictive analytics, which can also include things like finding surges in content consumption (used to identify intent) or recommending the best content to send an individual. As the Guide defines it:

Vendors in this category use statistical techniques to select target accounts. The models most often predict whether an account will make a purchase, but sometimes predict events such as renewing a contract or becoming an opportunity in the sales pipeline. Scores can be built for individuals as well as accounts, although account scores are most important for ABM. Many scoring vendors gather external data from public or commercial sources (or both) to gain more inputs for their models. They may or may not share this data with their clients, and they may or may not provide net new records. Target scoring is more than tracking intent surges, which do not capture other factors that contribute to likelihood of purchase.

The vendors in this category include the specialized scoring firms (Infer, Lattice Engines, Leadspace, Mintigo, Radius) plus companies that do scoring as part of a data offering (Avention, Datanyze, Dun & Bradstreet, InsideView, GrowthIntel) or for message targeting (Demandbase, Everstring, Evergage, Mariana, MRP, The Big Willow). Beyond those fundamental differences in the vendor businesses, specific differentiators include:

• the range of data used to build models, including which data types and how much is proprietary to the vendor 
• amount of client data (if any) loaded into the system and retained after models are built
• advanced matching of unaffiliated leads to accounts (an important part of preparing data for account-level modeling)
• tracking movement of accounts and contacts through different segments over time (as opposed to simply providing scores or target lists on demand)
• self-service model building (as opposed to relying on vendor staff to build models for clients)
• separate fit, engagement, and intent scores (as opposed to a single over-all score)
• range of model types created (fit, engagement, behavior, product affinity, content consumption, etc.)
• limits on number of models included in the base fee
• implementation time (for the first model) and model creation time (for subsequent models
• sales advisory outputs including talking points, intent indicators, product recommendations, content suggestions, etc. 

Even though target selection is obviously a core ABM process, target scoring is distinctly optional.  Most firms already have target account lists that were built by sales teams based on their own marketing knowledge.  An ABM program can easily get started using that list.  Chances are, though, that target scoring will find some high-potential accounts that aren't on the old list and find some low-potential accounts that are on the old list but shouldn't be.  Scoring can also do a better job of prioritizing accounts within the list, often by incorporating event and intent information to uncover opportunities that would otherwise remain hidden.  So although account scoring may not be the first thing you do when setting up your ABM program, it is something to consider adding as you move along the ABM path.