What does ABM segmentation actually decide?
ABM segmentation splits a named-account list into tiers. The tier sets three things for every account: how deep the research goes, how many people you contact there, and how many touches each person gets. Most teams work with three tiers: 1:1 for a handful of top accounts, 1:few for small clusters, and 1:many for the wide list.
Common Room's ABM playbook uses the same three tiers and puts rough sizes on them: up to 10 accounts for 1:1, 10 to 100 for 1:few, and hundreds or more for 1:many. Treat those as starting points. Your real numbers come from rep hours, as the capacity section below shows.
Two inputs place an account in a tier. One is what a deal there could be worth. The other is a live signal you can see for free. Two more set how many accounts each tier holds: a rep's real hours, and the hours one account in that tier takes. Size the tiers by capacity first, then fill them by value and signal.
Tiering works on a list you already have. Whether to go after named accounts at all, or a broad market, is an earlier choice. Settle the trade-off between account-based or volume outbound before you tier anything, and build the list itself first with this ABM list guide.
The usual starting point is a list that is too long. Forrester's best-practice report on named-account segmentation starts from this problem. The account list is often long, and sales and marketing need a data-driven way to refine it so both agree which accounts get the effort (Forrester).
I never give a client a ready-made tier model, because no two businesses sell the same way. Everything below is built from your own hours and your own closed deals.
How do the three tiers differ in practice?
1:1 treats one account as its own campaign. You research the company, map the people who buy, and write to each of them about that company's situation.
1:few groups a small cluster of accounts that share one problem or trigger: the same sector, the same kind of project or the same recent event. You write one message per cluster and add one line for each account.
1:many is one offer to one audience at volume. The precision comes from who is on the list, and the copy stays nearly the same for everyone.
| Tier | Accounts per rep (worked example below) | Research per account | People contacted per account | Message | Owner |
|---|---|---|---|---|---|
| 1:1 | About 10 | Deep: people map, current projects, recent news, any past contact | 5 to 8 | Written for each person | One named rep |
| 1:few | About 40, in clusters of about 10 | Shared cluster research plus one fact per account | 2 to 3 | One per cluster, plus one custom line | One named rep |
| 1:many | Hundreds to a few thousand, capped by sending capacity | Fit check only | 1 to 2 | One offer for the whole audience | One rep per territory |
Why does 1:1 need so many contacts? Forrester's 2024 State of Business Buying research found an average of 13 people taking part in each purchase decision (Forrester). Its Buyers' Journey Survey, 2024 lists these roles in those groups: the user, the ratifier (who signs off at the end), the influencer, the champion and the decision-maker (Forrester).
So a 1:1 account gets a row for each role in your sheet, even if some cells stay empty for a while.
You map everyone you can find. Then you contact the 5 to 8 who decide or carry the decision, with the champion and the decision-maker first and the ratifier later. Buying groups differ by deal size, so check that range against your own closed deals.
A 1:few cluster is only real if you can name its shared problem in one sentence. "Contractors that won a public hospital job this year" is a cluster. "Mid-size companies in Europe" is a filter.
1:many needs a narrow focus. A 1:many campaign written for everybody lands with nobody. So I keep it to one offer for one audience, and I keep that offer running for a year rather than swapping it every month.
Which criteria put an account in a tier?
Your current customers are the best template. I use that client list as the template and pull the target companies that resemble your best current clients. Then I go back through those past deals to see who actually signed off, because that title is the one to search for.
Five criteria do the work for most B2B lists:
- Firmographic fit: industry, headcount band, revenue band and country, judged against your past clients.
- Lookalike strength: how closely the account resembles a client you have already won.
- Deal value: the size of the deal this account could sign, measured in bands taken from your own closed deals.
- Reachable buying group: you can name at least three real titles at the account and find the people who hold them.
- Live signal: something dated and public that says this account has a reason to talk now.
Judgment beats software here, even at large companies. In a 2026 study of 40 companies, most of them large enterprises (an online survey plus 16 interviews, February to April 2026, run by the ABM consultancy Inflexion Group, so the publisher's own research), prioritization leaned on future growth potential (97%) and current revenues (77%). Only 16% relied on predictive scoring (Inflexion Group).
Deal value has to drive placement. One relationship with a very large contractor can be worth more than dozens of small sites, so accounts like that belong in 1:1 or 1:few. Smaller regional contractors buying the same equipment belong in 1:many, where the volume lives.
Titles follow the same rules in every tier. I cut the generic titles first: Assistant, Associate and a bare Manager all go. On a construction list, the titles that stay are the ones that buy on site: head of procurement, contracts manager, project director and health and safety director. The tier changes how many of those people you contact. It never changes who counts as a real buyer.
Now cross the two placement inputs. Tier 1 means 1:1, Tier 2 means 1:few and Tier 3 means 1:many. Set the value bands from your own median closed deal. The 3 times and half-median cut-offs below are example starting points, so reset them from your own deals:
| Deal value (from your closed deals) | Live signal in the last 90 days | No live signal |
|---|---|---|
| High: 3 times your median deal or more | Tier 1 | Tier 2, first in line for a Tier 1 slot |
| Mid: from half your median up to 3 times | Tier 2 | Tier 3 |
| Low: under half your median | Tier 3, sent first | Tier 3 |
These signals are free and visible:
- A new hire in the role that buys from you, visible on LinkedIn.
- An open job post for that role, or for the team your product supports.
- A public win, such as a tender award, a new project, a new site or an acquisition.
- The person has been active on LinkedIn within the past month, which roughly shows that a human checks the account.
- Past contact: an earlier reply, a lost deal, an introduction or a referral.
One rule keeps the table honest. A signal gets a date and a link in the sheet. No date means no signal.
How many accounts can one rep handle in each tier?
The numbers below are assumptions, not benchmarks. Log your own hours for one month, then replace every figure.
- Hours: one rep has about 100 hours a month for outreach once calls, CRM work and internal meetings are taken out.
- Split: 40 hours go to Tier 1, 30 to Tier 2 and 30 to Tier 3.
- Tier 1: assume 4 hours per account per month for mapping, writing to each person and handling replies. 40 hours divided by 4 gives 10 accounts.
- Tier 2: assume 45 minutes per account per month, with the cluster research shared. 30 hours divided by 0.75 gives 40 accounts.
- Tier 3: the hours go to list building, testing and replies. The real cap is sending capacity.
That capacity comes from sending accounts, and each one belongs to one person. Scaling Tier 3 means adding more LinkedIn or email accounts to what already works, and every added account needs its own named owner.
Ownership has to be just as clean. Each account gets exactly one owner, and no rep ever writes into another rep's accounts. An account appears once in the whole sheet, with one tier and one owner, so a Tier 1 champion never also receives a Tier 3 cold email from a colleague.
Ten Tier 1 accounts with 5 to 8 people each is a pool of 50 to 80 people. At volume rates, about 1% of people reply and 10% to 50% of those replies become meetings, so one booked call takes roughly 200 to 1,000 people contacted. At those rates, the whole Tier 1 pool gives you less than one booked call a month.
Reach 50 to 100 ideal-fit people in a month and most of them will be busy, not ready or not interested. That list cannot be your whole plan. Personalization lifts Tier 1 reply rates, as the next section shows, but a higher rate on 60 people is still a handful of replies, not a full calendar.
That job falls to Tier 3. The decision rule follows. If your calendar target is more than a few calls a month, 1:many has to carry most of it, and Tier 1 is reserved for the deals that justify the hours. Our campaigns tell the same story. Booking 30 to 35 calls a month needs 10,000 people contacted. That is the minimum, every month.
What email and LinkedIn sequence does each tier get?
| Tier | People per account | Touches per person | Channel order | Personalization | Who sends |
|---|---|---|---|---|---|
| 1:1 | 5 to 8, champion and decision-maker first | 6 or more | LinkedIn connection and note, then warm email, then follow-ups across both | Every message written for that person and that account | The account owner, from their own profile and inbox |
| 1:few | 2 to 3 | 5 to 6 | LinkedIn first, warm email after | One cluster message plus one line about the account | The account owner |
| 1:many | 1 to 2 | A short cold email sequence, plus a LinkedIn touch where the profile has room | Cold email track, LinkedIn as follow-up | One offer, one audience, a custom first line at most | The territory owner's sending accounts |
On a top account, the email never opens the relationship. A LinkedIn touch goes first, so the email that follows is a warm one: it reaches someone who has already seen your name. One or two messages almost never start a conversation, so every Tier 1 person gets at least 6 touches across both channels. The message that most often gets the reply in our campaigns is follow-up number three, once enough days have passed since the second.
Set reply expectations per tier before you judge anything. Tier 3 sits at the volume rates from the capacity section, and often nearer 0.5% replies at thousands of emails a day.
Tier 1 is different. When each account is mapped and every note is written for one named person, I have seen replies reach 15% to 17%, with roughly three in four repliers keen to talk. Those ranges come from my own campaigns, and the bigger the volume, the lower they sit.
Take a Tier 1 pool of 60 people as an illustration: at the top of that range, a full sequence would give about 9 replies, 6 or 7 of them interested. That pays when each deal is large.
In 1:many, I keep emails short and human, the way a person writes, because copy that reads like a template dies at volume.
Here is a made-up Tier 1 example: a supplier of site welfare cabins writing to a contractor's head of procurement, after connecting on LinkedIn.
Subject: Site cabins for the Leeds hospital job
Hi Sarah, thanks for connecting last week. Your post about the new hospital package in Leeds caught my eye, because site welfare cabins are most of what we hire out to main contractors on UK sites. The question site teams ask us first is how fast a cabin arrives and moves between phases. Worth 15 minutes on a call before those packages get placed? If someone else owns this, I will happily write to them instead.
It runs about 75 words, names one project and one problem, and asks for one small thing.
On construction lists the split between tracks is easy to see. Large contractor groups justify LinkedIn plus email for several people each, while the long tail of regional contractors runs on the 1:many track.
When should an account move up or down a tier?
Promote an account when one of these happens:
- A second person at the account replies, or the first person copies in a colleague.
- A new decision-maker joins in the role that buys from you.
- An introduction or referral into the account arrives from a client, partner or contact.
- Anyone at the account books a call, which keeps the account in Tier 1 for the rest of that sales cycle.
Demote an account when one of these happens:
- Every contact at the account has finished the full sequence with no reply from anyone.
- A call shows the account cannot buy, because of size, budget or need.
- Your champion leaves, or the company is bought or cuts the unit you sell to.
A demoted account drops to 1:many. It only leaves the list when a call has proved it can never buy.
Tiers are full at their capacity number, so moves come in pairs. Promoting one account into Tier 1 means demoting another out of it.
Every move goes into the CRM with the date and the reason. Use any CRM you like, as long as no account changes tier only in somebody's head. When an account moves, check that its owner stays the same, so the one-owner-per-account rule still holds after the move.
Can a small team run tiered ABM without a platform?
Yes. You need a spreadsheet with one row per account, a CRM, verified contacts, and LinkedIn and email accounts set up for sending. The sheet needs columns for value band, signal, signal date, tier, owner, the people mapped and the last touch. That is the whole stack.
Intent data is optional, and sometimes useless. I use it when the market is large enough for signals to show up across a real share of it. If you have 2,000 target companies and intent signals on only about 10, skip it and build an offer that fits all 2,000. The free signals in the criteria section cover most small markets better anyway.
Tiers cannot rescue a bad list, and neither can copy. Fix the list first. Then tier it. A broken foundation sinks even thousands of "personalized" messages a week.
Before any account gets a tier, I check the titles, the countries, the headcount range, when and how the data was verified, signs of life and overlap between reps. Verified addresses only, with bounces under 0.5%. If your replies are thin, the problem is usually the list, and fixing it comes before tiering it.
Open your list and give every account a value band from your own median deal and a dated signal or a blank. Then count how many Tier 1 accounts your hours can really carry.
If you have a named-account list and cannot decide where the tier lines go, book a call and bring it. We can place your top accounts together.
Questions founders ask about ABM tiers
What are the four types of segmentation, and which matter for ABM?
The classic four are demographic, geographic, psychographic and behavioral. In B2B they become firmographic (industry, size, revenue), geographic (country or region), needs-based (the problem the account has) and behavioral (what the account is doing now). For tier placement, firmographic value and behavioral signals do most of the work. Needs-based segmentation mostly shapes how you write 1:few cluster messages.
How is an ICP different from an ABM tier?
An ideal customer profile (ICP) decides whether a company belongs on the list at all. A tier decides how much effort it gets right now. Two accounts can match your ICP equally well and still sit in different tiers, because one just hired a buyer and the other shows nothing. An ICP should hold steady across a year of campaigns. Tiers move whenever the evidence does.
Should the founder or a sales rep send Tier 1 messages?
Send from whichever profile a buyer would take seriously. I would rather see an average message from a strong profile than a perfect one from an empty profile. A founder can open the LinkedIn touch on the top accounts and hand over once a call is booked. Whoever sends needs a profile built for buyers, so fix the reps' profiles too.
What if my named-account list has fewer than 30 companies?
With a list this short, a three-tier split usually adds admin without changing the work. Run every account as 1:1, divide them between reps, and map each buying group fully. Then build a separate 1:many list of similar companies in neighboring sectors or countries, so your calendar does not depend on 30 inboxes answering in the same quarter.
How often should the whole tier list be reviewed?
Re-score the full list once a quarter as a default, and sooner if your sales cycle is short. Deal sizes drift, so recalculate your median deal and the value bands from fresh closed-won data. Re-run the hours split whenever a rep joins or leaves, or when a month of time logs shows your hours-per-account assumptions were wrong.
