Most account-based marketing advice is written for enterprise teams with a dedicated ABM manager, an intent data contract and a six-figure ad budget. A SaaS company between $1M and $15M ARR has none of that, but it often has a sharper reason to do ABM: a small number of accounts that are worth far more than the average deal. This is the playbook I use at that stage. It is deliberately lean. It covers how I choose accounts, how I tier them, the plays I run per tier, how I size the budget and what I report. Paid channels are only one part of it; if you want the paid media side on its own, that is covered on my B2B paid advertising page.
Step 1: Build the account list from closed-won data, not ambition
The most common ABM mistake I see is a target list made of logos the founder admires. Start instead with your last 12 to 24 months of closed-won deals and look for what the best customers share: industry, employee range, region, the tool they replaced, the trigger that started the deal and the deal size. Then look at closed-lost and churned accounts for the patterns you want to avoid. The output is a written ICP with firmographic filters you can actually run against a database, not adjectives. My ICP and buyer persona template walks through that exercise. Once the filters exist, pull every account that matches them from your CRM and a data provider. At this ARR range a workable first list is usually a few hundred accounts, small enough that sales can review every name. Ask the account executives to remove accounts with an open opportunity they own, current customers (those belong in an expansion motion) and anything they know is a poor fit. A list sales has edited is a list sales will work.
Step 2: Tier the accounts so effort matches value
I use three tiers, and the definitions are about effort, not just fit. Tier 1 is a short list of accounts where a single deal would materially change the year. These get one-to-one research, a named owner in sales and marketing, and custom content. Tier 2 is a larger group that shares an industry or use case, so you can build one-to-few plays per cluster. Tier 3 is everything else on the list: programmatic, lightly personalised, mostly to build familiarity before a sales touch. If you run HubSpot, it ships two company properties for this: Target account, a checkbox, and Ideal customer profile tier, with Tier 1 to Tier 3 values you can rename. HubSpot's knowledge base notes that these ABM tools sit in Marketing Hub or Sales Hub Professional and Enterprise. In Salesforce or Zoho CRM, create the same two fields on the Account object. Whatever the tool, the tier must live on the company record so every report can be cut by it. Review tiers each quarter and move accounts based on engagement and new triggers such as funding, leadership hires or a competitor contract coming up for renewal.
Step 3: The plays I run per tier
For Tier 1, the core play is a researched account plan: the buying group mapped by role, the business problem in their own words (from earnings calls, job posts or interviews), and a sequence where the account executive, an executive sponsor and marketing each have a defined touch. Marketing's job here is assets: a short point of view written for that account, a relevant case study, and an invitation to something small such as a roundtable or a working session. For Tier 2, build one cluster play per industry or use case: a landing page and a guide written for that segment, targeted LinkedIn ads to the cluster, and an SDR sequence that references the segment problem rather than the individual. For Tier 3, run awareness ads and nurture email to known contacts, and let engagement promote accounts into Tier 2. Every play needs an owner, a start date and an exit rule. A play without an exit rule runs forever and nobody can say whether it worked. Keep the plays few. Three plays executed well beat ten plays executed once.
Step 4: Sizing the budget without guessing
I do not have a universal ABM budget number for you, and I am wary of anyone who does, because the right figure depends on deal size and how many accounts you can actually work. I size it bottom up. Start with the number of accounts per tier, then cost each play: hours of research and content for Tier 1, ad spend and content per cluster for Tier 2, and ad plus tooling spend for Tier 3. Check the result against the value of the pipeline you expect, using your own historical win rate and average deal size. One platform constraint shapes the paid line: LinkedIn's help centre states the minimum audience for an ad set is 300 member accounts and suggests a minimum of 50,000 to drive results. A tight Tier 1 list will often fall below that, so for Tier 1 I lean on one-to-one outreach and content rather than ads, and use LinkedIn matched audiences for Tier 2 clusters and Tier 3 where audiences are large enough. HubSpot's ABM tools can sync target accounts or specific ICP tiers to a LinkedIn audience if you have the ads account connected.
Step 5: Metrics that tell you whether ABM is working
Lead-based metrics break in ABM because you are deliberately working fewer accounts. I report at account level, by tier, in four layers. Coverage: how many target accounts have at least one known contact in each key buying role. Engagement: how many target accounts showed meaningful activity this month, such as multiple contacts visiting high-intent pages, replying to outreach or attending an event. Pipeline: opportunities created and pipeline value from target accounts versus non-target accounts. Outcomes: win rate, deal size and sales cycle length for target accounts versus your baseline. The comparison with non-target accounts is the point. ABM is working if target accounts convert to pipeline at a higher rate, close larger or close faster than similar accounts you did not target. Give it at least two of your normal sales cycles before judging outcomes, and judge coverage and engagement monthly in the meantime. If coverage stays low, the problem is data, not the plays.
Where ABM goes wrong at this stage
Three failure patterns show up repeatedly in the companies I work with. First, ABM is launched as a marketing campaign without sales agreeing to work the list. Fix it with a written agreement on which accounts each rep owns and what they will do in the first 30 days, the same way I structure a sales and marketing SLA. Second, the team buys an intent data or ABM platform before the account list and tiers exist. Tooling amplifies a strategy; it does not create one. Many teams at this ARR range can run the first two quarters on their CRM, LinkedIn and a spreadsheet. Third, results are judged on lead volume, which falls by design, and the program is cancelled before a full sales cycle has passed. Agree the metrics in Step 5 before launch so nobody is surprised when MQL counts drop. If the basics are not in place, especially clean company records and owner assignment, fix those first. ABM on a messy CRM produces messy reporting and an argument about whether it worked.
Sources
HubSpot Knowledge Base, Get started with account-based marketing in HubSpot: https://knowledge.hubspot.com/branding/get-started-with-account-based-marketing-in-hubspot LinkedIn Marketing Solutions Help, Target audience size best practices: https://www.linkedin.com/help/lms/answer/a423690 LinkedIn Marketing Solutions Help, Matched Audiences best practices: https://www.linkedin.com/help/lms/answer/106247
FAQ
Start with as many as sales can genuinely review and work. In practice that is usually a few hundred accounts across all tiers, with a short Tier 1 list of accounts that would change the year. If reps cannot name what they will do with an account in the next 30 days, the list is too long.
No. Build the account list, tiers and plays first using your CRM, LinkedIn and a data provider. Tools amplify a strategy that already works. Buy one when you can name the specific job it does that your current stack cannot, such as account-level ad targeting at scale.
Paid advertising is one channel inside ABM. ABM is the operating model: an agreed list of accounts, tiers, coordinated sales and marketing plays, and account-level measurement. Ads mostly support Tier 2 and Tier 3, while Tier 1 relies on one-to-one research and outreach.
Coverage and engagement should move within the first month or two. Pipeline and win-rate comparisons need at least two of your normal sales cycles, because you are comparing target accounts against a baseline and need enough closed deals to make that comparison fair.