Feature · Deal Matching Algorithm

A deal matching algorithm the rep actually trusts because the receipts are on the card.

Deal matching algorithms tend to arrive as an opaque score the sales lead was told to trust. The rep sees a number, has no way to inspect why the match fired, and ends up ignoring it. Strkr builds the matching algorithm inside the CRM with the typed signals, the similarity method, the playbook routing, and the receipts on the opportunity card. The rep sees why the match fired and takes the next step with context rather than reading a black box. The pricing is a flat seat with no per-match fee.

Start free trial See pricing
How the matching actually works

Seven signals the algorithm reads before it fires.

A matching score that reads one signal is a sorting rule in a trench coat. One that reads seven typed signals and lands the right playbook on the opportunity is a tool. The signals below cover the common sales matching jobs: find similar deals, route to the right rep, suggest the right playbook, surface the right competitor.

Account firmographics

Segment, size, industry, geography.

The algorithm reads the typed account firmographics (segment, employee count, industry, geography) and finds prior opportunities on accounts with similar firmographics. The common failure where a match surfaces a mid-market deal for an enterprise rep because the segment field was ignored does not happen.

Product mix

Quoted products on the opportunity.

The algorithm reads the quoted products on the opportunity and finds prior opportunities with overlapping product mix. A multi-product deal matches other multi-product deals rather than a single-product match that would send the rep down the wrong playbook. The product match is on the typed opportunity-product rows.

Stage history

How the opportunity moved through stages.

The algorithm reads the stage history (how long each stage took, which stages the opportunity skipped) and finds prior opportunities with similar paths. A slow-at-pricing deal matches other slow-at-pricing deals so the suggested playbook is the one that cleared the pricing block before.

Activity cadence

Call, email, meeting rhythm.

The algorithm reads the activity cadence (calls per week, meeting frequency, email thread depth) and finds prior opportunities with similar cadence. The common failure where a quiet account matches an active account and the rep gets the wrong playbook does not happen because the cadence signal is read.

Competitor signal

Named competitor on the opportunity.

The algorithm reads the typed competitor field on the opportunity and finds prior opportunities with the same competitor named. The suggested playbook pulls from the battlecard that worked in prior deals against that competitor rather than a generic competitor response.

Contact seniority

Decision maker and champion roles.

The algorithm reads the typed role on the contacts attached to the opportunity (economic buyer, champion, user) and matches on the role mix. A deal with a champion but no economic buyer matches other deals with the same gap so the suggested playbook is the one that landed the economic buyer meeting.

Source attribution

Inbound, outbound, partner, event.

The algorithm reads the source attribution on the opportunity (inbound lead, outbound SDR, partner referral, event) and matches on source. Inbound deals have a different closing pattern than outbound deals, so the suggested playbook comes from the right pool rather than the whole-pool average.

What the match drives

Six actions the matching algorithm fires once it finds a hit.

Finding similar deals is the easy half. Driving the right action is the harder half. The matching algorithm fires a playbook, routes to the right rep, suggests the right competitor response, and leaves the receipts on the card. The six actions below cover what the rep expects after the match fires.

Suggested playbook

The sequence that closed similar deals.

The matched prior deals drive a suggested playbook: the sequence of outreach, meetings, and content that closed the similar wins. The rep sees the playbook on the opportunity card with a Start button rather than hunting through a Confluence page. The common failure where the playbook lives in a wiki nobody opens gets avoided.

Rep routing

Lead reassigns if the match points elsewhere.

If the matched deals clustered on a different rep (because that rep won the segment before), the lead router flags the reassignment to the sales lead for approval. The lead router does not auto-reassign without the sign-off, but the signal lands on the lead's home so the conversation happens.

Competitor battlecard

The response that worked before.

When the matched deals share a competitor, the competitor battlecard surfaces on the opportunity card with the response language that landed the win before. The rep reads the response in context rather than searching the enablement library mid-call. The battlecard is a typed asset so the version is on record.

Pricing anchor

What similar deals signed at.

The matched deals carry the signed amount, the discount tier, and the term length. The pricing anchor shows up on the opportunity card so the rep knows what similar deals cleared. The common failure where a rep guesses a discount and misses the pattern does not happen because the anchor is on the record.

Risk flag

If the matched deals mostly lost.

If the matched deals mostly lost, the risk flag fires on the opportunity card with the pattern explained (lost at pricing, lost at legal, lost at champion departure). The sales lead sees the risk before the forecast call rather than finding out in the postmortem. The flag is read, not noise.

Receipts on the card

Click the match, see the prior deal.

The match carries the receipts: which prior opportunities drove the score, which signals contributed most, and a one-click drill-through to the matched record. The rep inspects the receipts and takes the next step with context rather than reading a black box.

The engineering underneath

What keeps the matching honest at scale.

A matching algorithm that works on 100 opportunities and overfits on 100,000 is a demo. One that holds across quarters, as segments shift and product mix changes, is a tool. Strkr ships the primitives that keep the match honest.

Typed signals

The algorithm reads fields, not free text.

The algorithm reads typed fields rather than free-text parsing. The segment field is an enum, the amount is a currency, the stage is a controlled value. The common failure where a matching algorithm pulls garbage from a notes field and surfaces nonsense matches does not happen because the inputs are typed.

Explainable score

Each signal's contribution is on the card.

The score is the sum of weighted signals. Each signal's contribution shows on the card: segment match contributed 20 points, product mix 15, stage path 10. The rep can inspect why the match fired. The common failure where a black-box score loses trust the first time it surfaces a weird match gets avoided.

Permission scope

The rep only matches against visible deals.

The matching scope honors the rep's record permission. A rep who cannot see enterprise opportunities does not get enterprise matches surfaced. The common failure where a matching score leaks information about deals the viewer should not see does not happen because the row filter runs first.

Weight tuning

Ops lead adjusts signal weights.

The ops lead adjusts the signal weights from the admin surface based on what landed. If product mix matters more in a given tenant than segment, the ops lead cranks the product-mix weight. The common failure where a matching algorithm is tuned once at vendor install and never adjusted does not happen.

Backfill history

The algorithm reads the full typed history.

The matching algorithm reads the full typed opportunity history, not a nightly export. A deal closed yesterday contributes to today's match. The common failure where the matching tool lags the live record and surfaces matches that are already stale gets avoided because the inputs are live.

Audit log

Every match surface lands on the audit.

Every match surface lands on the audit log keyed on the opportunity, the viewer, and the matched records. A compliance query covering every opportunity the match flagged as high-risk in Q3 is a filter on the audit page. The audit log is append-only.

Where teams land in practice

Three deployment patterns from live customers.

A 70-seat SaaS, a 160-seat professional services firm, and a 350-seat security software vendor each tune the matching algorithm differently. The platform plumbing stays the same. The weights and the surfaced actions flex to the business.

SaaS tuning

Product mix dominates the score.

A 70-seat SaaS cranks the product-mix weight because the multi-product deals follow a different closing pattern than single-product deals. The suggested playbook changes substantially based on the mix. The sales lead cites the match as a factor in the forecast call.

Services firm

Account firmographics dominate.

A 160-seat professional services firm cranks the firmographic weight because the deal pattern tracks the account industry heavily. A healthcare prospect matches other healthcare wins and the suggested playbook ships the healthcare case study and the HIPAA posture statement.

Security vendor

Competitor signal dominates.

A 350-seat security software vendor cranks the competitor weight because the response language differs sharply by competitor. A deal against Vendor A matches other Vendor A deals and the battlecard surfaces on the opportunity card for the discovery call.

Monthly review

Sales lead reviews surfaced matches.

Monthly, the sales lead reviews the matches that fired against the deals that closed. If the match predicted a win and the deal lost, the lead looks at the signal receipts. The feedback loop tunes the weights without a vendor engagement.

Quarterly audit

Ops lead inspects weight drift.

Quarterly, the ops lead inspects the weight configuration against the deal pattern. If the segment distribution shifted (more mid-market, fewer enterprise), the weights adjust. The adjustment is a form on the admin surface rather than a vendor ticket.

Playbook refresh

Enablement updates the suggested sequence.

When enablement updates a playbook, the match fires the new sequence on the next opportunity that hits the pattern. The common failure where the playbook update sits in a Google Drive folder for a quarter before anyone adopts it gets avoided because the match surfaces the new version on the card.

A deal matching algorithm the rep trusts because the receipts are on the card.

Strkr reads typed signals, scores similarity with Strkr AI, fires playbook routing, and surfaces the receipts on the opportunity. Flat seat price with no per-match fee. HIPAA BAA available, SOC 2 Type II in progress.

Start free See pricing
Common questions

What buyers ask about this feature.

How is the Strkr deal matching algorithm different from a generic AI scoring bolt-on?

A generic AI scoring bolt-on reads a flattened CSV export, surfaces an opaque score, and leaves the rep to trust a black box. The Strkr algorithm reads typed signals from the live record, surfaces an explainable score with every signal's contribution visible, and lands the suggested playbook on the opportunity card. The rep can click through to the matched prior deals and inspect why the match fired. The pricing is a flat seat rather than a per-match fee, so the sales team runs the matching on every opportunity without a cost conversation.

Can the ops lead tune the signal weights?

Yes. The ops lead adjusts the signal weights from the admin surface based on what lands in the given tenant. If product mix matters more than segment in a given business, the ops lead cranks the product-mix weight. The adjustment is a form on the admin surface rather than a vendor ticket. The weight configuration carries versioning so a quarter-end audit can reconstruct the match logic that was in effect at any point. The common failure where a matching tool is tuned once at vendor install and never adjusted does not happen on Strkr.

How does the matching algorithm honor record-level permissions?

The matching scope honors the rep's record permission. A rep who cannot see enterprise opportunities does not get enterprise matches surfaced. The row permission filter runs before the matching aggregation so the match score cannot leak information about deals the viewer should not see. The common failure where a matching tool reveals aggregated information that lets the viewer back out individual deals does not happen because the permission scope applies at the row filter boundary.

Does the matching algorithm reassign deals automatically?

No. If the matched prior deals cluster on a different rep, the lead router flags the reassignment to the sales lead for approval rather than auto-reassigning. The signal lands on the lead's home surface so the conversation happens with the rep rather than behind their back. Auto-reassignment is a toggle the admin can enable per routing rule, but the default is sign-off because the human context (relationship, in-flight meeting, split logic) matters more than the algorithmic signal.

Where does the matching algorithm store its receipts?

The receipts live on the typed opportunity record. Every match fires with the matched prior opportunities, the contributing signals, the weighted score components, and the suggested next action attached to the opportunity. The audit log carries every match surface keyed on the opportunity, the viewer, and the matched records. The receipts are inspectable by the rep on the card and by the sales lead on the audit page. The common failure where a matching tool surfaces an opaque score with no inspection path does not happen on Strkr.

Does the matching algorithm carry a per-match fee?

No. Strkr is $50 per seat flat with no per-match fee, no per-score fee, and no per-playbook fee. A sales team that runs the matching on every opportunity pays the same as a team that runs it on a subset. The common failure where a matching vendor charges per score and the sales lead ends up rationing the signal to the top deals does not happen on Strkr. The seat price on the pricing page is the full price for the deal matching capability.

Related features

Pair this with...

AI Lead Scoring AI Sales Forecasting Automated Lead Routing Deal Tracking Deal Pipeline Software AI Sales Automation AI Sales Assistant Activity Tracking

Try it free. Bring your team next week.

No sales call, no migration consultant, no four-month implementation. Enter your card, get 14 days of the full Pro tier, cancel any time before day 14 with zero charge. Spin up a workspace, import your CSV, and have something useful before lunch.

Start free trial See pricing