How to build a target account list

A list is a set of claims, not a set of names

Most target account lists are built backwards: export everything that matches the filters, sort by employee count, call it strategy. The result is a list nobody trusts, which means a list nobody works — reps drift back to whoever answers, and the “target list” becomes a spreadsheet with a date on it.

A list worth working is different in kind: every account on it carries evidence — which of your best customers’ problems it demonstrably has, what proves the problem is live, when a human last verified each claim. Building one takes six steps, and none of them requires software to start.

---

Step 1 — Define fit from your best customers

Which problems do your best customers share?

Before touching a database, spend an hour on the accounts you already won and love: what were the three to five of them actually dealing with when they bought? Not their industry — their situation. A throughput stall. A new leader with a mandate. A consolidation with a deadline.

Write those problems down in plain sentences. That list is your fit test, and it’s the step almost everyone skips — which is why almost every target list is really a firmographic export wearing a strategy costume. Filters describe what a company is; fit describes what it’s trying to change, and only the second predicts buying. (The full argument: which accounts do we double down on?)

Step 2 — Assemble the universe

What’s the raw material?

Now use the filters — industry, size, geography, whatever genuinely disqualifies. Pull everything that passes into one place. This is the universe, and the correct emotional relationship to it is suspicion: it’s a pile of candidates, most of which you’ll cut, not a list.

The universe’s only job is completeness. If your addressable market is finite enough to name — and if you’re running account-based sales it should be — this step is a morning, not a project.

Step 3 — Score fit with evidence

Which candidates demonstrably have your problems?

For each candidate account, look for your fit problems in their words: announcements, job postings, leaders’ interviews, initiative pages. Record what you find as claims with receipts — which problem, what source, what date.

The discipline that makes the list durable: no dated evidence, no fit score. An account that “feels right” scores zero until something checkable says otherwise. This is slower than a lookalike model and it is the entire difference between a list that survives contact with reps and one that doesn’t — because reps can argue with a feeling, but a dated source ends the meeting.

Step 4 — Score timing with evidence

What proves it’s live now?

Fit without timing wastes the quarter. For each account that scored on fit, record what proves the problem is being worked now: an initiative with a named owner, budget that moved, a hire whose job description is your problem statement, a deadline the account didn’t choose.

Be strict about halo: growth, headlines, and a good quarter are not timing. And date these claims too — timing evidence rots fastest of all, which is how target lists quietly become fiction.

Step 5 — Cut to the working cohort

What can the team actually cover?

Rank by fit-plus-timing and draw the line where coverage math says to: the number of accounts a seller can research, map, and multi-thread properly — tens, not hundreds. This is the cut most teams refuse to make, because a short list feels like a small ambition. It’s the opposite: the two hundred–account “list” guarantees every account gets a two-hundredth of the attention, which is how the year lands light.

Everything below the line isn’t deleted. It waits — deliberately.

Step 6 — Set re-entry signals and run the weekly pass

What reopens an account, and what keeps the list true?

Every below-the-line account gets a named re-entry event: a leadership change in the right function, a stated initiative, money moving. When one fires, the account re-enters on evidence. (What a signal is actually for.)

Then the maintenance that separates a living list from January wallpaper: a fifteen-minute weekly pass that re-dates stale claims, processes fired signals, and moves accounts across the line in both directions. A list without this pass has a lifespan of about one quarter.

The template

One account, one row

The columns that make a target list defensible. Copy the labels; the values show one illustrative account.

Identity & verdict

• Account · ownerMeridian Fabrication · owner: J. Alvarez
• VerdictWorking cohort — rank 6
• Reviewed onSep 26

Fit evidence — which best-customer problems, with receipts

• Problem match 1Throughput stall — COO interview, trade press, Sep 4
• Problem match 2Systems consolidation — careers page: two integration roles, Sep 18

Timing evidence — why now, dated

• Initiative & owner"Operational excellence program," VP Ops named — earnings call, Aug 30
• Money / deadlineBudgeted this FY per COO interview — verify in first conversation

If below the line

• Re-entry signalNew ops or engineering leadership · funded modernization initiative

Where lists die, and what to do about it

Everything above works in a spreadsheet, and should — for the first cohort. Then the failure mode arrives on schedule: evidence rots silently, re-entry signals fire unseen, and the weekly pass loses to the quarter. The list was right the day it was built and quietly stops being true, which is worse than never building it, because concentration multiplies whatever the record gets wrong.

Systematize the evidence layer, keep the verdicts human. Narrative AI keeps each account’s fit and timing claims current, dated, and watched for re-entry events, so the weekly pass is minutes of judgment instead of an afternoon of re-research — the leader still draws the line, every message stays rep-initiated, and every AI-suggested CRM change waits for the rep’s approval.

Six steps, one template, one weekly pass. The list is the strategy — build it like it.