ABM Orchestration: Lists and Plays Sales Actually Closes

The first ABM motion I ran had 6,000 accounts on the list. By the end of the quarter, the number we actually worked was 41. Sales had quietly muted the shared Slack channel by week three. The Sendoso budget was 78% spent. Pipeline sourced from "ABM" was a single deal that the AE swore he would have closed anyway.

That's the gap most demand gen managers walk into on day one of an ABM motion. Marketing builds a list of every company on the planet that vaguely fits the ICP, sales gets handed a spreadsheet, and three weeks later everyone's back to inbound triage. The reframing that fixes it is small but loud: ABM isn't a list of who could buy. It's a list of who should buy this quarter.

That sentence will save you a year of pain. Read on for how to actually build the list, design plays sales will run, and bring numbers to QBR that hold up.

The 6,000-account ICP that's actually 600

Here's the math nobody wants to do upfront. Your TAM is not your ICP. Your ICP is not your ABM list. Your ABM list is not your active target set. Each layer cuts the previous one by 60-80%.

A real example from a Series B SaaS I worked with (sales-led, $14M ARR, mid-market focus, $40K-$120K ACV):

• TAM: ~6,000 US companies in their three target industries between 200 and 2,000 employees.
• ICP fit (firmographic): ~1,200 after filtering by tech stack (BuiltWith showed they ran a competing point solution), revenue band ($25M-$500M), and geo (US + Canada only because the AEs didn't speak Spanish or Portuguese).
• Showing intent: ~600 after a 6sense surge filter on three priority topics over the previous 90 days.
• In an active buying window: ~150 after layering Clearbit funding signals, LinkedIn Sales Navigator job-change alerts (new VP of Sales, new Director of RevOps), and G2 buyer-intent visits to their category.

Six thousand becomes 150. That's where you start tiering. If your "ABM list" still has 4,000 accounts on it after the cut, you don't have an ABM list. You have a marketing automation segment.

Why most ABM dies in month two

I've seen this pattern enough times to name it. Three failure modes, every time:

1. The list is too big. When sales sees 1,500 accounts in the "ABM target" view, they treat it like a regular territory. They prioritize the names they already know. The actual targeting math you did is invisible to them. List size has to be small enough that an AE can hold every account name in their head. For Tier 1, that's 8-15 accounts per AE. Anything more becomes wallpaper.

2. The plays don't fire. Marketing builds a campaign and ships it on a date. That's not a play, that's a launch. A real ABM play fires off a trigger: new VP of Sales hired, funding announced, a specific intent topic spikes, a competitor's contract is up for renewal. No trigger logic, no play. You're just running broadcast outbound with extra steps.

3. Reps don't run them. This is the one that quietly kills 70% of ABM motions. The play is good. The list is tight. The trigger fires. And then half the AEs don't run the sequence because nobody told their VP this was a quarter-level priority. ABM without QBR-level sales-leadership commitment is theater. We'll come back to this.

ICP filter math — building a real list

Here's the layered filter to run, in order. Each layer is a SQL filter or a saved view in your data tool. Don't skip layers because "we already know our ICP." You don't, until the data confirms it.

Layer 1: Firmographic floor (kills 80%)

The non-negotiables. Revenue band, employee count, geography, industry. Pull from ZoomInfo, Apollo, or Clearbit Reveal. If you sell into RevOps teams, add a "has a RevOps function" filter — LinkedIn Sales Navigator titles will show you which orgs have a Director of RevOps or VP of Operations. If you sell a tool that sits alongside Salesforce or HubSpot, run BuiltWith or Wappalyzer to confirm the tech stack. Companies on Pipedrive don't buy enterprise CPQ tools.

Layer 2: Intent layer (kills another 50%)

This is where you separate "fits the profile" from "actually thinking about buying something." Three sources, layered:

• 6sense or Demandbase surge: account-level intent score on your priority topics over the trailing 30-90 days. Set the threshold high. "Showing some research" is noise. "Surging" is signal.
• G2 buyer intent: if a buying committee at an account is looking at your category page, comparison pages, or a competitor's profile, that's a much warmer signal than a topic search.
• Bombora topics: good for category-level interest, weaker for specific buying intent. Use as a tiebreaker, not a primary filter.

Don't use all three with equal weight. 6sense surge + G2 visits is the strongest combo I've seen. Bombora alone produces a lot of false positives.

Layer 3: Buying-stage signals (kills another 60%)

Now you're looking for change. Companies in steady state buy on their renewal calendar. Companies in change buy on the change.

• Job changes: new VP of Sales, new CRO, new Head of RevOps, new CMO. LinkedIn Sales Navigator alerts catch these in the week they happen. A new VP of Sales will replace 30-50% of their stack in the first 90 days.
• Funding: Series B and beyond. Crunchbase or Clearbit's funding signal. Newly funded companies suddenly have a budget and a mandate to grow.
• M&A: acquired companies get integrated and re-tooled. Acquirers consolidate stacks.
• Leadership turnover at the buying-committee level: Director of Demand Gen leaves, your champion's manager moved on, the CFO replaced the COO. All of these reset the stack conversation.

Layer all three and you'll typically end up at 100-200 active accounts. That's your real ABM list. Now you tier.

Tier 1, 2, 3 — what each tier actually means

The tiering isn't just budget allocation. It's play depth and who runs the play. Get this wrong and you'll either over-invest in accounts that aren't ready or under-invest in the ones that are.

Tier 1 is where you spend your judgment. Tier 2 is where you spend your process. Tier 3 is where you spend your tooling. Don't confuse them.

A common mistake: starting with all three tiers in month one. Don't. Pick 30 Tier 1 accounts, ship one play, prove the motion, then layer in Tier 2 in month four. Tier 3 last, once your data infra is solid.

Play design sales actually runs

A real play has a trigger, a sequence of touches, an owner per touch, a channel per touch, and a clear "done" definition. Here's one in full.

Play: New VP of Sales hired at a Tier 1 account

Trigger: LinkedIn Sales Navigator job-change alert. Verified by SDR within 24 hours (sometimes the alert fires on a lateral move or interim title — sanity check before activating).

Why this play works: A new VP of Sales replaces 30-50% of their stack in their first 90 days. We have 14-day window before they're locked into vendor evaluations. After day 30, they've already made first-pass decisions and you're playing catch-up.

14-day sequence:

That's a play. Trigger, owner per touch, channel per touch, content brief per touch, exit criteria. You can run this. A "VP of Sales sequence" with a vague description in a Notion doc is not a play. It's a note.

One more thing about play design

The biggest mistake I see is plays that require marketing to manually fire each touch. If your play needs you to log into HubSpot every Tuesday and queue up emails, it'll stop running the first week your campaign launch slips. Plays should be sequenced in your engagement platform (Outreach, Salesloft, or your CRM's native sequencer) and triggered automatically off the data signal. Marketing's job is to design and load. Sales' job is to run.

Coordinating AE, SDR, and CSM motion

The play above has three owners. Without a coordination layer they will double-tap, contradict each other, or worst, not show up at all.

What works:

• One Slack channel per Tier 1 account. Sounds heavy until you realize Tier 1 is 30 accounts, so you need 30 channels. Auto-archive after 90 days of inactivity. Each channel has the AE, SDR, ABM-PM, and (if expansion is in play) the CSM.
• Weekly 15-minute ABM standup with sales. Not a 60-minute marketing meeting. Fifteen minutes. Three things only: (1) which Tier 1 accounts triggered a play this week, (2) which plays stalled and why, (3) what creative or content sales needs from marketing for next week.
• Shared Notion or Confluence play library. Every play documented with the table format above. New AEs onboard against it. When a play gets a 6% reply rate, that becomes the working version. When it gets 1%, it gets archived with a post-mortem note. You're building institutional memory, not running campaigns.
• First-touch ownership rule. SDR owns first touch on Tier 2 and Tier 3. AE owns first touch on Tier 1, full stop. If the SDR touches a Tier 1 first, the AE will deprioritize it. Make this rule explicit.

The "play didn't work because reps didn't run it" failure mode

This is the section nobody writes about because it's awkward. Here's how to diagnose it honestly.

When a play underperforms, the reflex is to blame creative. Subject line, video script, exec email tone. Before any of that, pull the activity data:

• How many AEs were assigned the play? (Should be all of them holding accounts in the target set.)
• How many AEs enrolled their accounts in the sequence?
• How many AEs completed the day-3 personal video step? Day-7 follow-up?
• What percentage of touches were actually executed vs scheduled?

If fewer than 60% of reps ran the play to completion, the play isn't broken. Adoption is. Fixing creative when adoption is the problem is malpractice. It wastes another six weeks and burns sales' patience further.

The fix for adoption is upstream of marketing:

1. Confirm the VP of Sales committed to this play in the QBR. If not, you don't have a play, you have a marketing project.
2. Make play execution a metric in the AE's weekly 1:1. Their manager owns this, not you. Your job is to give the manager the dashboard.
3. Make it dead simple to run. If the play requires the AE to write a custom email from scratch, half won't. Pre-load three approved variants in the sequence, let the AE pick and personalize the opener.

When I diagnosed a stalled motion at one company, only 4 of 11 AEs had enrolled their Tier 1 accounts. The play was fine. The CRO had told his team "support the marketing thing" without making it a number on their scorecard. We added "Tier 1 play execution rate" as a weekly metric and adoption went to 91% inside three weeks. Same play, same creative, same trigger logic. Just adoption.

Measurement that holds up in QBR

The CRO doesn't care about open rates. They care whether ABM is making the number bigger or just spending budget. Bring a four-layer scorecard.

Two notes on this:

• Sourced vs influenced pipeline is a religious war. Pick one definition with your CRO and stick with it for at least two quarters. I usually advocate for "ABM influenced" = account had at least one ABM play touch in the 90 days before opp creation. Cleaner than fighting over multi-touch attribution models.
• Win rate is the real number. If your ABM accounts close at the same rate as inbound, you're not running ABM, you're running expensive outbound. The whole point is that the win rate is materially better because you're targeting accounts in active buying windows.

Bring those four numbers to QBR every quarter. Not 47 vanity metrics. Four.

The first 90 days of an ABM motion

If you take one thing from this piece, take this: don't try to launch all three tiers at once.

• Days 1-30: Build the data infra. 6sense or Demandbase contracted, intent topics configured, Sales Navigator alerts wired, Clearbit enrichment running. Cut the list to 30 Tier 1 accounts. Get the CRO to commit in writing (or in the QBR slide) that AEs will run the plays.
• Days 31-60: Ship one play. Just one. The "new VP of Sales" play above is a great starting point because the trigger is clean and the time pressure is real. Run it on the 30 accounts. Track adoption weekly.
• Days 61-90: Diagnose. Pull the activity data. Was it a play problem or an adoption problem? Iterate accordingly. Show the CRO the four-layer scorecard.

After 90 days, if the meeting rate is in the 15-25% band and adoption is above 70%, you have permission to add Tier 2. Don't ask for it before that.

ABM is three things, all owned by the demand gen manager IC: list math, play coordination, and measurement. You don't get to outsource any of them. You don't get to do two of them well and skip the third. The motion only works when all three are running.

That's why most ABM motions fail. And why the ones that work win the year.

Learn More

• A Day in the Life of a Demand Gen Manager
• Your First 30/60/90 Days as a New Demand Gen Manager
• Funnel Math: MQL to SQL to Opportunity Forensics
• Demand Gen Tools and Tech Stack
• Demand Generation Manager JD