The Research Constraint Just Died. Your ABM Strategy Doesn't Know Yet.

Every ABM tier model I've ever seen was a capacity plan wearing a strategy costume.

I've sat in the quarterly meetings where the tiers get debated. Which accounts deserve Tier 1. Whether a hot Tier 2 account should get promoted. Whether Tier 3 should even exist. The conversation always sounds strategic. Account potential, fit scores, propensity models.

It was never about the accounts. It was about the team. Fifty accounts get real research because the team can research fifty accounts. Not because exactly fifty companies deserve it.

I've made this argument before: ABM tiering was a resource constraint dressed up as a strategy. And last month I showed where the constraint actually lives: not in your intent data or your orchestration platform, but in the research and writing work between the signal and the sent email.

This post is about what happens next. Because the constraint that produced tiering is gone, and most ABM strategies are still organized around its corpse.

Tiers are a capacity plan wearing a strategy costume

Strip the language away and here's what a tier model actually says.

We have more accounts than research hours. Real research takes 20 to 40 minutes per account. A good SDR produces 8 to 12 genuinely personalized emails a day, and spends 60 to 70 percent of the week on research to hit even that. So we will pick a small set of accounts to treat like buyers, a bigger set to treat like a mailing list, and a long tail to treat like an audience.

That's it. That's the model. The 50 / 250 / 1,000 structure isn't a theory of how companies buy. It's a description of how much your team can produce, with the shortfall named and laminated.

We didn't tier accounts because buyers come in three flavors. We tiered them because research time was scarce and somebody had to lose. The genius of calling it "strategy" is that nobody has to say the quiet part: the Tier 2 template isn't an outreach motion. It's a confession that the math didn't work.

None of this was wrong, by the way. Given the constraint, rationing was the adult move. The teams that tiered deliberately beat the teams that spread effort evenly and reached nobody. Every good operator I know built some version of this system, including me.

But "correct given the constraint" is doing all the work in that sentence.

Constraints write strategy. Then they die.

Here's the pattern that should make every ABM leader nervous.

Strategy is mostly constraint management. Sales territories were drawn around drive time, until inside sales made geography irrelevant. Media plans were built around scarce ad slots, until digital made inventory effectively infinite. Software shipped on annual release trains, until deployment got cheap and the train schedule became a liability.

In every case, the same thing happened. The constraint died quietly, the strategy built on it lived on loudly, and for a few years the org chart kept solving a problem that no longer existed. The companies that noticed early didn't just get more efficient. They got to play a different game while competitors optimized the old one.

The tell is always the same: a practice that feels like strategy but can't survive the question "would we design it this way today, from scratch?"

Ask that about your tier model. If research were free, would you invent tiers of treatment? Would you decide that 250 in-market companies should receive a template with their name dropped in, on purpose, as a positive choice?

Nobody would. Tiering exists because research wasn't free. That's the whole foundation.

The constraint behind ABM just died

So here's the claim, stated plainly: the research constraint is dead.

The work inside those 20 to 40 minutes was always reading, cross-referencing, and synthesis. What did leadership say about this year's priorities. Who runs the function now, who just joined, who left. What the hiring page says about where the pain is. What shipped last quarter. What the company already uses next to your category.

That is exactly the shape of work AI agents got good at. Not "write me a friendly email" parlor tricks. The unglamorous part: reading everything public about an account and reducing it to a working hypothesis about what that company is trying to fix this quarter. Done properly, with real checking on top, the machine version of that pass is as thorough at account 257 as it was at account 7, doesn't burn out, and doesn't get pulled into pipeline reviews. We run it across 150 to 200 accounts a week, and I wrote up how the architecture works if you want the system view.

The point of this post is not the plumbing. The point is the strategic consequence: when research stops being scarce, every decision you made because it was scarce is up for review.

Here's the shift, at the level that matters:

The constraint era assumed	What's true now
Research time is the scarcest resource in outbound	Research is the most automatable step in outbound
Personalization is a luxury you ration toward big logos	Personalization is the default for the whole book
Tiers decide who gets treated like a buyer	Signals decide who gets attention this week
The book is as big as the team can research	The book is as big as your ICP actually is
SDRs are researchers who occasionally converse	SDRs are conversation owners who never research

One nuance before anyone panics: the constraint that died is research, not judgment. More on that below, because it's the difference between this argument and the "AI SDR" pitch I keep watching fail.

What account-based strategy looks like after it

If treatment no longer needs rationing, what's left of account-based strategy? Three things, and they're better than tiers.

Timing replaces tiering. The useful question was never "how much is this account worth," which barely changes year to year. It's "is this account reachable this month," which changes weekly. A leadership hire in your buying function. A hiring spike on the problem you solve. A funding round, a product launch, an intent spike. Run the book as a queue ordered by live signals, where any account that did something this week moves to the front. Worth decides who's in the book. Behavior decides the order. Tiers decide nothing.

The ICP becomes the real strategy document. When the team could only research fifty accounts, a loose ICP was survivable, because human judgment filtered the list one account at a time. When every account in the book gets researched and touched, the book itself becomes the highest-stakes decision in the motion. Defining the ICP tightly stops being a marketing exercise and becomes the strategy. Garbage in, researched garbage out.

Treatment equalizes upward. The promise ABM platforms sold for a decade was every account treated like it matters. What got delivered was fifty accounts treated like they matter and a newsletter for everyone else, because the execution layer couldn't keep up. Post-constraint, the Tier 1 research brief isn't a scarce artifact anymore. It's just what an account gets for being in the book.

Notice what's not on this list: more volume. Killing the research constraint is not an argument for emailing more companies. It's an argument for treating the same right companies dramatically better. The teams that read "research is free now" as "blast everyone" are about to relearn why spray-and-pray died, at higher speed.

The hard question: does volume kill quality?

The objection I respect most goes like this: if every account gets "personalized" outreach at machine speed, doesn't personalization stop meaning anything? Won't it all converge on the same AI-flavored mush?

Yes, for most teams. And that's exactly where the next competitive line gets drawn.

The constraint didn't disappear. It moved. The scarce thing is no longer research hours. It's the quality bar you hold the machine output to. Whether anything verifies the research before it becomes an email. Whether the claimed company event actually happened. Whether the third touch argues something new or just rephrases the first. Whether the result reads like a person did 40 minutes of homework, or like a merge field with a vocabulary.

I've written about the difference between research-driven personalization and mail merge with extra steps, and the gap between those two is the entire game now. Buyers can't see your stack. They can only see whether the email knows something true and specific about their situation. The teams that win the post-constraint era won't be the ones who adopted AI research first. They'll be the ones who refused to let it ship anything a sharp SDR would be embarrassed to send.

Which is why the right mental model isn't "AI does outbound now." It's "the floor moved." Tier 1 treatment is the new minimum. The ceiling, the judgment about what's worth saying and to whom, is as human as it ever was.

What this means for the SDR role

The standing hot take applies: AI won't replace SDRs. It replaces the parts of the SDR job that SDRs hate.

Nobody got into sales to spend Tuesday cross-referencing hiring pages. Yet that's where 60 to 70 percent of the SDR week actually goes, and it's the part of the job with the least judgment in it. The part that is all judgment, reading a lukewarm reply and deciding whether to push, soften, or step back, knowing when a no is a not-yet, working a skeptical buying committee, never had a capacity problem. It had a time problem, because research ate the week.

Post-constraint, the SDR role gets better, not smaller. The researcher-who-occasionally-converses becomes a conversation owner with a full calendar of the work humans are actually good at. If I were running an SDR team today, my worry wouldn't be headcount. It would be whether my best people have spent so long doing research that nobody developed the conversation skills the role is about to be entirely made of.

The prediction

Here's the falsifiable version of this post.

Within a few years, "we template our Tier 2" will sound the way "we only personalize for enterprise prospects" sounds today. Like a confession. Not a strategy choice you defend in a QBR, but a thing you admit while explaining why mid-market win rates are soft.

Tier language won't vanish. Finance will still want account potential modeled, and that's fine. What disappears is tiered treatment: the idea that the quality of research and outreach an in-market company receives should depend on which bucket a spreadsheet put it in last January.

And the window matters more than the endpoint. Right now, researched-grade outreach to a Tier 2 account is an anomaly. That account's inbox is full of templates, so the email that knows what they're actually trying to fix this quarter stands out absurdly. The advantage exists precisely because most teams haven't noticed the constraint is dead. It shrinks as they do. Early movers get the gap. Late movers get parity, later, at higher cost.

Five questions to ask your team on Monday

Not a checklist. Just the questions that expose whether your motion is built on a live constraint or a dead one.

1. If research were free, would our tier model survive a from-scratch redesign? What would we keep on purpose?
2. What did our Tier 2 accounts actually receive last quarter? Pull five real emails and read them out loud.
3. How many in-market signals did our platform flag last month, and how many got a researched first touch inside the window?
4. Is our ICP definition tight enough to be the strategy, or has tiering been quietly compensating for a loose one?
5. If our SDRs got their research time back tomorrow, do we know what we'd have them do with it?

If those questions are comfortable, your motion is already post-constraint. If they sting, the corpse is still on the org chart.

Frequently asked questions

Why do ABM tiers exist in the first place?

Because human research time was the scarcest resource in outbound. Real account research takes 20 to 40 minutes, and an SDR spending most of the week on it still tops out at 8 to 12 genuinely personalized emails a day. Tiers were the rationing system: a small set of accounts treated like buyers, the rest templated or ignored. The structure described team capacity, not buyer behavior.

Is ABM tiering still relevant in 2026?

For modeling account potential, yes. As a treatment policy, it's a legacy of a constraint that no longer exists. AI agents have made the research pass automatable, so the quality of outreach an account receives no longer has to depend on which bucket it was assigned to in January. Timing, driven by live buying signals, is the prioritization layer that still earns its place.

What replaces tier-based treatment in ABM?

A signal-ordered queue over a tightly defined ICP. Every account in the book gets researched-grade treatment as the default, and live signals like leadership hires, hiring spikes, funding rounds, product launches, and intent spikes decide which accounts get attention this week. Account value decides who's in the book. Behavior decides the order.

Doesn't automating research make every email sound the same?

It does for teams that treat AI as a writing shortcut instead of a research layer. The differentiator is the quality bar: whether claimed facts get verified, whether each touch argues something new, whether the output reads like a person did 40 minutes of homework. Personalization converging on mush is a quality-control failure, not a property of automation.

What happens to SDRs if account research is automated?

The role concentrates into what was always its hardest part: conversations. Reading replies, handling objections, working buying committees, knowing when a no means not-yet. Research consumed 60 to 70 percent of the SDR week and contained the least judgment of anything in the job. Removing it makes the role more senior, not redundant.

Do I still need my ABM or intent platform?

Yes. Identification and orchestration are real problems and the platforms solve them well. The dead constraint was never in the platform layer. It was in the execution layer between the signal and the sent email, which platforms were never built to close.

See it on one of your accounts

The argument above is falsifiable on your own book, this week.

Pick one account your platform says is in market, one that's currently slated for the template. Send me the company name. I'll show you what researched-grade treatment on that exact account looks like, and you can judge whether the gap between that and your current Tier 2 touch is worth reorganizing around. No charge, no obligation.

Our goal at steady state is 5 to 15 meetings per month for a client, and the honest reason that range is wide is that ICP fit and market math vary. We show output and let the output convince.

Book a 30 minute walkthrough and I'll have the account ready when we meet. If you'd rather read first, the Kill Your ABM Tiers playbook covers the re-prioritization framework.

Related reading: 6sense tells you who. Demandbase tells you when. Neither tells you what to say. • ABM tiering was a resource constraint, not a strategy • Marketing has been selling ABM for 10 years • Your SDR is spending 60 percent of their time on research • AI personalization vs mail merge