A founder of an early-stage SaaS company called me on a Wednesday last spring, asking how to build an ABM strategy from zero. He had $4M in ARR, eight reps, no marketing platform, and a board meeting in seven weeks where he had to commit to a number.

He’d already read the long-form ABM playbooks online. He had a 47-page guide open in one tab and a Demandbase quote in another. He told me he was paralyzed.

The longer guides aren’t wrong. They’re written for a B2B SaaS company that already has a working marketing function. For a company starting at zero, most of what they recommend is overhead you can’t yet afford.

What follows is what I’d actually do in his position. A 90-day path to a running ABM motion in a B2B SaaS company without a CMO, without an enterprise platform, and without time to build a council. The steps are small on purpose. The leverage is in the sequence.

Day 1 to 14: lock the math

Before any list, before any tool, before any meeting that has the word “alignment” in it, I’d ask the founder three questions and not move past day fourteen until they’re answered.

What’s the deal you most want to win, and what does it look like? Industry, headcount, tech stack, buying-committee shape. Not “any company that pays us.” A specific shape of company. If the founder has closed four deals that fit that shape and three that don’t, you have your data. If you’ve closed eleven deals that all look completely different, you don’t have an ICP yet, and ABM is premature.

What’s the average ACV of that ideal deal? If it’s under $10K, ABM probably isn’t the right motion. The math doesn’t work. ABM costs money to run, and the unit economics of the motion start making sense around $25K ACV and start working cleanly around $50K. SaaS companies with sub-$10K ACVs are usually better served by a strong product-led motion with selective ABM on the largest accounts only.

What’s the sales cycle? If it’s under 30 days, ABM is also probably not the right motion. The pre-sale orchestration time is longer than the deal cycle. ABM math works best at 60 to 270 day cycles, and breaks at the extremes.

If the answers don’t add up, the right answer is to not build an ABM motion. I’ve sent founders away. It’s the most useful thing I do some weeks. ABM done in the wrong context doesn’t underperform; it produces zero, and the zero costs $200K.

If the answers check out, you have your starting point. Write it down on one page. ICP shape, ACV target, cycle length, the deal you most want to win. That page is your North Star. Every decision in the next eighty days should reference it.

Day 15 to 30: the named account list

The most consequential output of the first month is a named account list of 50 accounts. Not 500. Not 5,000. Fifty.

Fifty is not a research target. It’s an attention budget. If your full sales team is eight reps, and each rep can hold roughly six to twelve named accounts in their head at any given time, fifty is the working ceiling for the company. Anything above it dilutes attention to the point where the program can’t be told apart from regular outbound.

Build the list in three steps.

Pull the candidate pool. Use Apollo, Clay, ZoomInfo, or even a manual LinkedIn search. Filter to companies that match the ICP page from the first two weeks. Cap the candidate pool at 200. Anything more is wasted research time.

Score the candidates. Three signals matter at this stage. Fit, signal, and reachability. Fit is the ICP match strength. Signal is whether the account is doing something you can attach to: hiring, recent funding, a product launch, a leadership change, intent on a relevant topic. Reachability is whether you have, or can get, a credible path in. Warm intro, existing relationship, mutual customer, conference contact, recent inbound interest. Score each candidate 1 to 5 on each. Sort.

Take the top 50. Walk through them with sales. Cut anything sales pushes back on with reason. Add anything they’re already chasing. Have the founder sign off on the final list.

That list is the program. Everything else is in service of running plays against it.

Day 30 to 45: the data layer

Most ABM playbooks tell you to set up the platform here. I’d skip the platform for now.

A 50-account list does not need 6sense or Demandbase to run well. It needs a clean spreadsheet, a working CRM, and a small set of free or low-cost tools that handle the work the platform would do at 10x the spend.

What you actually need.

A spreadsheet (or HubSpot/Salesforce list view) that holds all 50 accounts, the buying committee contacts you’ve identified at each, and the engagement state. Refresh it weekly.

A signal-monitoring layer. LinkedIn alerts on each named account. Google Alerts on the company name plus key phrases. Free or low-tier intent data from Bombora or Cognism if budget allows. A weekly sweep of company news and earnings.

A way to identify expanding contacts. UserGems, Champify, or just a quarterly LinkedIn sweep of the buying committee for job changes. The fastest path into a named account is often a champion who just joined.

A reverse-IP tool on your website. RB2B, Leadfeeder, or Clearbit Reveal. When a named account visits your pricing page, you should know within 24 hours.

The total cost of this stack is somewhere between zero and $1,500 a month for a small team. It does 80% of what an enterprise platform does, and you’ll know within six months whether the program produces enough lift to justify upgrading.

The discipline that matters in this phase is not the tools. It’s the weekly sweep. Once a week, somebody on the team has to look at every account on the list, update its engagement state, log new signals, and flag accounts that are surging. If that meeting doesn’t happen, the data layer is decoration.

Day 45 to 60: three plays, no more

The second-biggest mistake I see at this stage. Founders who decide to launch eight different account-based plays at once. Personalized direct mail, custom landing pages, gifting, paid ads, LinkedIn ABM, BDR sequences, executive dinners, a podcast pitch.

Pick three.

The three I’d usually pick at the small B2B SaaS stage:

A surge play. When a named account triggers a meaningful signal (intent surge, web visit on a key page, a buying-committee LinkedIn engagement), an SDR runs a researched, signal-referenced sequence within 24 hours. The trigger is the play. Without the trigger, the play doesn’t run.

A buying-committee expansion play. Marketing identifies the full buying committee at each named account using LinkedIn and Cognism or similar. The SDR’s outbound expands beyond a single champion to two or three additional roles. This is where most ABM programs at this size produce the biggest single bump in pipeline. Most companies are talking to one person at the account. ABM means talking to three.

A founder play, for the top 10 accounts only. The founder sends a personalized message via email and LinkedIn. Not a pitch. A specific, researched note tied to something happening at the account. Quarterly cadence at most. The conversion rate on this is high because the message is real, and it’s the one play AI hasn’t commoditized yet.

Each play has one owner. Each play has one trigger. Each play has one measurable outcome. Document each on a single page.

Three plays, run consistently, will outperform ten plays launched and abandoned. The path that works is boring on purpose.

Day 60 to 75: the cadence

The pod meeting is what holds the program together. At the early stage, it’s small. Founder, head of sales, head of marketing or marketing lead, and one SDR. Forty-five minutes, weekly.

The agenda doesn’t change.

Surge accounts this week. Which named accounts triggered a meaningful signal? What play was run? What happened?

Buying committee gaps. Which accounts are still single-threaded? What’s the plan to widen?

Stuck accounts. Which accounts have been on the list 60+ days with no engagement? Cut, demote, or escalate.

Pipeline impact. New opportunities sourced from named accounts this week. Movement on existing opportunities.

Friction. What broke this week. What needs to be fixed.

That meeting is the program. Skip it for two weeks and the program drifts. Skip it for a month and you’re back to general outbound with extra steps.

Day 75 to 90: report and adjust

By day ninety, you should have enough motion to put a real slide in front of the board. Not a strategy slide. A results slide.

What goes on it.

The named list, with engagement state and surge counts.

Pipeline opportunities sourced from the list to date, with a baseline-to-current arrow.

Buying-committee coverage on the top 20 accounts. How many seats are you actively engaging.

Two leading indicators trending in the right direction. SDR meetings booked from named accounts. Inbound from named accounts. Time-from-surge-to-meeting.

One honest piece of “what we got wrong this quarter.” This is the slide most teams skip and it’s the one that earns the program credibility. The board does not believe a program that has no failures in 90 days. Tell them what didn’t work and what you’re doing about it.

The board’s job is to decide whether to fund another quarter. Your job is to give them enough to decide. A program that produces three good slides at day ninety is in a much stronger position than a program that produces one extraordinary slide and seven hand-wavy ones.

What I’d skip in year one

A few things that look obvious in the playbooks and that I’d skip at this stage.

A six-figure ABM platform. Not yet. Fifty accounts don’t need it. Buy it when you’re past 200 active named accounts and the spreadsheet is breaking.

An ABM agency. Maybe in year two. In year one, the founder and the sales team need to learn the motion themselves. Outsourcing it before you understand it produces a vendor relationship that won’t survive any pressure.

Account-based advertising. Almost never works at 50 accounts on a small budget. Spend the ad budget on retargeting your inbound traffic and finding more of it.

A tiered list with 1:1, 1:few, and 1:many. At 50 accounts, it’s all 1:few or 1:many. The taxonomy slows the team down without adding clarity.

A separate ABM dashboard. The named list view in your CRM is enough. Build the BI dashboard when you have a year of data and a real reporting need.

The shape of the year

If the first ninety days produce real motion (a working list, three plays running, weekly pod, two real opportunities sourced), the next nine months is volume.

Add ten accounts to the list each quarter. Add one new play each quarter, only when one of the existing three is running cleanly. Add a CSM seat to the pod when retention becomes a real metric. Upgrade the data layer when the spreadsheet breaks. Hire a marketing operator dedicated to ABM when the volume justifies it.

By the end of year one, the program should be running 75 to 100 named accounts, four or five plays, a weekly cadence, and reporting both leading and lagging indicators to the board.

If you’re not there at the end of year one, the answer is not to add more. The answer is to subtract until what’s left is working.

The teams that win at ABM at this stage are not the ones with the most. They’re the ones with the smallest motion that actually runs.

That founder I mentioned at the top is on month seven now. He’s at 60 named accounts, three plays, and a 38% lift in named-account pipeline against the baseline. He’s not famous for it. He’s not on a podcast about it. He’s running it.

That’s the version that works.