A salesperson once told me her whole year came down to eleven companies. Not eleven thousand leads, not a funnel, eleven logos on a whiteboard, each one worth more than the rest of the pipeline combined. Pointing the usual marketing machine at that list felt absurd: why run ads to a "segment" of eleven? That instinct, stop marketing to a crowd, start marketing to a named company, is the whole idea behind account-based marketing.

The quick version

  • ABM flips the funnel. Instead of casting wide and filtering down to a few buyers, you start with a named list of high-value accounts and build everything around them.
  • It treats one company as a market. Each target account gets its own research, message, and joint sales-and-marketing plan, "a market of one."
  • The evidence is encouraging but vendor-led. Practitioner surveys report higher ROI than other marketing; the strongest numbers come from firms that sell ABM software, so read them as directional.
  • It only pays off when accounts are few, large, and complex. ABM is expensive per account. Use it where deals are big and decided by a committee, not where you need volume.

The idea in depth

The phrase was coined around 2003 by Bev Burgess and colleagues at the Information Technology Services Marketing Association (ITSMA), the B2B marketing body that first gave the practice a name and a method (see the history of the term). Their definition is the one still worth memorising: treat individual accounts as markets in their own right. A market of one.

That sounds like a slogan until you sit with what it rejects. Classic demand generation is a numbers game shaped like a funnel: attract a large audience at the top, qualify the curious, and convert the survivors at the bottom. It works beautifully when your buyer is plentiful and self-selecting. ABM inverts the geometry. You begin with the accounts you have decided are worth winning, then work outward into the specific humans inside them. Sangram Vajre, who built the #FlipMyFunnel movement around this, frames it as turning the funnel upside down: identify the accounts first, then engage everyone who matters inside them. In practice that means one concrete thing, before you write a single campaign, agree a target list with sales and put names, not personas, on it.

flowchart TB
  subgraph TRAD["Traditional demand gen"]
    direction TB
    A1(["Wide audience"]) --> A2(["Leads"]) --> A3(["Qualified"]) --> A4(["A few buyers"])
  end
  subgraph ABMF["Account-based marketing"]
    direction TB
    B1(["Named target accounts"]) --> B2(["Map the buying group"]) --> B3(["Tailored engagement"]) --> B4(["Grow the account"])
  end
ABM starts where the funnel ends, with the named accounts, and works outward. Leaders Loop

Why the committee, not the contact, is the unit

The reason ABM exists is that big B2B purchases are not made by a person. Gartner's research on the B2B buying journey finds that a typical complex purchase involves six to ten decision-makers, each arriving with their own information and agenda, looping back and forth rather than marching down a tidy path (Gartner, The B2B Buying Journey). A single well-nurtured lead is close to useless if the four other people who can veto the deal never hear from you.

The practical answer is to map the account as a group before you message it: who feels the problem, who controls the budget, who can block it, who has to live with it afterwards. ABM's job is to make sure the case for you reaches all of them, in language each one cares about, the CFO hears the payback, the IT lead hears the integration story, the end-user hears "this gets easier." That is also why ABM is run as a joint programme: marketing supplies the air cover and the tailored content, sales works the relationships, and they share one account plan rather than two scorecards.

What the evidence actually says

The headline claim you will hear is that ABM delivers higher returns than anything else marketers do. The most-cited version comes from ITSMA's annual benchmarking work (the body now trades as Momentum ABM), where most firms that measure ABM report it outperforming their other marketing on ROI. Forrester's 2024 study found the same pattern across regions: most decision-makers it surveyed reported ABM programmes returning 21–50% higher ROI than non-ABM efforts, and roughly a quarter reported gains of 51% or more (Forrester, December 2024).

Treat individual accounts as markets in their own right.

The honest limitation: almost all of this evidence is self-reported and a good deal of it is produced by, or commissioned with, companies that sell ABM platforms. Self-reported ROI is generous by nature, nobody volunteers that their flagship programme is underwater, and there is little independent, peer-reviewed work isolating ABM's effect from the simple fact that firms point ABM at their best accounts, which were always more likely to convert. Read the numbers as directional, not as proof. The defensible claim is narrower and sturdier: focusing scarce sales-and-marketing effort on a small set of high-value accounts tends to beat spreading the same effort thin. The economics underneath are old. Acquiring a new customer can cost several times more than keeping and growing an existing one, Harvard Business Review's Amy Gallo puts the often-quoted range at 5 to 25 times, depending on industry and study, in a piece that also draws on Bain & Company's retention research. ABM is, in large part, a disciplined way of pointing your effort at the accounts where that math is most favourable.

A worked example

Picture a mid-sized software company that sells logistics tools. Last year it generated four thousand leads and closed thirty deals, almost all of them small. Three enterprise accounts it really wanted, a national retailer, a freight carrier, a hospital group, never moved, because each is a committee of eight people and the company was only ever talking to one of them.

It runs an ABM pilot. The numbers below are illustrative, drawn to show the shape of the trade-off, not a real case. Marketing and sales agree the three accounts and spend a fortnight on each: who the buying group is, what each member's pain looks like, what would make them say no. For the retailer they build a tailored micro-site, a short ROI model using the retailer's own store count, and a roundtable invitation aimed at the operations VP and her two deputies. Sales works the relationships in parallel, off the same plan.

flowchart LR
  S(["Select 3 accounts"]) --> M(["Map each buying group"])
  M --> C(["Build tailored content
per account"])
  C --> E(["Sales + marketing
engage together"])
  E --> R(["Review: pipeline,
meetings, deal velocity"])
  R -.->|"keep / cut / add accounts"| S
An ABM pilot is a loop: pick, map, tailor, engage, review, then re-pick. Leaders Loop

Six months in, the pilot has produced fewer "leads" than a normal campaign, a handful of meetings rather than hundreds of downloads, but two of the three committees are now actively evaluating, and the retailer deal alone is worth more than last year's thirty combined. That is the ABM trade-off in miniature: less volume, more concentration, and a cost per account that only makes sense because the accounts are large. Run the same playbook against three hundred small accounts and it would simply bankrupt the marketing budget. As we cover in marketing strategy & STP, the prior question is always whether your market is shaped for concentration or for reach.

Frequently asked questions

How is ABM different from just good key-account selling?

It overlaps heavily, strong salespeople have always treated big accounts as their own little markets. What ABM adds is marketing as a deliberate partner: tailored content, campaigns, and "air cover" built for one account's buying group, planned jointly rather than sales asking marketing for a brochure. The novelty is organisational, not conceptual.

How many accounts should we target?

It depends on how much tailoring you can sustain. "One-to-one" ABM (a bespoke programme per account) might cover a handful; "one-to-few" clusters similar accounts into small groups; "one-to-many" applies lighter personalisation to a few hundred using software. Start narrow. A pilot of three to ten accounts you can genuinely tailor for beats a list of a hundred you can only spam.

Isn't ABM just expensive marketing for a few logos?

It can be, if you pick the wrong accounts or never tie it to revenue. ABM earns its cost only where deals are large, buying groups are complex, and the named accounts are genuinely winnable. For a high-volume, low-price product, traditional demand generation is usually the smarter spend.

How do we measure it when there are no leads to count?

Change the scoreboard. Lead volume is the wrong metric for a programme built on a dozen accounts. Track account-level signals instead: how many of the target buying group you are reaching, meeting velocity, pipeline created within the named accounts, and deal progression, not form fills.

Whose job is ABM, marketing or sales?

Both, on one plan, or it fails. The most common reason ABM disappoints is that marketing runs "account-based campaigns" sales never asked for and never works. Shared target list, shared account plan, shared definition of success. Treat that alignment as the precondition, not a nice-to-have.

Related in the Toolkit

Where to go next