data-monetization-model-selection

Selects and justifies a suitable data monetization model (direct: DaaS/Insight-as-a-Service/data exchange vs. indirect: product enrichment/resource optimization/risk mitigation/Data Flywheel) with a d

SKILL.md

GitHub

Data Monetization Model Selection

Purpose

Once data has been diagnosed as a strategic asset (see ../data-role-diagnosis/SKILL.md), the next question isn't "should we monetize" but "with which model." Direct and indirect monetization require different capabilities, a different risk profile, and a different timeline — mixing the two without a deliberate choice typically leads to an organization trying to sell data externally before internal productization has even been tested, or, conversely, leaving a clear external demand unexploited. This skill produces a justified model choice and separately checks whether a "data flywheel" claim is real or wishful thinking.

Anchored in research

  • Direct and indirect data monetization models (industry consulting practice, a synthesis of multiple sources, 2026): Data-as-a-Service, Insight-as-a-Service, data exchange/IP licensing (direct); product enrichment, utilization/resource optimization, business-risk mitigation (indirect).
  • The Data Flywheel mechanism in AI business: unique data → model training → better product → more users → more data. As a concept, an extension of Collins's (2001) flywheel mechanism (see the anchoring in ../data-role-diagnosis/SKILL.md) applied to the data/AI context.
  • Business model innovation patterns from the ../../../business-design-frameworks and ../../../specialisation-packs/business-model-canvas packs (e.g. financial.rev.data_monetization, operating.resources.leverage_customer_data) — this skill deepens one pattern family with data-specific detail.

Method

  1. Walk through the direct-monetization decision tree first, since it rules out options quickly:
    • Is selling the data externally legally/contractually permitted (customer contracts, GDPR and other privacy law, third-party IP rights)? If not, the direct model is ruled out without a separate legal review — don't proceed to a direct model before this check.
    • Is there an identified external buyer or market for this data/insight TODAY (the resale test, see ../data-role-diagnosis/SKILL.md)? If not, the direct model would first require validating the market, not just productizing it.
    • Is the data unique enough that a buyer couldn't acquire the same thing more cheaply elsewhere (the defensibility test)? If the data is widely available, pricing power in direct sale is weak. If all three pass, consider a direct model: DaaS/raw-data sale (the lowest processing level, requires the strongest data governance), Insight-as-a-Service (analytics/forecasts instead of raw data, protects the raw data better), or data exchange/IP licensing (data traded for another party's data, technology, or market access — a common model in innovation ecosystems).
  2. If the direct model didn't pass point 1, move to an indirect model — this is usually both more profitable and lower-risk:
    • Product/service enrichment: data enables a new, paid feature or a better user experience within an existing product (e.g. hyper-personalization).
    • Utilization/resource optimization: data reveals bottlenecks and enables dynamic pricing or more efficient resource allocation — shows up directly as cost savings or added revenue.
    • Business-risk mitigation: data enables proactive detection (e.g. customer churn, fraud, equipment failure) before the event — "a euro saved is a euro earned."
  3. If a "data flywheel" model is claimed, check it separately with a four-point checklist — don't accept a flywheel claim without this:
    • Unique collection channel: is there a genuine, ongoing way to collect data that a competitor doesn't have at the same scale?
    • Measurable model improvement: does the model/product provably (not just presumably) improve as more data accumulates — is there already evidence of this, or is it still an assumption?
    • Observable user-experience improvement: does the user notice the model improving, or is the improvement so small it doesn't affect behavior?
    • The loop closing as growth: does the observed improvement provably attract more users, who generate more data — or does the loop break somewhere (e.g. the improvement isn't enough to attract new users)? If even one of the four points fails the check or is still just an assumption, name it explicitly as a risk rather than presenting the flywheel as an already-working, self-sustaining mechanism.
  4. Compare the model on speed, risk, and competitive advantage, not just revenue potential: a direct model often produces a faster but more easily copied revenue stream (the data itself is no longer your exclusive once it's been sold); an indirect model builds more slowly but toward more defensible advantage.
  5. Produce a justified recommendation in one sentence per option, naming WHY the model fits or doesn't fit this situation — not just the model's name.

What this skill does NOT do

  • Doesn't provide a legal assessment of privacy or contractual obstacles to selling data externally — identifies that a check is needed, but doesn't replace separate privacy/legal expertise.
  • Doesn't calculate exact pricing or ROI for the chosen model — see ../../../business-case-and-analysis/skills/roi-npv-sensitivity-model/SKILL.md.
  • Doesn't claim an indirect model is always better than a direct one — context decides; the skill produces a structured comparison, not a ready-made answer.
  • Doesn't confirm figures, market demand, or competitor data from memory — uses the inputs you provide, or marks an assumption clearly ([assumption — verify]).

Refinement notes

Areas to keep deepening with real practice:

  • your own examples of when a direct vs. indirect model turned out to be the right choice (and why the wrong choice would have failed)
  • a concrete decision-tree template as a visual tool (into ../../references/)
  • your own observations on which of the four flywheel checklist points most often fails in practice

Once this section is filled in and validated in practice, update the maturity field in skills_index.json to draft, validated, or canonical (see ../../../meta/maturity_levels.md). Don't add new fields to the frontmatter — name and description are the only ones allowed (see ../../../meta/frontmatter_schema.md).

Continue from here

  • Before this: ../data-role-diagnosis/SKILL.md — first confirm that the data passes the asset tests.
  • Before this (prioritization before model selection): ../data-ai-strategy-design-and-prioritization/SKILL.md
  • Related skill in another pack: ../../../specialisation-packs/business-model-canvas/skills/bmc-innovation-pattern-matching/SKILL.md — if the chosen monetization model still needs to be fitted into a broader business model (Financial Model patterns).
  • Related skill in another pack: ../../../business-case-and-analysis/skills/roi-npv-sensitivity-model/SKILL.md — translates the chosen model into a financial calculation.
  • A ready-made skill chain for this situation: see ../../../playbooks/
  • This pack's shared guardrails: ../../CLAUDE.md

References

  • ../../references/data-monetization-frameworks.md — a broader comparison of direct and indirect monetization models
  • ../../references/ — the pack's shared background material
  • ../../CLAUDE.md — the pack's shared guardrails