Account Based Marketing Campaigns: How to Plan, Run and Measure Them

A targeted campaign narrows the audience. An account based campaign changes the unit of measurement. In a targeted campaign you are still counting leads, you have just chosen better leads. In an account based campaign the account is the thing that succeeds or fails, and an individual response is only interesting insofar as it moves the account forward. That distinction sounds academic until you try to build a report, at which point it changes everything.

The practical consequence is that you plan around buying groups rather than individuals. If four people at a target company need to be reached and you have reached one, the campaign is a quarter done, not done. This is why account based campaigns feel slower and produce fewer vanity metrics. You are deliberately declining to count the easy responses that do not move a decision.

The second consequence is that the message is built for the account's situation rather than for a persona. Persona messaging says what a head of operations generally cares about. Account based messaging says what this company's head of operations is likely dealing with given their recent expansion, their new facility, their published commitments, or the regulatory change affecting their sector. That specificity is the entire value proposition and it is what most programmes quietly skip.

The third consequence is coordination. Marketing and sales work the same accounts at the same time with a shared view of who has been touched and how. Without that, an account receives a nurture email, a cold call and a LinkedIn message from three different people in the same week with three different framings, which reads as chaos rather than as attention. Our account-based marketing service is structured around that coordination problem because it is where most in-house attempts come apart.

The new logo campaign targets accounts you have never sold to and where no relationship exists. This is the hardest type, requires the most research per account, and has the longest window. Expect to run it over a quarter rather than a month, and expect the first measurable outcome to be a meeting with one member of the buying group rather than an opportunity. The value of the type is that these are usually the accounts worth the most.

The expansion campaign targets existing customers where you have a relationship in one part of the business and want another. These convert far better than new logo campaigns because you have proof, references and an internal advocate, and they are systematically under-run because they are less exciting than hunting. If your customer base has multiple divisions, geographies or product needs, this is the highest return campaign type available to you and it is usually sitting untouched.

The revival campaign targets accounts that went cold, whether a lost deal, a stalled evaluation or a lapsed customer. The message here is different in kind because you are not introducing yourself, you are explaining what has changed since. Something must genuinely have changed, either in your product, in their situation, or in the market, and campaigns of this type fail when the only change is that a quarter has passed and you need pipeline.

The event campaign is organised around a moment rather than a segment. A conference, a trade show, a regulatory deadline, an industry milestone. The moment supplies a reason to make contact that does not require you to manufacture urgency, which is why event campaigns often outperform their effort. We build these around industry events because a date on the calendar does more for a campaign than any subject line ever will.

An annual target account list and a campaign account list are different objects and conflating them is a common mistake. The annual list is your strategic set, possibly a few hundred names. A campaign list is a subset selected because they share a situation the campaign speaks to, and it is usually much smaller. Twenty to fifty accounts is a workable campaign, because that is roughly how many a small team can genuinely personalise for within a quarter.

Select on a shared trigger rather than a shared attribute where you can. All manufacturers with over five hundred employees is an attribute. All manufacturers who have announced a new European facility in the last nine months is a trigger, and a trigger gives you something to say. The campaign writes itself when the selection criterion is already the reason for the outreach.

Where triggers are unavailable, select on shared pain rather than shared demographics. Companies who have recently posted three or more roles in the same function are signalling something about capacity. Companies who have changed a relevant leader are signalling something about direction. Public hiring and leadership data is imperfect but it is directional, and it is more useful than firmographics alone.

Get the sales team to cut the list before it is finalised. They will know which accounts have a live relationship you would damage by cold contact, which have a history that makes the campaign's premise wrong, and which are already in a process. A campaign list that the sales team helped build is a campaign list they will work, and a list handed to them fully formed is a list they will politely ignore.

B2B decisions are made by groups, and the group is larger than most campaign plans assume. Gartner's work on the B2B buying journey describes a process in which multiple stakeholders each pursue their own information, and where the difficulty is less about persuading one person than about getting a set of people to agree. A campaign that reaches one name in an account has not really reached the account.

For each target account, identify four roles rather than four people. The economic buyer who controls the budget. The technical or functional evaluator who will judge whether the thing works. The user or operational owner who will live with the consequences. And the blocker, usually procurement, legal, IT security or finance, who can stop the process for reasons unrelated to whether your product is good.

Then decide what each role needs to hear, because they do not need to hear the same thing. The economic buyer needs the commercial case. The evaluator needs enough substance to form a technical opinion. The operational owner needs to believe this will make their job better rather than worse. The blocker needs their objection anticipated before they have to raise it. Sending all four the same message is the most common reason ABM campaigns produce polite non-answers.

Buying groups are also where the digital ceiling shows up. Gartner has separately found that a majority of B2B buyers report preferring a rep-free buying experience, which means the people you most need to influence are the least likely to accept a cold meeting. Reaching them often requires a route that is not a sequence, which we come to later.

Write the campaign thesis first, in one sentence, before any copy. Something of the form: companies in this situation are facing this specific problem, which usually shows up as this symptom, and here is what changes if it is solved. If you cannot write that sentence without vagueness, the campaign is not ready and no amount of clever copy will rescue it.

Then build three layers of personalisation. The campaign layer is shared by every account and carries the thesis. The account layer is two or three sentences specific to that company, drawn from something observable and recent. The person layer is a single line reflecting what this role is likely to care about. This structure keeps personalisation sustainable at fifty accounts, which fully bespoke writing is not.

Be disciplined about what counts as account-level personalisation. Referencing a company's recent funding round or new office is a fact, not an insight, and buyers have seen enough automated versions of it that the technique now signals automation rather than attention. The version that works connects the observable fact to a consequence they are probably experiencing, which requires you to actually think about their business for two minutes.

Finally, write the offer, and make it proportionate to the relationship. A first-touch offer of a strategic partnership discussion is not credible from a stranger. An offer to share what you have seen three comparable companies do about a specific problem is credible, costs the recipient fifteen minutes, and gives them a reason to reply that is not about buying anything.

McKinsey's research into B2B buying has consistently found that buyers now move across in-person, remote and digital channels rather than staying in one, and that omnichannel engagement has become the leading approach to B2B sales rather than a differentiator. A campaign that lives entirely in email is competing for one narrow slice of attention against everyone else who made the same decision.

A functional ABM campaign usually runs four channels. Email carries the substance and the repeatability. LinkedIn builds familiarity so that the email is not arriving from a complete stranger. Phone creates the conversations that written channels cannot start. And physical presence, whether a visit or an event, closes the gap with the people who never respond to anything digital.

The channels should not run independently. If someone has viewed a LinkedIn profile, the call script should acknowledge it. If someone has replied to an email, the LinkedIn sequence should stop rather than continuing to send connection prompts. If a call reached a gatekeeper, the email should be timed to arrive when the name has been mentioned. This coordination is unglamorous operational work and it is most of what separates a real campaign from a set of parallel activities.

The number of touches required is higher than most plans allow. A campaign that gives up after four touches across two channels has tested very little. Twelve to eighteen touches across four channels over eight to twelve weeks is closer to what a serious ABM campaign involves, spaced so it reads as persistence rather than pestering.

Email in an account based campaign behaves differently from broad outbound. Volume is low, so the marginal cost of writing well is small, and the tolerance for a generic message is close to zero. These are accounts you have deliberately chosen, which means a lazy email does more damage than no email, because you have spent a first impression on a name you cannot easily replace.

Keep the technical hygiene the same as any outbound programme even at low volume. Authentication, secondary domains and verified addresses still matter, because a bounce or a spam complaint against a target account is disproportionately expensive. The mechanics of running this properly are the same ones covered in our cold email outreach work, just applied to a much smaller list.

Set expectations from published benchmarks rather than hope. HubSpot's email benchmark data provides a reasonable frame for what typical performance looks like, and it is worth remembering that open rates have become a less reliable signal since privacy features started inflating them. In an ABM campaign the meaningful numbers are reply rate and accounts engaged, and with fifty accounts those numbers are small enough that you should read them individually rather than in aggregate.

Multi-thread deliberately rather than accidentally. Emailing four people at the same company on the same day with near-identical messages is noticed and it looks automated. Staggering them over two weeks with genuinely different angles for each role reads as a company that has done its homework, which is the impression the whole campaign is trying to create.

LinkedIn's role in an ABM campaign is not primarily to start conversations. It is to make you a name the recipient has seen before, so that the email and the call land differently. Profile views, thoughtful engagement with what the person has published, and a connection request without an immediate pitch attached, all do more for the campaign than a message sequence fired at a cold contact.

Where messaging is used, the sequence should be short and the ask should be small. LinkedIn's own marketing solutions guidance covers the targeting mechanics, but the behavioural point is simpler: the platform is social, and messages that read like emails get treated like emails. Our LinkedIn outreach work treats the channel as a relationship layer rather than a second inbox for the same sequence.

Paid, where budget allows, has one specific job in an ABM campaign, which is to make sure that when someone in a target account searches for you or lands on your site, the experience is coherent with the campaign message. Account-targeted advertising is expensive per impression and cheap per outcome only when it is reinforcing an active campaign rather than running as a standalone activity.

Be realistic about what the familiarity layer can achieve on its own. It raises the response rate of the channels that actually ask for something. It does not, by itself, produce meetings, and campaigns that over-invest here tend to report high engagement and low pipeline, which is a comfortable place to be and a bad one.

The telephone remains the fastest way to find out whether a target account has a live problem, and it is systematically under-used because it is uncomfortable and because it does not scale. In an account based campaign neither of those objections holds, since the list is small by design and the effort per account is supposed to be high.

Calling into an ABM list is a different discipline from volume dialling. Research from The Bridge Group on sales development shows how activity-heavy the standard model is, but that model is built for breadth. In an account campaign the useful behaviour is fewer calls with more preparation, where the caller knows the company's situation well enough to have a conversation rather than deliver a script.

Gatekeepers should be treated as part of the buying group rather than an obstacle to route around. They know who owns the problem, they know when that person is reachable, and they remember how they were spoken to. A campaign that treats the person answering the phone as an information source rather than a barrier gets a materially better hit rate on the second attempt.

The phone is also the channel that reveals when a campaign's premise is wrong. Two weeks of calls will tell you whether the thesis you built the campaign on actually matches what these companies are experiencing, and that feedback arrives far faster than email metrics. Our cold calling programmes are often run early in a campaign for exactly that diagnostic reason.

Every digital channel in an ABM campaign is competing for attention with every other vendor running the same playbook against the same accounts. Physical presence is not, because almost nobody does it. Being in the building, attending the event, sitting in the meeting rather than appearing in a video window, is the one move in an account based campaign that your competitors have structurally decided not to make.

This matters most at the two ends of the process. Early, because a visit or an event conversation reaches the senior people who never respond to a sequence, and the buying group members you most need are usually exactly those people. Late, because deals stall for reasons that do not surface over email, and a person who can turn up and ask what is actually holding this up recovers opportunities that would otherwise quietly expire.

Practically this means designing the campaign with a physical touch built in rather than bolted on. Which accounts justify a visit. Which event in the next two quarters concentrates the most target accounts in one place. Who is going, what they are going to ask, and what happens in the two weeks afterwards. Our on-ground sales rep service exists because most companies want this and have no way to staff it, particularly when their buyers are in a different country.

The objection is always cost, and the answer is always arithmetic. If your average contract value is large enough to justify an ABM campaign at all, it is almost certainly large enough to justify a small number of visits to the accounts that would move your year. If it is not, you should be running broad B2B lead generation rather than ABM, and the campaign design question answers itself.

A workable structure runs roughly as follows. Weeks one and two are the familiarity layer only: profile views, engagement, connection requests, no asks. This costs almost nothing and materially changes how the next phase is received, which is why skipping it is a false economy.

Weeks three to five introduce the substance. The first email to the primary contact carrying the campaign thesis, a differently angled email to a second role, the first calls, and a short LinkedIn message to anyone who connected. The goal in this phase is not a meeting, it is a reply of any kind that tells you whether the premise landed.

Weeks six to eight escalate and multi-thread. Additional buying group members are contacted, the message shifts from introduction to specific value, calls become more frequent to accounts showing any signal, and the in-person plays are scheduled against the accounts that have engaged. Accounts with zero response across all channels by week eight should be moved out rather than pushed harder.

After week eight, split the list three ways. Engaged accounts move to sales-led pursuit with the campaign providing air cover. Partially engaged accounts move to a slower nurture with a revisit scheduled. Unresponsive accounts go back to the pool for a different campaign with a different thesis, because the failure was probably the premise rather than the account.

The first breakage is definitional. Marketing considers an account engaged when someone opened three emails and visited the pricing page. Sales considers it engaged when a human said something. Both definitions are defensible and the campaign cannot function while both are in use, so pick one before the campaign launches and write it down.

The second breakage is timing. A reply that sits for three days loses most of its value, and in a fifty-account campaign there is no excuse for that latency. Agree a response window, agree who covers it when the owner is away, and treat a missed window as a campaign failure rather than an individual one. Salesforce's State of Sales research has repeatedly documented how much of a seller's week is consumed by non-selling activity, which is usually the real reason replies go stale.

The third breakage is visibility. If sales cannot see what marketing sent to an account, they will open a conversation that contradicts it, and the account will notice. A shared record of every touch by account, visible to everyone working the campaign, is a minimum requirement rather than a sophistication.

The fourth breakage is post-campaign. A campaign ends, the accounts sit in a state that is neither pursued nor closed, and three months later nobody remembers what was said to whom. Building the end state into the campaign plan, including who owns each account afterwards, prevents an expensive campaign from evaporating the week it finishes.

Measure at the account level or you are not measuring ABM. The primary metric is accounts engaged, meaning accounts where at least one member of the buying group has had a real interaction. With a fifty-account campaign this is a number you can read directly rather than a rate, which is an advantage rather than a limitation.

The secondary metric is buying group penetration, meaning how many distinct roles inside each engaged account you have reached. This is the metric that predicts whether an opportunity will survive contact with procurement, and it is the one almost nobody tracks. An account where you have reached one enthusiastic person is considerably more fragile than it looks in a pipeline report.

Then the commercial metrics: opportunities created, pipeline value, and eventually closed revenue, all attributed at account level. Accept that these lag, and resist the temptation to judge a campaign on them at week eight. Judge weeks one to eight on engagement and penetration, and hold the commercial judgement until the sales cycle has had time to run.

Track the negative signals too, because they are diagnostic rather than embarrassing. Accounts that engaged and then went silent tell you where the message stops working. Accounts that never engaged at all tell you the selection criteria or the thesis was wrong. A campaign report that contains only the good numbers cannot teach you anything for the next one.

Too many accounts is the most common. A campaign of two hundred accounts cannot be personalised by a team of three, so the personalisation degrades into merge fields and the campaign becomes broad outbound with a smaller list and a bigger budget. If you cannot name the reason each account is on the list, the list is too long.

Single-threading is the second. One enthusiastic contact feels like progress and produces most of the false optimism in ABM pipeline reporting. That person can leave, lose influence, or simply fail to persuade their colleagues, and the campaign has no other route in. Penetration of the buying group is not a refinement, it is the thing that makes the opportunity real.

Treating ABM as a marketing programme is the third. If the campaign is planned by marketing, executed by marketing and reported by marketing, sales will treat its output as somebody else's leads. Joint ownership of the list, the message and the number is what makes the difference, and it has to be established before launch because it cannot be retrofitted mid-campaign.

Stopping too early is the fourth. Account based campaigns produce very little visible signal in the first month, which is uncomfortable when the previous campaign produced lots of low-quality replies quickly. Judging an eight-week campaign at week three and switching approach is the most reliable way to get nothing from either approach, and it is worth agreeing the review point before the discomfort arrives.