Growth & Digital/ICP Targeting
EDIT VIA GITHUB · /content
DRAFT / NOT FINAL — content pending review

ICP Targeting

The Redis ICP definition — verbatim from the Keyplay data dictionary (Jan 2026): core fields, tiering logic, and every enrichment field the R500 candidate calculation runs on.

Source DOCX ↓
MARKETS
2
Digital Native · Traditional Enterprise
TIERS
A–D
Keyplay model score → AI-reviewed final tier
R500
candidate rule
the one field computed via explicit logic
GRAIN
1 row / CRM record
upload key: CRM ID · canonical: Final Domain

Data Dictionary — Keyplay Core Fields

Overview

Purpose: Core Keyplay enrichment fields exported for CRM upload + account analysis. Grain: 1 row per CRM company record (multiple records may map to the same company via Final Domain). Primary upload key: CRM ID

Core Fields

FieldTypeExampleDescriptionNotes
CRM IDstring0013z00002LZXQFAA5Unique CRM record identifier used as the key for uploads.Use this to map back into Salesforce.
Input Domainstringexample.comThe domain pulled from the CRM “Website” field (raw input domain).May include redirects / non-canonical domains.
NamestringAcme CorpCompany name resolved via LinkedIn lookup.Best-effort enrichment (source: LinkedIn)
Final Domainstringacme.comCanonical domain after following redirects to the final landing domain.Used for deduping and consistent enrichment across CRM records.
Dupe?boolean (0 / 1)0Flags whether this CRM record is a duplicate of another record that resolves to the same Final Domain.0 = primary record (chosen by revenue or alphabetical). 1 = duplicate. All rows still get enriched. Deduping used to analysis and counting.
MarketstringDigital NativeMarket classification for the company.Traditional vs Digital Native.
SegmentstringFinancial Services & FintechKeyplay segment assigned to the company.Single select
R500 CandidateBoolean (TRUE / FALSE)trueIndicates whether the company qualifies as a Redis 500 candidate.Described as the only field computed via a specific logical rule.
ICP Tier (Final AI Reviewed)single selectAFinal ICP tier after an AI review pass.Single select
ICP Tier (Keyplay Model)single selectATier produced directly by the Keyplay scoring model.Deterministic based on ICP Score (Model) for the best fit Market (digital or traditional)
ICP Score (Model)number86Numeric ICP fit score output by the Keyplay model.Input used to compute ICP Tier (Keyplay Model).
ICP Fit ReasoningLong textLarge regulated financial enterprise (81k employees, \~12k engineers) with very high real-time dependency for payments/authorization and fraud detection, public-cloud deployments and active AI hiring — clear need for millisecond latency, geo‑replication, SLA/compliance support and platform consolidation; research gaps: current caching/datastore stack and PCI/regional certs.Short explanation of why the company received its score/tier.Intended to be rep-friendly CRM context (2–3 sentences).

Key Relationships / Logic

  • Upload key: CRM ID
  • Company canonical identifier (for grouping): Final Domain
  • Deduping for analysis: Use Dupe? = 0 to avoid double counting across records that share a Final Domain.
  • Tiering logic flow: ICP Score (Model) → ICP Tier (Keyplay Model) → ICP Tier (Final AI Reviewed) → R500 Candidate

Data Dictionary — Additional Enrichments (Optional)

Overview

Purpose: Optional enrichment fields available in the export. Primary use: These fields feed the R500 Candidate (Final) calculation. Note: Outside of the R500 candidate logic, these fields are optional and used for deeper segmentation, filtering, and rep context.

Additional Enrichment Fields

FieldTypeExampleDescriptionNotes
Final LinkedInURLhttps://www.linkedin.com/company/microsoftCanonical LinkedIn company profile URL used for enrichment.Source anchor for several fields.

Digital Native Model Outputs

FieldTypeExampleDescriptionNotes
Digital Native Fit Scorenumber (0–100)74Overall fit score for the Digital Native ICP model.Higher = better fit.
Digital Native Most Similar Tostring / domainstripe.comReference company the record is most similar to within Digital Native clustering.Used for explainability.
Digital Native Signal Scorenumber (0–100)74Score contribution from signals (rules/attributes) for Digital Native fit.Component score.
Digital Native Similarity Scorenumber (0–100)0Score contribution from vector similarity / nearest neighbors for Digital Native fit.Component score.
Digital Native Signalsmulti-select (text list)Software & Cloud Tech, Hiring, Using Public Cloud...Human-readable list of signals detected for Digital Native evaluation.Used in R500 logic + debugging.
Digital Native Tiersingle select (A/B/C/D)ATier bucket for Digital Native fit.Derived from Digital Native scores.

Traditional Enterprise Model Outputs

FieldTypeExampleDescriptionNotes
Traditional Enterprise Fit Scorenumber (0–100)90Overall fit score for the Traditional Enterprise ICP model.Higher = better fit.
Traditional Enterprise Most Similar Tostring / domainapple.comReference company the record is most similar to within Traditional Enterprise clustering.Used for explainability.
Traditional Enterprise Signal Scorenumber (0–100)90Score contribution from signals for Traditional Enterprise fit.Component score.
Traditional Enterprise Similarity Scorenumber (0–100)60Score contribution from similarity for Traditional Enterprise fit.Component score.
Traditional Enterprise Signalsmulti-select (text list)Employees 5001+, Traffic Rank Top 100K...Human-readable list of detected signals for Traditional Enterprise evaluation.Used in R500 logic + debugging.
Traditional Enterprise Tiersingle select (A/B/C/D)ATier bucket for Traditional Enterprise fit.Derived from Traditional Enterprise scores.

Intent Model Outputs

FieldTypeExampleDescriptionNotes
Intent Tiersingle select (A/B/C/D/none)noneBucketed intent strength for near-term buying signals.none means no intent detected.
Intent Scorenumber (0–100)0Numeric intent score.Often 0 for “no intent.”
Intent Signalsmulti-select (text list)Jobs mention Real-Time, Hiring spike...Detected intent-related signals that contribute to the score.Used in R500 calc when present.

Company Profile & Categorization

FieldTypeExampleDescriptionNotes
Business OverviewtextMultifaceted business model...Short company description / summary.Rep context + model explainability.
Primary Categorymulti-selectMedia & Entertainment, SaaS & CloudPrimary industry/category tags.Category taxonomy is Keyplay-defined.
Secondary Categorymulti-selectCollaboration & Productivity...Additional product/industry categories.Used for filtering and segmentation.
Audiencemulti-selectB2B, B2CPrimary go-to-market audience.Often multiple.

Company Size, Growth, Hiring

FieldTypeExampleDescriptionNotes
Employeesnumber225865Estimated employee count.Typically from LinkedIn or similar sources.
Year Foundednumber (YYYY)1975Company founding year.Used for maturity heuristics.
Headcount Growth (6 mo)number (ratio/%)\-0.05Headcount growth over the last 6 months.Negative indicates shrinkage.
Headcount Growth (1 mo)number (ratio/%)0.02Headcount growth over the last month.Short-term momentum signal.
Hiringnumber2464Count of open roles or hiring signal volume.Interpret based on your data source.
Open Roles / Total Headcountnumber (ratio)0.01Open roles divided by employee count.Hiring intensity proxy.
Recruiting Velocity Scorenumber21.82Composite score for hiring velocity.Used as a normalized hiring metric.

Traffic & Social Signals

FieldTypeExampleDescriptionNotes
Traffic Ranknumber3Website traffic rank bucket/value.Lower often implies higher traffic (depends on source).
Follower Countnumber27062289LinkedIn follower count.Company popularity / reach.
Follower Change (6 mo)number (ratio/%)0.06Change in followers over the last 6 months.Growth signal.
Follower Change (1 mo)number (ratio/%)0.01Change in followers over the last month.Short-term growth signal.

Public Company Info

FieldTypeExampleDescriptionNotes
Stock TickerstringNASDAQ:MSFTPublic market ticker identifier.Optional; only populated when applicable.

Headquarters & Locations

FieldTypeExampleDescriptionNotes
HQ CountrystringUSHQ country.ISO-like shorthand is common.
HQ CitystringRedmondHQ city.
HQ StatestringWashingtonHQ state/region.US-specific field when applicable.
HQ Postal Codestring98052HQ postal code.Optional.
Location Citiesmulti-select (text list)Bratislava, Reading, Singapore...Cities where the company has locations.Can include noisy/unicode entries.
Location Statesmulti-select (text list)Ontario, Dublin, Tokyo...States/regions where the company has locations.Source-dependent formatting.
Location Countriesmulti-select (text list)US, CA, GB...Countries where the company has locations.Useful for geo filtering.
Location Countnumber44Count of distinct locations detected.Proxy for global footprint.

Funding

FieldTypeExampleDescriptionNotes
Total Fundingnumber1000000Total funding raised (if known).May be blank for public/bootstrapped firms.
Last Funding Datedate12/9/22Date of most recent funding round.Format may vary by export settings.

Internal Flags & CRM Metadata

FieldTypeExampleDescriptionNotes
Tagsmulti-select*(blank)*Freeform tags associated with the company.Optional annotation layer.
Statussingle selectOkProcessing/status flag for the enrichment output.Useful for QA.
In SAMboolean (yes/no)yesIndicates whether the account is in the Serviceable Available Market.A gating filter before ICP.
In CRMboolean (yes/no)yesIndicates whether the company already exists in CRM.Useful for backfills.
Account Sourcesingle selectCRMWhere the account record originated.Example: CRM, imports, etc.
FieldTypeExampleDescriptionNotes
Logo UrlURLhttps://media.licdn.com/...Company logo URL (often LinkedIn-hosted).Useful for UI/display.
Job Board URLURLhttps://www.linkedin.com/company/microsoft/jobs/Primary jobs page link.Hiring validation + sales trigger.

AI-Derived Categorization Fields (with reasoning)

FieldTypeExampleDescriptionNotes
Digital Native vs Traditional (Category)single selectTraditional EnterpriseAI classification of company as digital native vs traditional.Used in segmentation and R500 logic.
Digital Native vs Traditional (Category) ReasoningtextFounded in 1975...Explanation for the category assignment.Rep + audit trail.
Key Segments (Category)single selectSoftware & Cloud TechAI classification into a key segment category.Higher-level version of Segment.
Key Segments (Category) ReasoningtextPrimarily develops software...Explanation for segment assignment.Auditability.
Software Engineers (Team Size)single select50+Estimated engineering team size bucket.Bucketed, not exact.
Software Engineers (Team Size) ReasoningtextFound 115,740...Explanation for the team-size estimate.Useful for trust/debugging.
AI-Native (Boolean)booleanFALSEWhether AI is foundational to the business model.Not “uses AI” — “built on AI.”
AI-Native (Boolean) ReasoningtextPredates AI era...Explanation for AI-native assignment.Auditability.
Engineering Ideology (Rating)rating (1–10)5Qualitative score for engineering philosophy/approach.Interpretable heuristic score.
Engineering Ideology (Rating) ReasoningtextBalances proprietary builds...Explanation for ideology rating.
Real-Time Data Dependency (Rating)rating (1–10)9Degree to which the company depends on real-time data.Strong R500 signal.
Real-Time Data Dependency (Rating) ReasoningtextCritically depend on...Explanation for real-time dependency rating.
Verbatim content from the source DOCX (downloadable above). These definitions feed the 6sense segments, the Redis500 list, and tier A/B targeting — changes ripple into 004 · Signal-Based Orchestration and 002 · Paid Media.