ABM marketing tools: the six categories, the trial that produces a decision, and what to buy first

ABM tools fall into six categories, and most teams buy them in the wrong order. The categories are account data, buying signals, advertising reach, orchestration, sales alerting and measurement. The order that works is to prove the programme with a CRM, a target account list and one advertising platform, then buy signals once you have more accounts than your team can read manually, then buy orchestration once several channels are running at the same time. A platform bought before there is a list and an owner becomes an expensive reporting layer over work nobody is doing.

The six categories, and what each one answers

Category	The question it answers	What to test in a trial
Account data and enrichment	Who are these companies, how big, which locations, which subsidiaries	Coverage and accuracy against fifty accounts you already know well
Buying signals and intent	Which target accounts are researching this category now	Whether the accounts it flags match the ones your sales team already feels are warm
Advertising and reach	Can I get a message in front of the buying group at a named company	Deliverable audience size at your real account list, before you commit budget
Orchestration	Can ads, email, web and sales activity follow one sequence per account	Whether it writes back into your CRM without a consultant
Sales alerting	What should a rep do today, at which account, and why	Whether a rep can act on an alert without opening another tool
Measurement	Which programmes moved accounts towards pipeline	Whether the report shows account progression rather than lead counts

The minimum stack that actually runs a programme

For a first ABM programme covering twenty to fifty accounts, four things are enough. Your CRM, holding the target account list as a real object rather than a spreadsheet. A written buying group map per account, which is a research task rather than a software purchase. One advertising platform able to target by company and role. And a weekly meeting where marketing and sales look at the same list and decide what happens next at each account. That stack costs very little and it exposes the only question that matters early, which is whether anybody is doing the account research.

LinkedIn is the usual advertising layer at this stage. Its own ad targeting page, checked on 12 October 2026, names company targeting, job title, industry, seniority and professional interests, plus matched audiences covering website, contact and account targeting, with contact targeting done by securely uploading a list of contact identifiers or connecting a contact management platform, and retargeting by action taken. The reference is LinkedIn's ad targeting page. Between company targeting and an uploaded account list, that covers the reach job for a programme of this size.

For the programme design that sits above the tools, including how to choose accounts and what to report, see our page on account based marketing in practice.

Where the platforms fit, described from their own documentation

The large ABM platforms bundle several of the six categories, which is their appeal and also the reason they are hard to compare. Two examples, described as each vendor describes itself, checked on 12 October 2026.

6sense, on its own site, says it captures buying activity across channels including keyword level intent, resolves identity at the buying group level and de-anonymises website visitors, uses per customer models to predict buying stage and identify who is involved in a decision, targets advertising using live buying intelligence rather than static lists, and delivers that intelligence into its own applications and third party tools through APIs and data warehouses while running workflows across ads, email, web and sales from one canvas.

Demandbase, on its own site, describes itself as a pipeline engine that turns signals into coordinated action across the go to market team. It names modules for marketing, advertising, sales intelligence and data, and features including website visitor lists with engagement scores, campaign influence reporting with stages labelled Reached, Visited, Interacted and Opportunity, and account hierarchy views it calls Family Tree.

Read those two descriptions side by side and the pattern is clear. Both sell the same promise with a different centre of gravity, one leaning on predictive models of the buying stage and the other on coordinating the whole go to market team around account signals. That is the real shortlist question: which half of your problem is worse, knowing which accounts are in market or getting everybody to act on the ones you already know about.

How to run a trial that produces a decision

• Bring your own list. Fifty accounts you know well, including five you are certain are in market and five you are certain are not. A platform that cannot separate those two groups on your own data is not going to do it on the rest.
• Check the match rate before the features. Ask what percentage of your fifty accounts the tool can resolve, and how many of the named contacts are current. Everything downstream depends on this number.
• Test the write back. Have the vendor demonstrate a signal arriving in your CRM, on the account record, in a field your reps already look at. If that needs professional services, price the services into the comparison.
• Put one rep in front of it for a week and ask them one question afterwards: did this change what you did on any day. A no from the rep is the most reliable signal in the whole evaluation.
• Price it against the list, not the seats. Ask how the price changes if your account list doubles, and what happens to your data at the end of the contract.
• Agree the one report you will run monthly before you sign, and have the vendor build it during the trial.

The data question underneath every tool in this category

All six categories depend on the same foundation: accounts identified consistently in one place. If your CRM holds three records for the same company with different spellings, intent data will attach to the wrong one, advertising audiences will be short, and measurement will double count. So the first week of any ABM tooling project is deduplication and a naming standard for accounts, including how you treat subsidiaries and parent companies. This is unglamorous and it decides whether the rest works. Account hierarchy features exist in the platforms precisely because this problem is universal.

The second foundation is a definition of a buying group that your team agrees on: which roles have to be engaged for an opportunity to be real. Write it as a list of job functions per deal type. That list becomes your advertising targeting, your contact gap report and your definition of account coverage, and it is free.

ABM tool questions that come up in a shortlist

Do we need an ABM platform if we already have a CRM and LinkedIn?

Not for a first programme. Buy a platform when the manual version is working and the constraint has become volume: more target accounts than your team can research, more channels than one person can sequence, or a sales team asking for signals you cannot produce by hand.

Is intent data accurate enough to act on?

Treat it as a prioritisation signal rather than a fact about one person. The practical test is the one above: run it against accounts whose status you already know and see whether the ranking matches reality. Where it does, use it to decide the order of outreach, and keep the message based on your own research rather than on the signal.

What does ABM tooling cost?

Pricing in this category is quoted per account list size, per module and per seat, and vendors publish very little of it, so get written quotes against an identical scope. Include the implementation cost, the integration work and any annual uplift clause, because the licence is often the smaller half of year one.

Can we run ABM advertising without a platform?

Yes. Company and job title targeting on LinkedIn, an uploaded account list as a matched audience, and retargeting of the pages your target accounts visit will cover a programme of fifty accounts. The platforms earn their place when you need the same sequence running across several channels at hundreds of accounts.

How do we measure whether the stack is worth it?

Three numbers, quarterly: the share of target accounts with the full buying group identified, the share showing engagement from more than one role, and pipeline created from the list against pipeline created outside it. If the tools are working, the first two rise before the third does.

The research visit that leaves no trace in any of these tools

Every category above is built on the same assumption, which is that the buying group leaves evidence: a page visit, a content download, a search captured by an intent provider. Part of the research now leaves none of that. A procurement lead or a technical evaluator describes the requirement to an assistant, asks which vendors handle it, and reads a short answer naming two or three companies with a sentence each. No visit, no form, no signal in the stack, and the shortlist may be set before your first email goes out. The useful response is to find out what that answer currently says in your category, which is the job AI Knows Us does: it runs the buying questions, records which vendors get named, and shows the sources behind each answer so you can work on the ones doing the describing.