ABCDX Segmentation: How to Grade Accounts by ICP Fit for ABM

Short answer: ABCDX segmentation grades every account in your market by how closely it matches your ideal customer profile. A accounts are ideal fit, B are strong fit, C are partial fit, D are poor fit, and X are explicitly excluded: competitors, disqualified segments, or do-not-contact accounts. The grade decides how much sales and marketing investment each account gets, and the X bucket stops wasted spend before it starts.

ABCDX borrows its letters from classic ABC analysis, the Pareto-style method of ranking inventory or customers by value, and adds two buckets B2B teams need: a D grade for accounts that technically qualify but rarely close, and an X grade for accounts you should never target. It is a fit model. It answers "should we ever sell to this company?" before anyone asks "is this company buying right now?"

What each ABCDX grade means

The X grade is the part most teams skip and the part that saves the most money. Every ad impression served to a competitor, every sequence sent to an existing customer by a new-logo rep, and every SDR hour spent on a disqualified segment is spend with no possible return. Our negative ICP filtering framework goes deeper on building exclusion rules.

ABCDX vs tier 1, 2, 3: fit grade vs investment tier

ABCDX and the familiar tier 1, 2, 3 ABM model are often confused, and the distinction matters.

The cleanest setup uses both: ABCDX grades the whole market for fit, then tiering selects accounts from the A and B grades based on timing and capacity. A typical mapping is that tier 1 comes from A accounts showing intent, tier 2 from remaining A accounts plus B accounts showing intent, and tier 3 from other B accounts and engaged C accounts. For the tiering side, see what account tiering is.

If your team grades fit in a spreadsheet and tiers by gut feel, the two drift apart fast. See how Abmatic AI keeps fit, intent, and tier on one account record.

How to build an ABCDX model in seven steps

1. Start from closed-won evidence. Pull customers from the last two years with revenue, retention, sales cycle, and expansion. Separate your best quartile from the rest, and pull closed-lost and churned accounts for contrast.
2. Pick fit attributes. Use firmographics such as industry, employee band, revenue band, region, and ownership, plus technographics such as CRM or cloud platform. If you need a refresher on which attributes are which, see demographics vs firmographics.
3. Classify each attribute value. For every attribute, mark values as must-have, strong fit, neutral, or disqualifying, based on how they distribute across winners and losers.
4. Write the X rules first. List hard exclusions before you score anything: competitors, current customers owned by account management, partners, restricted regions, legal or reputational exclusions, and contacts or domains on do-not-contact lists.
5. Score the rest. Give points for strong-fit values and remove points for weak ones. Keep it simple enough that a rep can explain why an account is an A.
6. Set grade cutoffs. Run the score across your full account universe and set thresholds. Then sanity-check: your best current customers should almost all grade A or B. If they do not, the model is wrong, not the customers.
7. Publish and sync. Write the grade to a field on the account object in your CRM so routing, ads, outbound, and reporting all read the same value. Re-score quarterly.

Example ABCDX scoring model

Illustrative cutoffs for this model: A at 80 or more, B at 60 to 79, C at 40 to 59, D below 40. The points and cutoffs are an example only; calibrate yours against your own closed-won data. Our explainer on ICP fit scores covers weighting methods in more detail.

Combining ABCDX with intent: the fit by timing matrix

A fit grade alone tells you who to sell to, not when. The highest-leverage view crosses ABCDX with intent and engagement.

This is where most ABCDX models stall: the grade lives in the CRM and the intent lives in a separate tool, so nobody sees both at once. Abmatic AI puts the grade and live first-party intent on the same account and acts on the combination automatically.

Common ABCDX mistakes

• Grading from aspiration. If the A grade describes the logos you wish you had rather than the customers who retain, the model will send budget to accounts that do not close.
• Too many A accounts. If a large share of your market grades A, the cutoffs are too loose and A stops meaning anything.
• No X bucket. Without explicit exclusions, competitors and existing customers leak into new-logo campaigns.
• Mixing fit and timing. Adding intent points into the fit score makes grades swing week to week. Keep fit stable and layer timing on top.
• Ignoring corporate hierarchy. Grade the parent and decide deliberately how subsidiaries inherit, or one enterprise ends up as a dozen separately graded accounts.
• Never re-scoring. Companies grow, get acquired, and change stacks. Re-run grades quarterly and whenever the ICP changes.

For a step by step on tiering once grades exist, see how to build account tiering. To avoid every one of these mistakes in one place, see how Abmatic AI grades, suppresses, and re-scores accounts automatically.

Running ABCDX segmentation in Abmatic AI

Abmatic AI is the most comprehensive AI-native revenue platform for B2B teams: 15+ first-party modules on one identity graph and one signal layer. It handles 50 to 50,000+ target accounts across tier 1 (1:1), tier 2 (1:few), and broad 1:many programs natively, which is exactly the range an ABCDX model spans. For ABCDX specifically:

• Account list building from firmographic, technographic, and intent filters on a first-party database, so every account in your market can be graded, not just the ones already in your CRM.
• Technology scraper that detects each prospect's tech stack, feeding the technographic part of the fit score.
• Account-level and contact-level deanonymization, so graded accounts are recognized when they visit, even before a form fill.
• First-party intent and third-party intent for the timing axis of the fit by timing matrix.
• Agentic Workflows that act on grade plus signal, for example: when an A account crosses an intent threshold, show a personalized banner, alert the AE in Slack, and enroll the buying group in a sequence.
• Web personalization and A/B testing by grade, so A accounts see a different experience from C accounts.
• Google DSP, LinkedIn Ads, and Meta Ads audiences built from grades, with X accounts suppressed everywhere.
• Salesforce and HubSpot bi-directional sync, so the grade lives on the CRM account object your routing and reporting already use.

Pricing starts at $36,000 per year with enterprise tiers available, and first-party signal capture goes live the same day the pixel does. Book a demo to grade your own market.

Frequently Asked Questions

What does ABCDX stand for?

ABCDX is a grading scale rather than an acronym. A means ideal ICP fit, B strong fit, C partial fit, D poor fit, and X excluded. Each grade maps to a level of sales and marketing investment, with X accounts suppressed from all targeting.

What is the X in ABCDX segmentation?

X marks accounts you should never target: competitors, existing customers handled by a separate team, partners, restricted regions, and do-not-contact lists. Grading them X keeps them out of ads, outbound, and lead routing, which removes spend that cannot produce pipeline.

How is ABCDX different from tier 1, 2, 3 account tiering?

ABCDX measures static fit with your ICP across your whole market. Tier 1, 2, 3 decides how much you invest in specific accounts right now, using fit plus intent, engagement, and capacity. Most teams grade with ABCDX first, then pick tiered accounts from the A and B grades.

How often should ABCDX grades be updated?

Re-score quarterly, and immediately whenever your ICP changes, for example after a new product launch or a shift in target segment. Fit attributes such as industry and size change slowly, so frequent re-grading mostly adds noise; timing signals belong in a separate intent layer.

What data do you need for ABCDX segmentation?

You need closed-won and closed-lost history to calibrate, firmographic data such as industry, employee band, revenue band, region, and ownership, technographic data on key platforms, and a list of exclusions. Abmatic AI supplies firmographic and technographic data from its first-party database and syncs grades to Salesforce or HubSpot.

Is ABCDX segmentation only for ABM?

No. The same fit grades work for territory planning, SDR coverage, lead routing, and customer success segmentation. It is especially useful for ABM because ABM depends on deciding which accounts deserve concentrated investment.