Automated Data Onboarding: Decision-making Tasks

Automated data onboarding: decision-making tasks

This is part 5 of our new blog series on automated data onboarding / list loading. Thus far we finished covering data preparation tasks. Today we will move right into decision making tasks.

If you need to read the previous parts of this client data onboarding blog series, you can find them here, here, here, and here.

B. Decision-Making Tasks

B.1    Segment

For a database to support any type of automation, whether it’s personalization, attribution, or routing, it needs to be properly segmented across multiple dimensions, especially if the input data is unstructured data. Unstructured data, like Job Title is nearly impossible to use to drive automation reliably. Common examples of segmentation includes:

  • Segmenting Job =Title into Job Function, Job Sub-Function, and Job Role, e.g., CISO → Job Function = “IT”, Job Sub-Function = “security”, and Job Role = “executive”
  • Assigning buyer personas based on Job Function, Job Sub-Function, Job Level, and other fields such as Industry and Company Size, e.g., Job Function = “IT”, Job Sub-Function = “security”, Job Level = “executive” → Buying Group  Role = “decision maker”
  • Segmenting Industry or SIC Code into 15 to 30 targeted Industries. Anything more than that is very hard to use to support automation. Many companies may also have custom Industry segments or custom Industry hierarchy.
  • Segmenting Company Size based on annual revenue, number of employees, or any industry specific metrics, e.g., number of beds in a hospital, number of open job openings, number connected devices, or number of cars in the fleet.

B.2    Score Lead/Contact/Person

Scoring helps to prioritize your sales team’s focus on the most promising leads, thus optimizing your sales resources to maximize revenue opportunities. This is especially important for list loading because this is when many leads get introduced in a very short period of time, thus overwhelming your available sales resources to perform timely follow up.

Scoring usually consists of two components: demographic and behavior. Behavioral scoring is usually not a big part of list loading process, partly because this is often when a lead is introduced to the system for the first time, thus without many activities to score on. If the lead being loaded already exists in the system with a substantial amount of past activities, then it’s best to score the lead once it’s has been updated in the system. As far as list loading is concerned, demographic scoring is the main scoring task. Demographic scoring is commonly based on the following segmentation dimensions:

  • Buyer Persona, Job Function, Job Sub-Function, or Job Level
  • Is the lead from a targeted account or an account with an active opportunity?
  • Is the lead from a targeted industry?
  • Is the lead from a high-value channel or source?

B.3    Score Anonymous Activities and Accounts

If the data being loaded is activity/intent data from anonymous leads, then behavior scoring takes center stage. In this case, the scoring task usually involves:

  • Normalize engagement/intent and account data from multiple sources
  • Score each engagement activity, which can include downloading an asset, performing a relevant search, or doing a “like” on a specific asset
  • Sum up the activity level score at the account level
  • If the activity is associated with an IP address, then you must first append company data to the IP address

B.4    Reject

Your data may contain leads and accounts you don’t want to work with or are not allowed to work with. These leads should be rejected outright and not even loaded into your systems or loaded into your system marked as disqualified. Common reasons for rejections are:

  • Countries that you don’t do business with, e.g., Sudan, North Korea, Iran, and Cuba
  • Industries that you don’t do business with, e.g., healthcare, education, and public sector
  • Competitors
  • Irrelevant job segments
  • Accounts too big or too small
  • Leads with personal email addresses and disposable email addresses
  • Leads with non-deliverable email addresses
  • Leads missing critical information like email or job title

B.5    Route and Assign

In order to initiate sales engagement, a lead needs to be routed to either a salesperson or a sales queue. The routing task is usually non-trivial for all but the smallest companies. In order to route properly, your list loading process needs to be able to handle these typical routing rules at a minimum:

  • Key/named accounts are routed to the account owner
  • Commercial accounts are routed by geography, which depends on Country, State/Province, County, City, and Postal Code
  • Specific industry accounts are routed to industry teams such as federal government, state and local government, higher education, K-12 education, intelligence agencies, etc.
  • Channel-registered leads are routed differently than direct accounts, and may need to be routed to both an account owner and a channel manager
  • SMB accounts are routed using round-robin or load balancing schemes to dynamically assign leads to the next/most available salesperson

Next and final set of client data onboarding tasks is action execution tasks. We will continue with that next time.

Leave a comment