Account-based marketing (ABM) is a B2B strategy in which sales and marketing pick a defined set of high-fit companies and treat each one, or each tight cluster, as a market of its own, with research, messaging and outreach built for that account. Instead of collecting as many leads as possible and sorting them later, you decide which companies are worth winning first and then work your way in to the people inside them.

It tends to go wrong in a predictable way. A team buys an ABM platform, uploads two thousand company names, points the usual nurture emails at them and reports MQLs at the end of the quarter. That is demand generation with a narrower audience and a bigger invoice.

The version that works rests on two decisions no software makes for you: choosing the accounts carefully, and agreeing on them with sales. The ads, the sequences and the intent data all come later, and they only multiply whatever those two decisions got right or wrong.

TL;DR

• ABM is a push motion aimed at named companies whether or not they've raised a hand, while demand generation is a pull motion that waits for interest to show up. Most B2B companies need both, pointed at the same ICP.

• The idea is old. ITSMA gave it a name in 2003, and enrichment, intent data and automation are what now let a small team give a few hundred accounts the attention that used to be reserved for a rep's three biggest logos.

• Run it at three depths: one-to-one for roughly the top 10 to 25 accounts, one-to-few for clusters of 5 to 15 that share a problem, and one-to-many for the hundreds that fit but don't justify bespoke work.

• Draft the target list from your ICP and an account score, then let sales strike and add names before it's final. Every account gets a named owner, and the plan covers the whole buying committee rather than one contact.

• When a target account shows activity (several people on your pricing page, a new sales leader, a known buyer in a new role), you know both fit and timing, so that's where the most coordinated effort goes.

• Settle the list and the owners before you buy a platform, and report on accounts: engagement across roles, pipeline from target accounts and win rate, with the marketing qualified account replacing the MQL.

Account-based marketing vs demand generation

The two motions start at opposite ends. Demand generation publishes, ranks and advertises to a broad audience, waits for buyers to raise a hand, and sorts the leads that come in. Its unit of work is the lead, and its scoreboard is usually lead volume.

Account-based marketing starts with a short list of companies chosen on purpose, works out who inside each one takes part in the decision, and goes to them. It doesn't wait for a form fill, and a target account that never fills one in is still a target account.

Neither replaces the other. Demand gen builds the familiarity that makes your ABM outreach feel less cold, and ABM makes sure your best-fit companies get more than whatever happens to reach them through the content machine. Trouble starts when a team runs one and calls it the other, or when the two point at different ideas of what a good customer looks like.

We've written about why a GTM motion beats a lead-gen motion, and ABM is where that difference is easiest to see. A quick test for your own program: if it would still make sense with the target account list deleted, it's demand gen.

An enterprise sales habit that software made scalable

The best enterprise reps have always worked this way. They'd take three or four big accounts, learn the org chart, read the annual report, find out which project was stuck and who was blocking it, and treat those accounts as their entire territory. It worked, and it was expensive, because a person did every bit of that research by hand.

ITSMA, the research and advisory firm for technology marketers, coined the term account-based marketing in 2003. For years afterwards, ABM mostly meant one-to-one programs for a company's largest strategic customers, because nothing smaller could pay for the effort.

What changed is the cost of research. Enrichment tools can pull headcount, tech stack and recent hires for a few hundred companies in an afternoon, intent data shows which of them are reading about your category, and automation can tailor a first touch per account. That brought the depth of the old key-account habit down to companies that would never have justified a dedicated person.

It also changed who does the work. Marketing brings the reach and the systems, sales brings what it knows about each account, and ABM is the one place the two have to run the same play against the same names.

One-to-one, one-to-few and one-to-many ABM

ABM runs at three depths, and the difference is how many accounts share a single campaign. ZoomInfo's guide to ABM puts the top 10 to 25 accounts in the one-to-one tier and suggests clusters of 5 to 15 accounts for one-to-few, which is close to what we see in practice.

Placing accounts doesn't need a model. If one win would change your quarter, the account goes in one-to-one. If five or ten accounts share a problem you can speak to with one story, they become a one-to-few cluster. Everything else that fits goes in one-to-many, where you keep the accounts warm and watch for any of them to heat up and move up a tier.

Getting it wrong costs you in both directions. One-to-one effort on a programmatic account burns hours you can't spare, and a templated email to a dream account wastes a first impression you don't get back.

Channel rules shape the tiers as well. LinkedIn's company list targeting needs at least 300 rows to upload, and LinkedIn recommends 1,000 or more companies, so a twelve-account cluster can't run as its own ad audience there. In practice, paid media does most of its work at the one-to-many level, while one-to-one and one-to-few lean on email, messages from the rep on LinkedIn, calls, events and the occasional parcel.

Building the target account list with sales in the room

Most ABM programs are won or lost here, before a single touch goes out. A good list is small enough to work with care, agreed by both teams, and grounded in fit rather than in which logos would look good on the website.

Start from your ideal customer profile, not from a revenue target. If you already run an ICP scoring model, its top tier is your first draft, and an account score that layers intent and engagement on top tells you which of those companies to start with.

Then hand the draft to sales before it's final. Reps know things no firmographic field holds: which account has a live project, where a former champion works now, which company burned a previous rep and won't take a call. Let them strike names and add a few strategic exceptions they can justify in one sentence.

Keep it small. Two thousand accounts is a database, and nobody works a database with any care. A few hundred well-chosen, tiered accounts will beat a long list, and if nobody can name the next step for an account, it doesn't belong in one-to-one.

Put a named owner on every company record. Without one, touches get dropped and follow-up stalls, because each team assumes the other has it. Then look at the list again every quarter, since companies grow out of your profile, others grow into it, and closed deals leave it.

Plan for the committee, not the contact

An account is a group of people. In its 2024 buying study, Forrester found that thirteen people, on average, have a hand in a company's purchase decision, and that almost nine purchases in ten cross at least two departments.

So map the buying committee for every one-to-one and one-to-few account before you write anything: who signs, who feels the pain, who will use the product day to day, who can block it. Then reach several of them with messages that fit their seat. Multi-threading is the reason to go account-based at all, because one sequence to one contact inside a big committee is just outbound with more steps.

Our inbound meeting workflow shows one way to do it. When one person books a call, the workflow finds the rest of the likely committee at that company and reaches them on LinkedIn and by email before the meeting takes place.

When a target account starts showing up

People hear "account-based" and picture cold outbound into large companies. That's half the motion, and the harder half, because cold outreach can't tell whether an account is in a buying window or a year away from one.

The warmer half watches your target accounts for activity. A fitting company appears in your website traffic through a reverse IP lookup. Three people from the same account read your pricing and integration pages in one week. A buyer you've sold to before starts as head of sales at a target account. Each of those tells you two things cold outbound never knows: the account fits, and something is happening there now.

That's signal-based selling applied at the account level, and it changes where the effort goes. Instead of working forty named accounts on a calendar, you put the most coordinated work into the four or five that are warming up this month and keep the rest on a lighter rhythm.

Here's how that plays out in a forty-account, one-to-few program for a Series B software company with logistics and wholesale as two of its clusters. A logistics account shows a fit and intent match. The rep gets an alert saying who visited and what they read, the champion and the economic buyer each get an email written around the logistics problem, the rep sends a note on LinkedIn, and a call follows within a few days. If nobody responds, the account drops back into the cluster's normal rhythm until the next signal.

Coordinating those touches is campaign orchestration, and it only works inside a wider GTM system where the signal, the CRM record and the sequences are connected. It's also why inbound-led outbound fits so well inside ABM. Inbound shows you which of your chosen accounts are moving, and the account-based setup makes sure the whole committee hears from you, not only the person who browsed.

Is it ABM if the platform does it?

Platforms such as Demandbase and 6sense are good at their part of the work: identifying companies, serving ads to them and surfacing intent. They can't pick your accounts or get sales to agree on them, and those are the two decisions that make a program account-based in the first place.

Skip those decisions and a platform produces the familiar result. The list is whatever got uploaded, every target account gets the same nurture email as everyone else, and the report shows MQLs. It's demand gen with an ABM label, and after a year of license fees the team concludes ABM doesn't work.

Teams that do the selection and alignment work tend to rate the approach highly. In the 2022 benchmark from Momentum ITSMA and the ABM Leadership Alliance, 72% of almost 300 ABM leaders and practitioners said ABM delivers higher ROI than other types of marketing. That's the self-reported view of people running ABM programs, published by an alliance that includes ABM vendors, so treat it as what practitioners believe rather than a measured return.

The order we'd follow is list first, owners second, a few plays run by hand third, and the platform last, once you know which plays deserve automating. Plenty of one-to-few programs run well on a CRM, an enrichment tool and a sequencer.

Reporting on accounts instead of leads

The surest way to sink an ABM program is to keep scoring it like demand gen. If the only number in the monthly report is MQLs, the team drifts back toward volume without anyone deciding to, because volume is what gets praised in the meeting.

Move the report up to the account. Four numbers carry most of it:

• Account engagement, meaning how many of the right roles at each target account are interacting with you, rather than how many clicks one person made.

• Pipeline from target accounts, reported on its own line. Pipeline from random inbound is welcome, but it isn't what the program was built to produce.

• Win rate and deal size on target accounts, compared with similar accounts you didn't target over the same period.

• The marketing qualified account in place of the MQL, so qualification happens at the level you're selling to.

Give it time, too. With complex deals, pipeline moves over quarters, so judge the first quarter on engagement and meetings with the right roles, and the second and third on opportunities and wins. A group of fitting accounts you deliberately leave out makes the result much harder to argue with.

In short

Account-based marketing is a decision about where effort goes. You pick a small set of companies worth winning, agree on them with sales, decide how deep to go with each one, and then judge the program on whether those companies turn into pipeline and revenue.

Most of what makes it work happens before any campaign runs: the list, the owners and the map of each buying committee. Tiers, signals and platforms come afterwards, and each of them amplifies the quality of those early choices. When a program stalls, the cause usually sits upstream, in a list built on ambition or a report that still counts forms.

Done properly, it compounds. Each quarter's results sharpen the list, the messaging gets more specific to each cluster, and sales and marketing get used to working from the same names.

When a client asks us to set one up, we build the target list, the account scoring and the signal-led plays in the CRM and outreach tools the client already runs, then hand everything over with a note on why each account and each play is there, ready for their own people to run.

Frequently asked questions