Account prioritization: how to decide which accounts to work first

Account prioritization is the step between building a target account list and starting outreach. The list says which companies are worth pursuing at all. Prioritization says which of them get your team’s hours this week, which wait, and which you leave alone.

It matters because selling time is short. Salesforce’s State of Sales research, a survey of more than 7,700 sales professionals published in December 2022, found that reps spend 28% of their time selling. Every hour spent on an account that is not ready is an hour taken from one that is.

Two related pages cover the parts this one builds on. Account scoring explains how to turn fit, timing and reach into a number, and the account scoring matrix is a template for doing it. This page is about what comes next: turning scores into tiers, deciding what each tier gets, and moving accounts between tiers as their timing changes.

• Key takeaway: prioritize on three separate questions: could we serve them, is there a reason to buy now, and can we reach who decides.
• Key takeaway: a tier is a decision about effort, not just a label. Each tier needs its own playbook.
• Key takeaway: tiers move. Timing changes by the day, so re-tier whenever a signal is added or ages.

Gate on fit, score what passes, tier by score, then match the work to each tier. Six steps cover it.

1. Write your ideal customer profile as must, prefer and exclude rules. A company that fails a must rule or matches an exclusion goes straight to tier X, with the rule it failed.
2. Score fit for everyone who passes: most points for verified must rules, the rest for preferred rules matched.
3. Score timing from dated events, such as a new executive, a new owner or a new site. Give each event type a window, for example 90 days after a new leader starts, and give more credit early in the window. See trigger events in sales for the common types.
4. Score reach: which buying committee roles you have found, each with a current, dated source.
5. Add the three into a score out of 100 and assign tiers A, B and C by cut-off.
6. Give each tier a playbook and a capacity, and re-tier every week, or whenever a signal is added or ages.

An account prioritization matrix crosses fit with timing, so you can place an account in a tier at a glance before you score it in detail. Fit sets the rows, timing sets the columns, and reach decides the close calls.

Two cells do most of the work. The top left is where to spend the week. The bottom row is where to spend nothing: a fresh event at a company you cannot serve is still a company you cannot serve. A strong-fit account with nothing happening is worth keeping close, because one dated event can move it to tier A.

Tiers A, B, C and X form a simple account tiering model. Each account is scored on fit (up to 45 points), timing (up to 40) and reach (up to 15), and the total sets the tier. X sits outside the score: those accounts failed a rule and are not scored at all.

Because timing carries 40 points, an account with no live signal tops out at 60 and cannot reach tier A. That is deliberate: fit says who to sell to, timing says when. Reach is weighted least because it is the gap research can close; a tier A account with thin reach needs its committee filled before outreach, not a lower tier. Forrester’s State of Business Buying 2026 reports that the typical buying decision includes 13 internal stakeholders and nine external influencers, so one named contact is rarely enough.

Marketing teams often describe tiers by treatment instead: one-to-one, one-to-few and one-to-many. The two fit together. Use fit, timing and reach to decide which accounts belong in each tier, then use the treatment to decide how much each tier gets.

A spreadsheet version needs one row per account and a few fixed point values. The values below are round numbers chosen to be easy to apply; the account scoring matrix has a finer version.

Keep the evidence in the next column: the rule each check passed, the signal’s date and source, and the people found. A score nobody can explain is a score nobody will follow.

Here is a fictional list for a supplier that sells to manufacturers, prioritized on Nov 9, 2026. Its must rules are: makes products in its own plants, 200 to 2,000 employees, and independent with no pending sale. Its preferred rules are recent capital investment and more than one site. The team gives a new leader a 90-day window and a new owner or an expansion permit a 180-day window. Halvorsen Castings is the site’s sample account; the other rows are unnamed, fictional candidates.

Three things stand out. First, Halvorsen ranks top because of two dated events, not because it is the best fit; the die-casting supplier fits better. Second, the die-casting supplier is tier A by score but cannot be worked yet: the next hour goes to finding its people, not to writing to its CEO. Third, the plastics molder has the freshest event on the list and still gets no time, because no event changes a failed must rule.

Want this on your own market? Get 25 qualified prospects, free: each comes with a dated reason to reach out and the people who decide.

Set tier sizes by the hours you have, not by the score alone. Work out how many new tier A accounts each seller can open properly in a week, and keep the tier A queue near that number.

An example with stated assumptions: a seller who sets aside 8 hours a week for new accounts, and spends about 2 hours on each new tier A account in its first week (research check, committee map, written first messages), can open about 4 a week. A team of three can open about 12. If the list holds 40 tier A accounts, most of their windows will close before anyone gets to them, so sort tier A by window close date and start with the soonest.

• Too few tier A accounts every week: widen the signal types or the windows before you lower the tier A cut-off.
• Too many tier A accounts: sort by window close date, or raise the strength needed for a signal to count as high.
• A large tier C: normal. Most good-fit companies have nothing happening at any given time. Keep them, and re-check them for new events.

Whenever its timing or its reach changes, which for timing is every day. Fit changes slowly; signals age through their windows on a fixed calendar. Here is Halvorsen Castings on four dates, with fit and reach unchanged.

The account was worth the same to the business on all four dates. What changed was how much a call was worth. Other moves to watch for: a pending sale that closes or falls through (an X account may come back), a contact who leaves (reach drops, and the account may need its committee refreshed), and a new event at a tier C account (it can jump straight to A). This is the core of signal-based selling.

• Prioritizing on fit alone, so the same companies top the list every month whether or not anything is happening there.
• Letting a fresh event rescue a company that fails a must rule. Fit is a gate, not a weight.
• Dropping an account to a lower tier because reach is thin. Reach is usually a research gap; fill it.
• Tiering once a quarter. Signals age by the day, so a quarterly tier list is out of date within weeks.
• Building tiers bigger than the team can work, so tier A becomes a list of missed windows.
• Removing tier X accounts without a reason. Keep the rule each one failed, so nobody researches it again by accident; see why accounts get rejected.

SalesOne scores each qualified account on fit, timing and reach, weighted 45, 40 and 15 by default, and an ICP can set its own weights. Accounts land in tiers A, B, C or X: tier A needs a score of 65 or more and a live signal, and a failed must rule or a matching exclusion gives X. Timing falls as a signal ages in its window, and extra signals add points.

Each trigger has a response window, for example 90 days after a new executive starts. SalesOne puts accounts with an open window first and shows the date to reach out by. Every company S1 turns down keeps the rule it failed and the evidence, so you can see why it isn’t on your list. Each account’s score is shown with its tier, best trigger and reasons. See account research for what each account record holds, and SalesOne for manufacturing for how this works when you sell to plants.

Source for SalesOne claims: SalesOne production research runs and internal tests, Sept–Oct 2026; method on the benchmark page.