Account scoring models

Account scoring gives outbound a repeatable way to decide which accounts deserve attention. Start with the decision the score should improve, then choose signals that answer it. A single score can mix whether you should sell to an account with whether you can reach it this year.¹ Keep fit and timing visible so a strong fit with weak timing does not look like a weak fit with active interest.

Set the decision

Use the model to make a choice. Before assigning points, write down what a high score should prompt and what a low score should rule out.

Account scoring sorts potential customers against an Ideal Customer Profile, or ICP.² An ICP consists of indicators that show whether a company or account could become a good customer.³ Ask which accounts should get focused attention and which should wait.

Keep this decision close to the score. Account tiering ranks accounts.⁴

Choose inputs

Pick inputs that describe the account and its buying situation. A field belongs in the model only if you can explain what it changes in your ranking.

Account assessment uses a collection of different data points.⁵ Most account scoring models combine firmographic fit, technographic fit, and possibly intent into a score from zero to one hundred.⁶ A fit score can cover firmographic, technographic, and operational criteria.⁷

Include business factors that change the value or cost of pursuing an account. These can include current revenue, account-level growth opportunity, cost to serve, and geographic location.⁸ Technographic data can also be a factor when you design equitable account books.⁹

Write a plain-language definition for each input. State what counts as a strong match, what counts as no match, and what happens when data is missing. The definitions are ready when someone else can apply them without asking what you meant.

Keep fit and timing separate

Fit and intent

If fit is high

Keep the account

If fit is weak

Investigate

Fit and intent

• If fit is high, keep the account
• If fit is weak, investigate

Keep structural fit and present buying activity as separate views until you decide how to combine them.

Account prioritization requires both fit and intent.¹⁰ A combined score can use actions and demographic information, while separate engagement and fit scores remain available.¹¹

This shows why an account ranks where it does. If fit is high and engagement is low, keep the account in a suitable tier while treating timing as unresolved. If engagement is high and fit is weak, investigate before spending focused effort.

Assign weights and scores

Weight inputs to match the decision. Give more influence to fields that predict the outcome you care about, and document the reason.

A weighted score can combine firmographic and technographic fit with intent or activity, then assign thresholds and examples to tiers.¹² Use the total to sort accounts, then inspect the component scores before moving an account into action. Score every account in the CRM on ICP fit and expected spend, then roll the scores up by territory.¹³

The score should determine the amount and type of work. A label alone is insufficient. Keep the component scores available so a total cannot hide missing or contradictory inputs.

Set account volume

An accurate score can still fail if the list is too large to work or too small to produce the target. Set account volume alongside scoring.

Use two calculations.¹⁴ The sales model asks, "How many accounts do you need to hit your target?".¹⁵ The capacity model asks, "How many accounts can one rep handle?".¹⁶

Compare the outputs and name the tradeoffs. Both calculations involve tradeoffs that need consideration.¹⁷ Calculate the account target from sales targets with a top-down approach,¹⁸ then adjust it for your business particulars instead of borrowing an industry average.^{19, 20} Once the formulas are clear, use them to choose tradeoffs suited to the business.²¹

The target is ready when it fits both the commercial need and the amount of account work the team can handle.

Turn scores into tiers

Use tiers to turn ranking into coverage decisions. Keep the boundaries visible so people can tell why accounts received attention.

Create tiers with weighted fit and intent or activity, then document thresholds and examples.¹² Target accounts are the highest-fit, highest-potential opportunities for a business.²² A focused program gives these accounts coordinated sales and marketing attention around their situation and buying context.²³

Sort high-value customers by attributes such as lifetime value or account tier to prioritize them.²⁴ Keep tier assignment separate from timing when the model gives you separate fit and engagement views. The score should tell you where to spend focused effort and what kind of attention the account warrants.

Keep the list honest

Scores become stale when the list follows memory. Test whether the ranking still reflects the account definition you set at the start.

The quality of a target account list determines where program resources go.²⁵ Keep target account lists shared across teams.²⁶ Use past success to guide expansion and forward-looking signals when refreshing the list.^{27, 28}

Review familiar accounts and ask whether they would still rank highly if familiarity were removed. Also check which high-fit opportunities are missing because nobody already knows them.

What not to do

Most scoring failures come from using the output without checking what it represents. Keep these failure modes visible while you build and run the model.

• Do not use industry averages to decide how many accounts each person should target.¹⁹
• Do not let sales intuition, existing relationships, or lightly updated prior-year lists determine account selection.²⁹
• Those choices skew the pipeline toward familiar names, leaving high-fit opportunities out.³⁰
• Do not treat industry and company size as the whole ICP; an ICP goes beyond firmographic data.³¹