Context-driven prospecting: what replaced the firmographic filter in 2026

TL;DR

Firmographic ICP filters still eliminate non-fit accounts. They no longer predict buying intent. Context-driven prospecting is what fills the gap.

• 70% of the B2B buyer journey is anonymous (6sense 2025); firmographic filters cannot see it.
• Seven signal categories replace static firmographics as the primary context: leadership changes, product launches, hiring intensity, tech-stack moves, funding, event participation, regulatory exposure.
• Agent-driven research compresses per-account context time from 10-15 minutes to under 2 minutes.
• Teams running context-driven prospecting see 2 to 4x higher connect rates and meeting conversion.
• The ICP of 2026 is a filter plus a context layer, not a filter alone.

Introduction

The modern B2B SaaS ICP document is an archaeological artefact. It names a company size range, two or three verticals, three or four personas, a budget band and a region. It stays frozen on an internal wiki page while the actual target accounts move in and out of buying windows at a cadence the document was never designed to capture.

The document is not wrong; it is incomplete. The ICP of 2026 is not a static filter. It is a filter plus a context layer that updates weekly. The filter eliminates the 90% of accounts that will never buy. The context layer prioritises the 10% of filtered accounts that are in market right now.

This article is for the CMO, head of growth or RevOps lead who is tired of the quarterly ICP review that produces no measurable change in pipeline. It covers the seven signal categories that have replaced firmographic filters as the primary prospecting lens, the agent-driven research pipeline that produces context at scale, and the implementation sequence that works in mid-market B2B.

Why firmographic filters stopped predicting intent

Three structural shifts between 2022 and 2025 degraded the predictive power of firmographic filtering.

The first shift is buyer anonymity. 6sense’s B2B Buyer Experience Report 2025 shows 70% of the buyer journey is anonymous and 81% of buyers have decided before contacting sales. Firmographic filtering sees whether an account matches your profile; it cannot see whether that account is currently anonymous-researching your category. The signal-rich moment of research is invisible to firmographics.

The second shift is buying-cycle fragmentation. The average B2B SaaS deal in 2026 touches 10.1 stakeholders across functions. Different functions trigger buying at different moments. A security tool is triggered by a new VP of Engineering hire, a trade show, a compliance deadline, a competitor breach. None of these triggers show up in a firmographic filter.

The third shift is volume economics. Modern outbound at scale requires precision. Firmographic-only targeting treats the top 100 and bottom 100 accounts in a 10,000-account list identically. Teams that cannot prioritise within their filtered list burn their TAM in 8 to 12 weeks and leave with brand damage and no remaining audience, as we covered in the AI SDR failure modes piece.

As Jordan Crawford of Blueprint GTM has argued for three years:

“The list is the message.”

Firmographic filters build identical lists for identical ICPs. Context-driven qualification builds different lists for different weeks because the context has moved. The second approach wins.

The seven signal categories

Across our deployments and the published operator benchmarks, seven signal categories reliably predict buying intent for B2B SaaS. The specific signals within each category depend on your product and vertical; the categories apply broadly.

Category one: leadership changes. New hires at VP level or above in the buying function. New CMO = MarTech rebuild window. New VP of Engineering = security and developer-tooling window. New CFO = financial-stack audit window. The signal is public (LinkedIn, press release) and the window of relevance is 60 to 180 days post-hire.

Category two: product launches. The target company launched a new product, feature or vertical in the last 90 days. For infrastructure products (hosting, security, analytics), a product launch creates demand upstream. For enablement products (sales tooling, marketing tooling), a launch creates demand downstream. Window: 30 to 120 days.

Category three: hiring intensity. The target company is hiring aggressively in a specific function. Hiring 10 engineers in 60 days = infrastructure tooling demand. Hiring 5 sales reps = sales tooling demand. The signal is observable through job posting data and reveals internal priority.

Category four: tech-stack moves. The target company adopted or deprecated a specific tool. Switched from Marketo to HubSpot = integration demand. Added a Snowflake deployment = data-tooling demand. BuiltWith, Clearbit and vendor usage databases expose this signal.

Category five: funding events. Series A, B, C close. Grant funding. Debt facility. Funding creates budget cycles and typically triggers a 60 to 90 day investment window in tooling categories. Window: 30 to 120 days post-close.

Category six: event participation. The target company sent speakers or sponsors to a trade show, industry event or conference. The event choice reveals internal priority (a company sponsoring a security conference is signalling priority) and provides a conversation hook. Window: 30 to 60 days around the event.

Category seven: regulatory or market exposure. Tariff changes, EU AI Act exposure, new compliance regimes, major regulatory shifts. For compliance software, a tariff change creates immediate demand for importers and exporters. For AI tooling vendors, EU AI Act Article 50 enforcement creates a demand window. See the EU AI Act audit piece for the regulatory calendar through 2027.

A high-performing context-driven list in 2026 scores each account on three to five of these categories, prioritises by composite score, and refreshes the scoring weekly.

The firmographic filter still matters (but differently)

Context-driven prospecting does not eliminate the firmographic filter. It subordinates it.

The firmographic filter still does the heavy lifting of eliminating non-fit accounts. Below 50 employees, above 5,000 employees, outside Benelux, outside B2B SaaS. The filter removes 90% of the TAM in one query, cheaply.

The remaining 10% is where the firmographic filter stops helping and the context layer starts. Among 1,000 fit accounts, which 100 are in a buying window this quarter? Firmographics cannot answer that. Context-driven signals can.

The practical architecture:

Jen Allen-Knuth of DemandJen frames the shift in buyer-centred terms:

“B2B buyers don’t trust vendors. They trust their peers and other customers.”

The buyer who gets a context-aware opener trusts the rep further because the opener proves the rep understands their current moment. The buyer who gets a firmographic-match opener dismisses the rep as another spray-and-pray outbound call. The context layer is the trust layer.

Agent-driven research is what makes this feasible

The historical constraint on context-driven prospecting was time. A human researcher produces one quality context profile in 10 to 15 minutes. A rep running 60 calls per day cannot do that work. The default fallback was firmographic-only lists with generic openers.

The 2026 constraint has lifted. AI research agents collapse per-account context time to under 2 minutes. A batch of 500 target accounts produces 500 enriched context profiles overnight. The research economics have inverted; what was previously too expensive to do at scale is now the default.

The architectural pattern that works in our Falora deployments is a two-phase pipeline.

Phase one: overnight batch research. For tomorrow’s call list, the agent researches each account against the three to five signal categories most predictive for the product. Output: a 60 to 100-word context briefing per account.

Phase two: real-time research augmentation. During the call, if a specific topic arises (a competitor mention, a timing question), the rep can trigger a 30-second real-time agent lookup that adds context mid-conversation.

The two-phase architecture produces context breadth (every account on the list has a baseline briefing) and context depth (deep research triggers only on live conversations). The economics work at any list size above 100 accounts per week.

This is a specific instance of the broader signal-based selling playbook applied to the prospecting phase.

The weekly prioritisation routine

The implementation that turns context-driven prospecting from a project into a routine is the weekly cadence.

Every Monday morning, the agent produces the top 100 accounts for the week, scored on the three to five signal categories most predictive for the product. The sales team works that list for the week, with daily refreshes if major signals surface mid-week.

The scoring formula is specific to the product and vertical but typically combines:

• Firmographic fit: 0 or 1 (gate)
• Signal count: 1 to 7 (how many categories are active)
• Signal recency: 0 to 1 (how fresh is the strongest signal)
• Signal weight: 0 to 1 (how predictive is this category for this product)

Composite score drives ranking. Top 100 enters the weekly list. The rest cycles for future windows as their signals mature.

Chris Walker’s broader argument from Revenue Vitals applies to the cadence specifically:

“Sales metrics like MQLs and Stage 1 pipeline are grossly misaligned to the actual goal of demand generation.”

The team optimising for MQL volume treats all firmographically-matched accounts equally. The team running weekly context-driven prioritisation treats accounts asymmetrically: the top-scored 100 get full attention, the rest wait. The second team produces fewer MQLs and more pipeline.

What the compound effect looks like

A representative mid-market B2B SaaS team compared under two scenarios.

The underlying firmographic fit is held constant. The delta is the context layer plus the weekly prioritisation routine. The ratio of 2.9 to 14.2 meetings per rep per week compounds through the quarter into the pipeline value gap.

Three failure modes

Failure one: context without filter. Teams add context-driven signals but forget to maintain the firmographic filter. The result is a prospect list where the context looks great but half the accounts are too small to buy, in the wrong geo, or in a vertical that cannot use the product. Keep the filter.

Failure two: too many signal categories. Teams define 15 signal categories and the scoring model becomes unexplainable. Three to five categories, weighted deliberately, is the operational sweet spot.

Failure three: context without openers. Teams build the context pipeline, produce great briefings, and then let reps revert to generic openers because the opener design never caught up. The research is wasted. Pair context investment with 1-minute opener training.

A 60-day implementation plan

Days 1 to 10: signal workshop. Workshop which three to five signal categories most predict buying intent for your product. Document hypotheses with sales, marketing, and customer success.

Days 11 to 25: data pipeline. Stand up the data sources for each category (LinkedIn hiring data, BuiltWith for tech-stack, Crunchbase for funding, press-release scrapers, event-attendee lists).

Days 26 to 40: agent-driven research. Deploy the overnight batch research pipeline. End-state: every Monday the sales team has a scored top-100 list with 60 to 100-word briefings per account.

Days 41 to 55: opener training. Train reps on prospect-centric opener design that uses the briefing. Pair with the 1-minute rule and the connect rate playbook.

Days 56 to 60: measure and iterate. Compare connect rate, one-minute rate, and meeting conversion against the pre-implementation baseline. Expect 2 to 4x on all three metrics within the first 60 days.

Frequently asked questions

Can we run context-driven prospecting without AI agents? Yes, manually, for lists under 50 accounts per week. Above that volume, the research economics force agent adoption.

Which signal categories matter most for enterprise outbound? Leadership changes and funding events dominate for enterprise. For deals above €100K ACV, the buying trigger is typically a strategic change at the top of the organisation, not a tactical operational signal.

Does this work for PLG motions? Yes, with modified signal categories. Product-usage signals (free-tier activation, feature adoption, team-expansion events inside the product) replace or augment the external signals. The architecture is the same.

How often should the signal weights be recalibrated? Quarterly. The predictive power of each signal category shifts as the market shifts, and the quarterly cadence catches drift before it degrades pipeline.

Does Falora provide this capability? Falora is the orchestration layer that consumes signal data from your sources, runs the agent research, produces the weekly top-100 list and feeds it into the outbound execution. The signal data sources (Common Room, Clay, Apollo, Cognism) remain best-of-breed.

Conclusion

The firmographic ICP filter stopped predicting buying intent in 2024. It took most B2B SaaS teams until 2026 to notice. The teams that adopted context-driven prospecting early produce 4 to 8x the pipeline output per rep of teams that are still running quarterly ICP reviews on static filters.

The seven signal categories above are the starting point. The three to five that matter most for your product become your weekly prioritisation input. The agent-driven research pipeline makes the whole thing economically feasible at scale.

If you want help identifying the signal categories that matter for your product and standing up the weekly prioritisation routine, book a 30-minute prospecting diagnostic with Falora.

Sources

• 6sense, B2B Buyer Experience Report 2025
• Jordan Crawford, Blueprint GTM
• Jen Allen-Knuth, DemandJen
• Chris Walker, Passetto / Revenue Vitals
• Gartner, Future of Sales 2026
• Kyle Poyar & Maja Voje, State of B2B GTM 2025
• Falora / Stretch Innovation Internal Benchmark Q2-Q3 2026 (n=11 B2B SaaS deployments, BE/NL)

Related reading on Falora

• The signal-based selling playbook
• The 1-minute cold call rule
• Connect rate: the forgotten lever in B2B outbound
• AI SDRs vs human SDRs: 7 conditions where AI loses
• The anatomy of a GTM engineering system

About the author

Stijn Van Daele is co-founder of Falora and a partner at Stretch Innovation. He has deployed context-driven prospecting pipelines in 18+ B2B SaaS scale-ups and writes about GTM engineering, autonomous revenue and prospecting craft on LinkedIn.