Account-Based Marketing (ABM) Playbook

Account-based marketing (ABM) means sales and marketing working one agreed list of named accounts together, with the effort each account gets matched to what it is worth. It pays when deals are large and several people sign off, and it works only when the two teams run it as one program.

Most B2B marketing starts with volume and sorts the leads afterwards. ABM starts with the accounts, then reaches everyone who decides inside them. This playbook is about running the program. Choosing and scoring the accounts comes first, in the ICP & Account-Scoring Playbook; here I start from a scored list. For the overall budget, the Marketing Budget Benchmarks by Stage 2026 report gives ranges by stage.

ABM fails as a marketing side project, because the accounts belong to sales. Before you choose a single account, agree four things in writing:

• A sponsor. One executive who signs off the investment and settles disputes; at a startup, usually the CEO.
• A program owner. One person who runs the roadmap, the plays and the numbers.
• An owner for every account. Each named account has one seller and one marketer, and SDRs work the accounts their sellers own.
• A goal in pipeline, not leads. Opportunities and pipeline created in target accounts; then win rate, deal size and cycle length against comparable accounts you are not targeting; and expansion in customers.

Then fix three meetings, each with one decision (Figure 1). The weekly one is where alignment actually happens: sales says what it heard on calls, marketing says which accounts are engaging, and the two agree the next move for each. If you also pass individual leads to sales, keep the handoff rules in the Lead Management Playbook; ABM adds one promise at account level, a deadline for acting when a target account shows a buying signal.

Start from a scored list, not a brainstorm: the tier-A and tier-B accounts from your fit score. Then decide how much attention each one gets. The tiers most programs use come from ITSMA, where Bev Burgess first developed the approach in 2003: strategic ABM, ABM Lite and programmatic ABM, usually called 1:1, 1:few and 1:many^[1].

• 1:1. A handful of accounts where one win would change your year. Each gets its own research, a team from both sides and a program built for it alone.
• 1:few. Small clusters of accounts that share an industry, a use case or a trigger. Each cluster gets messaging and content built on what its members have in common.
• 1:many. The rest of the scored list, reached with rule-based personalization by segment.

Most programs mix tiers: in Momentum ITSMA’s 2024 benchmark of 308 respondents, 61% ran more than one type of ABM^[2]. Burgess’s 2025 handbook now splits ABM into five types^[3]; for a first program, three tiers are enough.

Size the tiers by the hours your team actually has, not by ambition. Estimate the hours one account in each tier needs per quarter, multiply by the number of accounts, and cut the 1:1 tier until the total fits. Then lock the list for the quarter: accounts move between tiers only at the quarterly review.

An account does not buy; a group of people inside it does, and the group is larger than most sales teams assume. Forrester’s 2026 buyer research puts the typical purchase at 13 internal stakeholders and nine external influencers, and finds procurement among the decision makers in 53% of buying cycles^[4].

For every 1:1 account and the largest 1:few accounts, map the roles before anyone reaches out: the economic buyer who holds the budget, the people who own the problem, the users, the champion who wants you to win, and the reviewers in IT, security, legal and procurement who can stop the deal. For each person, note what they are measured on, their likely objection and where they pay attention. The B2B buying committee sets out the usual roles and the question each one asks.

Then add what is happening in the account: its strategic initiatives, recent news (funding, acquisitions, leadership changes, new regulation), the systems and vendors it already runs, and who is for or against you. For a 1:few cluster, one brief on the industry replaces the per-account dossier. Measure coverage, not contacts: the share of roles in each account where you know a name and have had a reply.

A play is a structured, multi-channel program for a set of accounts and the people inside them. Four motions cover most needs:

• New logo. Open opportunities in accounts that are not customers yet.
• Expansion. Sell more to existing customers: another product, team, region or site.
• Deal acceleration. Unblock a stalled deal with senior people and proof.
• Win-back. Revive closed-lost or dormant accounts that still fit.

Start with one or two plays that serve the goal from Step 1, and write each on one page with seven elements: the target accounts and roles; a measurable objective; the message (problem, value, proof); the offer; the channels; a sequence of two to eight weeks; and the measures of success. The offer decides whether anyone replies. In my view, a senior buyer will give an hour to a working session that leaves them with something, such as a diagnostic, a benchmark or a short roadmap, sooner than to a demo.

Run the expansion play

An expansion play turns one team, department or site into the whole account. The pilot or first contract is not the goal; it is the evidence for the larger decision, which is usually made by different people. Plan for that decision from day one.

1. Agree success in writing before you start. The buyer’s own metric and today’s baseline, the target, who will use it, the decision date, the rollout price and who signs. A pilot without these often ends in “interesting, let’s keep going”.
2. Map the second buying group early. The department head who said yes to the pilot rarely signs the rollout. Find out who decides for the wider account (often finance, IT, procurement and a committee) and meet them before the pilot ends, not after.
3. Report in the buyer’s language. Turn results into the one-page summary your champion can forward unchanged: the metric they chose, before and after, cost per unit of result, and what rollout would take.
4. Ask for the next introduction. When the result is in, ask the champion to introduce you to the peer team or the committee that decides for the other sites. Make the ask specific: a name, a meeting, a date.
5. Price the rollout before the pilot ends. Put the per-site or per-team price, volume tiers and the start date in the pilot agreement, so the decision is yes or no, not a new negotiation.

For medical devices, Step 8 of the MedTech Marketing Playbook works this play through a hospital evaluation and a health-system rollout.

Get the data a pilot needs

Buyers say no to data requests that look open-ended, risky or like extra work. Make the request small, safe and easy.

• Ask for the least data that proves the metric. Counts and averages often do; records about individual people rarely need to leave the buyer.
• Use reports they already run. A monthly report the department already produces is faster to approve than a new data feed.
• Bring the paperwork first. Send your data-handling summary, security questionnaire answers and draft agreement before you ask, so their IT, privacy and legal teams start in parallel, not after the pilot clock has started.
• Know which agreement applies. If personal data will reach you, expect the buyer to require a data-processing agreement before any data moves; if de-identified or aggregate data will do, say so, because it can shorten review. For patient data in US healthcare, Step 8 of the MedTech Marketing Playbook covers business associate agreements. This is general information, not legal advice: check with a qualified professional.
• Name who does the work. Offer to build the extract or the report yourself, with their analyst checking it.

In a play, every channel tells the same account the same story in the same weeks. What changes by tier is how many channels you use and how personal each one is (Figure 6):

• Seller and executive outreach: short emails, calls and LinkedIn messages written for each role and built on the play’s message, plus a note from an executive for 1:1 accounts.
• Account-list ads: ads shown only to people at your target companies, tailored by industry and role.
• Web pages: a page for each 1:1 account or each cluster, with the case studies that fit it.
• Events: roundtables, dinners and webinars co-hosted with a customer from the same industry.
• Direct mail: for 1:1 accounts only, tied to the offer and a clear next step.

Account-list ads have a floor. On LinkedIn, a company-list audience serves once it matches at least 300 members, and LinkedIn suggests lists of at least 1,000 companies for best results^[5]. A ten-account 1:1 tier is far below that suggestion, so there I treat ads as reinforcement; they become a main channel only in 1:many.

Test mail before you scale it. I found no public, methodical study of how often business buyers respond to mail, so treat it as a touch to test, not a channel to trust. The closest public evidence is about households: a 2019 report from the Postal Service’s Office of Inspector General found that advertising mail from a business with a past relationship is six times as likely to get a response as mail from one with none, measured as stated likelihood of responding^[6]. So send to people who have a reason to recognise you, and have a person follow up within five working days. Postage is the smallest line: 150 stamped letters, three people at each of 50 accounts, cost $123 at the single-piece First-Class price in USPS Price List Notice 123^[7]. The research, the item and the follow-up cost more. Test on 50 accounts, 25 mailed and 25 not, and count meetings booked, not opens. Be honest about how little 25 accounts a group can show: if 4 of the 25 unmailed accounts book a meeting, the mailed group needs 12 of 25 for the gap to clear the usual 5 percent threshold on a Fisher exact test (my arithmetic), so a small gap is a reason to repeat the test, not to scale it. Direct Mail to Target Accounts sets out the evidence, the costs and the test.

Sales needs more than the list. For each 1:1 account, give the seller a one-page dossier and a talk track. For each cluster, give a cheat sheet per role: what they care about and what to avoid saying. And have one piece of content for each stage the accounts are at: an industry benchmark to open the conversation, a use case or calculator to compare options, and a case study plus a business-case template the champion can take to the CFO.

Judge ABM by what happens inside the accounts, in four layers, each read by tier:

1. Coverage: the share of target accounts with the buying group mapped, and the number of relevant contacts per account.
2. Reach: ad reach and frequency inside target accounts, and their visits to your site.
3. Engagement: accounts above an agreed engagement threshold, meetings booked and event attendance.
4. Pipeline and revenue: opportunities and pipeline from target accounts; win rate, deal size and cycle length; expansion and renewal in customers.

The first three are leading indicators. Report the fourth upward. In 6sense’s 2025 survey of B2B marketers, 39% of ABM programs reported marketing-qualified accounts to the board and only 13% reported closed-won revenue (Figure 8)^[8]. 6sense sells ABM software; in my view, a program judged only on engagement cannot defend its budget.

To know what ABM added, compare like with like. Keep a group of accounts with similar fit scores that you do not target, track the same four layers for them, and compare opportunity rates. Without that comparison, ABM takes the credit for deals that would have come anyway. Use the quarterly review from Step 1 to act on the result: re-tier accounts, keep the plays that open opportunities and retire the ones that do not.

What changes by stage?

What mistakes should I avoid?

• Running ABM from marketing alone. Give every account an owner in sales (Step 1).
• Treating every account the same. Match effort to value with tiers (Step 2).
• Selling to one contact. Map and reach the buying group (Step 3).
• Leading with a demo. Offer a working session the buyer keeps something from (Step 4).
• Judging the program on engagement. Report opportunities and pipeline against a comparison group (Step 6).