The statistics behind account-based marketing read like they belong to companies with six-figure platform budgets. G2's roundup of account-based marketing statistics reports companies running ABM see 60 per cent higher success rates, but most proof points come from enterprises with dedicated ABM teams.
The gap between the method and the platform layer is wider than the vendors imply. Purpose-built ABM platforms run USD 50,000 to USD 250,000 a year while most startups hold USD 0 to 15,000 for ABM tooling, per Abmatic's tooling guide for startups, a vendor selling ABM software to that segment. In between sits a program a small team can run: 20 winnable accounts, each treated as a market of its own, drafted with Gemini and routed through Pipedrive.
One boundary so you land on the right guide. This post covers choosing a small set of named accounts and running per-account campaigns without an ABM platform. Routing incoming leads is a separate job, covered in our guide to AI lead routing for sales and marketing teams, and list building at scale has its own LinkedIn lead generation guide. Here: which 20 accounts, and what each gets that a bulk sequence never could.
What account-based marketing asks of a small B2B team
Account-based marketing means treating a named company as a market of one: researching it, tailoring messages to its situation, and measuring whether that account moves. That is the definition in Salesforce's account-based marketing guide, kept without the enterprise machinery.
The method makes four demands, and none of them is software: a short list of named accounts, account-level knowledge, messaging written for the account, and measurement at the account level. Knowledge is the steepest, because most B2B deals involve six to ten stakeholders, as Demandbase's account identification guide notes, so one contact per account is never enough.
The real cost is attention. HubSpot's account-based marketing guide, from a company selling marketing software, works from one marketer to up to ten salespeople, each managing ten accounts. B2BLead's account selection guide puts one-to-one ABM at 20 to 50 accounts a year for two or three marketers plus an SDR.
Why 20 accounts beats 2,000 names in a list
A list of 2,000 names rewards templates, because nobody can write 2,000 researched first lines. A smaller set, worked deeply, lets each message name the account's actual situation.
The capacity maths is blunt. An hour a day across four weeks is twenty hours a month, an hour per account when twenty are in play, enough to research, draft and review. Two thousand accounts at that depth would need eighty working days a month.
Abmatic's target account selection framework argues small teams can only afford 30 to 200 accounts rather than thousands, with a fit-score gate of 70 and above keeping the list honest. Tomba's lean ABM playbook, from an email finder vendor with a stake in tool-lite setups, estimates a spreadsheet, an email finder, LinkedIn and a sequencer deliver about 80 per cent of an ABM platform's value. The scarce resource is attention per account.
The 20-account filter: which accounts earn a personalised campaign
The filter has two halves, fit and winnability. Fit says the account looks like your best customers. Winnability says you have a realistic path in, because an account that recently renewed with an entrenched competitor is a beautiful file that will not move.
For fit, work backwards from revenue rather than a persona slide. B2BLead's account selection guide recommends building the ideal customer profile from your top 20 to 30 customers, analysed by annual contract value (the average yearly worth of a customer), retention and expansion revenue. Score candidates against the industry, size band, region and trigger those customers share, keeping only those that clear the line.
For winnability, three tests. Is one extra win worth the effort? Prospeo's small-business ABM playbook, from a vendor selling prospecting data, advises skipping ABM when the average deal sits under USD 5,000. Are there several people to reach? One decision maker and no champion makes it overkill. And is there a reason why now: a hiring spree, a funding round, a new market push, anything concrete?
DemandScience's tiering guide shows call intelligence company Invoca tiering 4,500 accounts by fit and intent. Small teams run the same idea smaller: top 20, plus 10 reserves.
Building the account file: firmographics, signals and the gaps you will fill
Each of the 20 accounts gets one file, which is what makes per-account outreach possible without an agency: firmographics (industry, size, region, the tools it appears to run), the stakeholder map, the signal that put the account on the list, the use case, and the current state of the relationship.
Keep the files where the work happens. In the Pipedrive setup we use, each account is an organisation record with contacts underneath, and custom fields carry the signal, the use case and the draft status, so a reply moves the account into a sales pipeline without retyping.
A generic worked example: a 40-person logistics software vendor hiring a partnerships manager is about to build a channel it never had. The use case is partner onboarding; the stakeholder map starts with the head of partnerships and the founder; the first message opens with the hire and the problem it implies.
The file also lists its gaps: names you do not know, emails you cannot guess. It is complete when a stranger could draft a credible first message from it without guessing. Judge completeness by that test.
Writing per-account outreach with Gemini: the draft and review loop
Drafting is where Gemini earns its place: sixty or more first messages across twenty accounts and their stakeholders, which is why teams return to templates. The loop: feed the account file to Gemini, get a draft, review it like a human, and send nothing you would not sign.
A working prompt looks like this, in plain words: "Here is what we know about [company]. They are hiring a [role], which suggests [situation]. We help companies like them [outcome]. Draft three versions of a first email to [person], under 120 words, opening with their situation, one specific conversation, no jargon, no claims we cannot back." Gemini drafts in the browser at Gemini's free web app, or inside Gmail with the Help me write button, which Google documents for business and personal accounts and keeps improving, per its May 2026 Workspace updates post.
The review keeps outreach honest. Check names, roles and facts against the file, because Gemini will confidently preserve a typo you fed it. Check claims, and keep the ask small: a reply or a 15-minute call. A human approves every message before it sends.
Personalisation depth scales with the file: the economic buyer hears about payback, the operator about implementation, the stack guard about integration. That stays separate from list-scale personalisation, covered in our cold email workflow guide.
Filling the gaps: Hunter.io enrichment and Pipedrive routing
Enrichment fills the gaps the file flagged. Hunter is an email finding and verification platform: name the person and the domain, and it returns the address it has on file or the pattern the company uses, with verification on the addresses it is confident about. Its pricing page shows a free tier of 50 credits a month, with paid plans from USD 34 a month billed yearly.
SalesHandy's 2026 review of email finder tools, from a company that sells email outreach software, reports 8 out of 10 such tools charge full price for emails that bounce, based on testing 15 tools across 95,000 cold emails. An unverified list quietly burns your sender reputation.
Routing is what Pipedrive does. Each account file lives as an organisation with its contacts, and Pipedrive's knowledge base documents the loop: Campaigns for templated sends, Web Visitors for spotting the companies most engaged with your site, and hundreds of integrations across the plans. Teams wanting sequence mechanics in a dedicated email tool can sync filter-defined contacts around the clock, the pattern Outfunnel, which builds Pipedrive-to-email sync, describes.
The logic is the same everywhere: a reply creates a deal and hands the account to the sales owner, silence schedules the next touch, an opt-out stops everything. Our guide to the MQL to SQL handoff in Pipedrive covers the moment a reply becomes a qualified deal.
The four-week rollout: what happens in weeks one to four
The rollout runs four weeks at about an hour a day, with one person owning the list and one approving sends; separate sittings for drafting and approving are what make the review real. The tooling cost is a CRM seat: Pipedrive Lite at AUD 19 per seat per month billed annually, AUD 49 for Growth, per Pipedrive's pricing page.
| Week | Focus for that week | Concrete output | The sign it is working |
|---|---|---|---|
| Week one | Choose the 20 and open a file for each account | A ranked list with 10 reserves, and a firmographic skeleton per account | You can say why each account is on the list without checking notes |
| Week two | Fill the gaps in the files | Named stakeholders with verified emails, and a signal and use case per account | A stranger could draft a credible first message from the file |
| Week three | Draft and send the first round | One researched message per account, approved by a human before it sends | Replies come back and the first meetings appear in the calendar |
| Week four | Follow up, log the results, pick the next wave | Follow-ups to quiet accounts, notes on what worked, and reserve accounts drafted | You know which accounts are in play and which ones to release |
An account that replies moves into the sales pipeline. One that never engages after a full sequence returns to the reserve pool, its slot going to the next candidate. One that engages but does not buy stays on a low-frequency list.
Measuring account-based marketing: meetings, pipeline and accounts in play
With 20 accounts you measure events rather than percentages: replies per account, meetings booked, pipeline value created, and accounts in play, meaning at least one stakeholder engaged by reply, meeting or site visits.
Compare the program against the previous quarter's results for the same hours. Two meetings from 20 targeted accounts can be strong or weak depending on deal size, so pipeline value outweighs the meeting count. If you already score engagement across site and CRM, our lead scoring guide shows how to fold account signals in.
Set the horizon honestly. Demandbase's ABM strategy template, from the company selling the platform layer this method runs without, says significant results typically take 6 to 12 months. Read that as first replies in the opening weeks, first meetings within a quarter, and pipeline compounding across two or three quarters.
Frequently asked questions
How many accounts can a small B2B team realistically work at once?
Twenty to fifty accounts is the working range for one-to-one programs, and twenty is the right start when one person runs it. HubSpot's rule of one salesperson managing up to ten accounts puts twenty at roughly two sellers' attention.
Do we need an enterprise ABM platform to run account-based marketing?
No. A CRM, an email finder and an AI drafting layer cover the mechanics of a 20-account program, the argument Pipedrive's own ABM software page makes. Move to a purpose-built platform when per-account advertising and website personalisation join the mix and the budget exists.
How long does account-based marketing take to show results?
Two to three quarters is the realistic horizon for a first program: first replies in the opening weeks, first meetings inside the first quarter, pipeline value building as follow-ups land.
What does the tooling cost for a 20-account program?
Roughly the price of a CRM seat plus optional enrichment: Pipedrive Lite at AUD 19 per seat per month billed annually, Hunter from USD 34 a month with a 50-credit free tier, Gemini free. The stack stays under a few hundred AUD a month for two seats.
Is cold outreach to named accounts legal?
It depends on where you and the recipient are based. GDPR in Europe, CAN-SPAM in the United States and local anti-spam rules elsewhere treat B2B cold email differently, and the rules change, so check what applies to your list first. What travels well: an honest subject line, a real reply address and an instant opt-out.
When should we move past 20 accounts?
When the first 20 produce replies and meetings and someone has capacity to run more, usually a second person owning part of the list or an SDR taking follow-ups. The next step is tiering, 50 to 200 accounts, the top tier getting the full treatment. Most small teams never need the enterprise platform.
Turn 20 accounts into a pipeline this quarter
The whole method reduces to four moves.
- Pull the last 12 months of won and lost deals and write down what the winners share: industry, size band, the situation before the purchase. That pattern is the filter for the 20.
- Run the arithmetic on one extra win. Average deal value times the two or three extra deals a good 20-account program should surface in a year. If the average deal is worth AUD 15,000, two extra deals is AUD 30,000 a year against a stack costing a few hundred AUD a month. The case writes itself, or the method is wrong for your deal size.
- Name the owner and the approver before week one: the owner keeps files current, the approver reads every message. If they are the same person, review in a different sitting than the drafting.
- Start on a Monday and treat the hour as fixed. The program compounds only while the weekly rhythm holds.
This is the pattern we wire up at Supernodes: the 20-account list, the Gemini review loop with a human approving every send, and the Pipedrive routing that turns a reply into a deal without retyping. If that sounds like the pipeline you want from the next quarter, tell us what your numbers look like and we will map the method onto your stack, with the foundation live in a two-week pilot.