TL;DR: Find in-market accounts by scoring fit, signal strength, recency, and source confidence, then require corroboration before high-touch outreach. Sales, Growth, and RevOps teams should use a 100-point confidence score, route accounts above 70, research scores from 45 to 69, and suppress low-fit or unsafe records.

Key Facts and Decision Thresholds

Methodology and Limitations

This framework combines the current Unify signal catalog, the Outbound Sweet Spot methodology, and named customer stories reviewed in September 2026. The score is an operating rubric, not a statistical probability model. Thresholds must be recalibrated against closed-won and disqualified accounts by segment, region, and deal size.

Customer results are individual outcomes, not a universal platform benchmark. First-party observations usually deserve more confidence than modeled third-party intent, but neither creates permission to contact a person. Privacy, consent, suppression, and regional calling rules still apply.

What does in-market mean in B2B?

An in-market B2B account is a qualified company showing recent, credible behavior consistent with an active problem or evaluation. Interest alone is not enough. An account must fit the commercial model, the signal must plausibly relate to the problem, and the timing must support action now.

A pricing-page visit can be strong first-party evidence, yet it may still come from a job seeker, investor, partner, competitor, or existing customer. A funding announcement can change capacity without proving category demand. The scoring model exists to prevent those observations from becoming false certainty.

Which signals indicate active demand rather than general interest?

Active demand appears when the signal is close to the product decision, recent enough to matter, and supported by another independent observation. Product usage, repeated commercial-page visits, a direct evaluation request, and verified category research tend to be more actionable than broad news or one anonymous visit.

• First-party commercial behavior: repeated pricing, security, integration, or implementation-page activity from a qualified account.
• Product usage: adoption, limit, team-invite, or feature behavior that reveals a credible expansion or sales-assist moment.
• Organizational change: a relevant executive hire, team build, or operating-model change connected to the problem you solve.
• Market event: funding, expansion, regulation, or technology adoption that creates capacity or urgency, after fit is confirmed.
• Third-party research: category or competitor investigation that is recent, attributable to the account, and validated by another source.

For a deeper distinction between observed and modeled signals, use the guide to first-party versus third-party intent. To avoid acting after a signal has lost meaning, pair it with the signal-decay framework.

How should you score fit, strength, recency, and confidence?

Score each dimension independently, then sum the result out of 100. Keep fit as a gate: a high-intent event at an account you cannot serve should never become a high-priority lead.

Use 70 to 100 as the action band, 45 to 69 as the research band, and below 45 as nurture or ignore. A score does not override ownership, open-opportunity, customer, suppression, or consent rules.

When should two weak signals beat one strong signal?

Two weak signals should beat one strong signal only when they are independent and jointly explain the same commercial hypothesis. A job posting plus a matching technology change may justify research. Two page views generated by the same anonymous session do not provide independent confirmation.

• Require independence: the signals must come from different behaviors or sources.
• Require coherence: both signals must support one plausible business problem.
• Require recency: each event must still fall inside its useful decision window.
• Require fit: corroboration cannot rescue an account outside the ICP.
• Cap the action: weak-signal stacking may trigger research or a task, not an unsupervised send.

How do you filter bots, competitors, job seekers, and customers?

Filter non-buyers before scoring, not after a sequence starts. Maintain explicit exclusion rules for known bots, data-center traffic, careers-page behavior, competitor domains, partners, employees, current customers, open opportunities, and active support cases.

What action should each score band trigger?

Match the action to confidence and ownership. The highest band gets prompt seller attention, the middle band gets verification, and the lowest band remains out of an active sequence.

• 70 to 100: route to the owner, show the evidence, and prepare a reviewed message within the team SLA.
• 45 to 69: run account and contact research, seek a second signal, and create a task only if the hypothesis survives.
• 20 to 44: keep the account in monitoring or nurture without seller interruption.
• 0 to 19: ignore or suppress until new evidence appears.
• Any compliance or ownership conflict: block outreach regardless of score.

How Unify covers in-market detection and activation

Unify gives every rep outbound agents that combine first-party engagement, third-party data, and AI-discovered signals across 40+ sources. Reps can ask which accounts are showing intent, inspect why, enrich the right contacts, and move a qualified account into a seller-controlled sequence from the same chat.

Unify Plays turns the score into governed action. Teams can set fit gates, exclusions, CRM ownership checks, and sequence rules before a signal triggers research, a task, an alert, or outreach. This is AI for SDRs, not an autonomous replacement for the seller.

Worked example: score a pricing-page visit without overreacting

A 300-person software company in the target market visits the pricing and integrations pages twice within one hour. Fit scores 27 of 30, signal strength 25 of 30, recency 20 of 20, and source confidence 16 of 20, producing 88 points.

The CRM shows an open opportunity owned by an AE. The account remains high confidence, but the action changes: notify the owner with the evidence and block a parallel sequence. That is the difference between detecting demand and creating operational noise.

Use this 30-second decision framework

• If the account is high fit and the signal is first-party commercial behavior, route it immediately after ownership checks.
• If fit is high but the signal is a broad market event, research the business hypothesis before messaging.
• If two independent weak signals align, create a research task and wait for confirmation.
• If the signal is strong but identity confidence is low, resolve the account before any person-level action.
• If the account is a customer or open opportunity, route to the existing owner.
• If consent, suppression, or regional eligibility is unclear, stop and review.

How should the model change by role and segment?

• Sales: weight ownership, role fit, and near-term commercial behavior most heavily.
• Growth: weight product usage, web behavior, and experiment traceability.
• Marketing: use broader research signals, but require sales-ready fit before routing.
• RevOps: own definitions, exclusions, decay windows, CRM fields, and audit logs.
• Enterprise: require multi-person or multi-source confirmation and route into account plans.
• SMB: use simpler bands and faster decay windows so work does not exceed account value.

Edge Cases and Disambiguation

• Job seekers versus buyers: careers-page concentration is not purchase intent.
• Opens versus engagement: opens alone are unreliable and should not trigger a high score.
• Funding versus need: funding creates capacity, not automatic relevance.
• Anonymous versus person-level intent: company identification does not reveal who has buying authority.
• Interest versus permission: intent signals do not remove consent or suppression obligations.

Stop or Adapt When a Red Flag Appears

Top 5 Mistakes to Avoid

• Treating a single page view as purchase intent.
• Letting a strong signal override poor account fit.
• Counting correlated events as independent confirmation.
• Ignoring CRM ownership and active-opportunity rules.
• Using a static score without recalibrating it against outcomes.

Ready to turn the framework into a seller-controlled outbound workflow? Try Unify free.

Frequently Asked Questions

How do I find accounts that are in market right now?

Combine account fit with recent first-party, product, organizational, and third-party signals. Score fit, signal strength, recency, and source confidence, then require at least 70 of 100 points for high-touch action. Apply ownership and suppression rules before outreach.

What is an in-market account?

An in-market account is a qualified company showing recent, credible evidence consistent with an active problem or evaluation. A single behavior is not proof of purchase. The account must fit, the signal must relate to your category, and the event must still be timely.

How many buying signals are enough?

One strong, auditable first-party signal can justify seller attention when fit is high. Two independent weaker signals can justify research if they support the same hypothesis. Correlated events from one session should count as one evidence cluster.

How recent should an intent signal be?

The useful window depends on the event and sales cycle. Pricing and product events decay quickly, while hiring or funding may remain relevant longer. Define a decay window for each signal and remove events when they age out.

Can intent data identify the right person to contact?

Account intent identifies a company, not necessarily the decision-maker. Resolve the buying role separately, verify current contact data, and check CRM ownership. Do not infer authority from a title alone.

How does Unify identify in-market accounts?

Unify combines first-party engagement, product and CRM data, third-party sources, and AI-discovered signals in one outbound workflow. Sellers can inspect evidence, research the account, enrich contacts, and activate approved Plays without leaving the same workspace.

Glossary

• In-market account: A qualified company showing recent evidence of active need or evaluation.
• First-party signal: Behavior observed directly in your website, product, CRM, or owned channels.
• Third-party intent: Modeled or observed research activity supplied by an external source.
• Signal strength: How closely an event maps to a credible buying action.
• Signal decay: The loss of decision value as an event gets older.
• Source confidence: The auditability and identity certainty of the evidence.
• Signal stacking: Combining independent observations to raise or lower confidence.

Sources

• Unify Signals, 2026.
• Unify Plays, 2026.
• Product Usage Signals: Turn Engagement Into Pipeline, Unify, 2026.
• Navattic customer story, 2026.
• How to Prioritize Signals in Your Outbound Motion, Unify Explore, 2026.

About the Author

Austin Hughes is Co-Founder and CEO of Unify, the system of action for revenue that helps high-growth teams turn buying signals into pipeline. Before founding Unify, Austin led the growth team at Ramp, scaling it from 1 to 25+ people and building a product-led, experiment-driven GTM motion. Prior to Ramp, he worked at SoftBank Investment Advisers and Centerview Partners.