ABM Is Being Rebuilt Around Buying Signals

Your account-based program is not underperforming because the target list is too small. It is underperforming because the list is static and the buying is not. The fix is a signal-triggered account motion: identify the company, read the live intent, and tier treatment by signal strength instead of funding every named account equally.

Start with the most uncomfortable number in B2B right now. In 6sense's 2025 B2B Buyer Experience Report, a global study of nearly 4,000 B2B buyers published in November 2025, the winning vendor was already on the buyer's Day One shortlist 95% of the time. That is vendor-published research, and it is worth treating as a behavioral read rather than gospel, but the direction is consistent with everything a six-figure deal cycle looks like from the inside.

Read what that means for your budget. Intent signals do not create demand. They tell you that a decision you are already part of, or already excluded from, has started moving. Signals are a timing instrument. Most ABM programs spend like they are a demand generator.

Your target account list is already out of date

The list ages faster than the buying committee changes. In 6sense's 2023 and 2024 research, buying groups ran to 10+ people, and the 2025 report puts median purchase value between $200,000 and $300,000 with group size unchanged this year. Buyers evaluated 5.1 vendors on average, up from 4.5 the year before.

A committee that size reshuffles constantly. Someone gets promoted, a new VP arrives with a vendor preference, finance joins late and reopens the shortlist. Your list was built last quarter against a contact map that no longer exists. You keep paying to reach titles that have moved.

The cycle is also compressing. Average buying cycles dropped from 11.3 months in 2024 to 10.1 months in 2025, and the point of first contact with sellers moved from 69% to 61% of the journey, roughly six to seven weeks earlier. Part of that compression is sample mix: the report notes physical goods purchases in APAC rose from 20% to 32% of the sample, at a 7.3 month average cycle, pulling the global average down. The bigger driver of earlier contact is AI inside the products being bought. Nearly 90% of buyers say AI features are part of the solutions they acquired, and 58% said needing to evaluate a vendor's AI implementation pushed them to engage sooner. Economic pressure is the other half: 62% of buyers said it drove them to engage sellers sooner.

Earlier contact is good news only if you are watching. If your program runs on a quarterly list refresh, six to seven weeks of movement happens entirely in the dark.

The three signals that actually predict a live deal

Not all intent is intent. Three patterns reliably precede a real cycle. The rest is noise dressed up as a dashboard.

1. Repeat pricing and solution page visits from the same company. One visit is research. Three visits to pricing in ten days is a committee building a business case. The repeat is the signal, not the pageview.

2. Multiple contacts from one domain inside a short window. This is the single highest-value signal in a six or seven-figure cycle, because it is the committee assembling in public. Three people from one company touching three different pages in two weeks is one account in motion, not three coincidences. Most teams score contacts individually and never see it.

3. Research behavior that shows up in AI assistants. 94% of buyers used large language models somewhere in the buying journey, with use peaking in the middle rather than at discovery, and the top use case being comparison of vendor offerings. If you are not present in those comparisons, you are being removed from shortlists in a room you cannot see.

And the signal that does not predict anything useful: raw traffic volume to your site.

What AI assistants changed, and what they did not

The panic says LLMs are eating vendor traffic and taking ABM with it. The data is calmer than the panic. In 6sense's follow-up analysis published in November 2025, whether a buyer used LLMs explained less than 2% of the variability in buying group size, point of first contact, vendors evaluated, cycle length, or number of interactions. Buyers still reported 16 interactions per person with the winning vendor, one fewer than the prior year and statistically flat against 2023.

In-market buyers still come to your site, and they still interact at double-digit volume. Here is the operational consequence nobody states. If your aggregate traffic falls while your buying-committee visits hold, every signal threshold you tuned to volume starts misfiring. Score concentration, not count. Three visits from one domain in a quiet week is a stronger signal than it was when your traffic was twice as high.

This is also why AI search visibility is now an account-based concern rather than a content-team concern. Buyers are comparing vendors inside assistants in the middle of the journey, which is exactly when your sales team thinks the deal is going well. Answer Engine Optimization, being the source those engines cite, is how you stay in the comparison. You cannot assert your position there by feel. It has to be measured.

Tier treatment by signal strength, not by list membership

Equal spend across a named list is the most common way ABM budget evaporates. Sequence it instead.

Tier 3, no signal. Named, fit-qualified, silent. Treatment: low-cost presence and content that earns shortlist inclusion before the cycle starts. No sales touch. Given that most winners are on the shortlist from Day One, this is the tier where the deal is actually won, long before anyone is ready to talk.

Tier 2, early signal. One or two contacts, non-pricing pages, repeat visits over weeks. Treatment: targeted paid against that account, relevant content, a nurture sequence. Still no cold outreach. Unsolicited SDR and BDR outreach plays a minimal role in when buyers make first contact. They deliberately ignore it until they are ready.

Tier 1, live signal. Multiple contacts, pricing or solution pages, compressed into days. Treatment: everything. Account-specific paid, a named owner reaching out with a reason that references the actual research pattern, executive air cover, a tailored proposal path.

Yes, paid advertising belongs in this motion. We run it. But account-level paid is sequenced after identification and signal strength, not sprayed across a list of logos because they made it into a spreadsheet. The budget follows the evidence.

The prerequisite most teams skip

Two things decide whether any of this works, and both get skipped.

Clean company-level identification. Company resolution is noisy. ISP traffic and remote-work addresses resolve badly, and a dirty match rate produces confident nonsense that sales learns to ignore within a month. Good looks like this: high-confidence matches only, ISP and generic-carrier traffic discarded rather than guessed at, and your own employees, vendors and existing customers filtered out of the signal set entirely. A smaller, trustworthy signal feed beats a large one that burns sales credibility.

One named owner for the signal-to-outreach handoff. Every article in this category says act on signals. Almost none assigns the handoff. Name one person, define the response window in hours rather than days, and write the routing rule down. Without an owner, Tier 1 signals sit in a dashboard until the window closes.

Measure pipeline per account, not reach inside the list

Target-list impressions and reach are reporting artifacts. They tell you the media ran. They tell you nothing about whether the motion works.

Measure three things:

• Qualified pipeline per target account. The only number that justifies the program.
• Win rate by segment. Which industries, sizes and signal patterns actually convert. This is where you find the tier worth doubling.
• Shortlist inclusion. Whether you are present in comparison research before the first conversation. 97% of buyers had prior experience with at least one vendor on their shortlist, and buyers average eight to nine prior purchase journeys in the same category. Familiarity is a measurable asset, not a soft one.

The honest version of ABM in 2026 is narrower and faster than the version most teams are running. Fewer accounts, better identification, tiering by evidence, one owner, and a visibility position inside the AI assistants where the shortlist gets built.

That is the work we do: mapping where your signals are actually coming from, where your AI search visibility stands, and which accounts deserve budget this month. If you want a read on your own account data before anyone spends a dollar, that is where we start.

Frequently asked questions

Do you run paid advertising as part of account-based marketing?

Yes, and it is sequenced rather than sprayed. Account-level paid spend goes to accounts showing live signals, repeat pricing visits or multiple contacts from one domain, while silent named accounts get low-cost presence instead of full-funnel media. The budget follows the evidence, not the logo list.

Can you review our account data before we commit to anything?

That is the right first step. Before any spend, the useful work is auditing company-level identification quality, checking whether your signal thresholds still hold as traffic patterns shift, and measuring where you currently stand in AI search visibility. Those three answers decide what the program should cost.

Is intent data still worth paying for if buyers decide early?

Yes, but for a different job than most teams assume. Research from 6sense in 2025 found the winning vendor was already on the buyer's Day One shortlist 95% of the time, so signals are not creating demand. They tell you a decision you are already part of has started moving, which is a timing instrument worth having.

Are AI assistants killing the traffic our ABM program relies on?

Not in the way most people fear. 6sense's 2025 analysis found whether a buyer used LLMs explained less than 2% of the variability in cycle length, buying group size or number of interactions, and in-market buyers still averaged 16 interactions with the winning vendor. The practical risk is that falling aggregate traffic makes volume-based signal thresholds misfire, so score concentration instead of count.

How many people are actually on the buying committee now?

The 2025 6sense buyer research put buying groups at more than ten members on roughly a $250,000 average deal size, with buyers evaluating 5.1 vendors on average, up from 4.5. That churn is why a static contact map ages quickly. Watching multiple contacts from one domain is more reliable than maintaining a named-person list.