Observing the Economic Value of Information (EVI)

The Transmutation (High Net Economic Value, Major Revenue Uplift, Very Resilient). Author using Gemini

Key Question: What is the net financial contribution from this data?

Level 1 dimensions

Revenue uplift
- Incremental revenue attributable to data‑driven decisions (e.g. price optimisation, cross‑sell).
- New products or services that depend on this data.
- Increased customer lifetime value where data plays a proven role.
Cost savings / avoidance
- Reduced operational costs (automation, fewer errors).
- Lower loss ratios / fraud / claims leakage.
- Avoided capex/opex by using data instead of physical experiments.
Risk reduction / capital efficiency
- Reduced regulatory fines / incidents.
- Lower capital charges due to better risk models.
- Improved resilience (e.g. fewer outages, better contingency planning).
Net value
- EVI_gross = revenue uplift + cost savings + risk reduction (in $).
- EVI_net = EVI_gross − CVI (including governance spend).
- NPV of EVI_net over expected data lifecycle.

EVI should be tied to your finance systems: map uplift and savings to P&L lines where possible.
Rank assets by EVI_net and EVI/CVI ratio to drive capital allocation.

To make this usable in the wild:

For each dimension (e.g. IVI: Validity), specify what 1, 5, and 7 look like with concrete numbers or examples.

Pick 3–5 known datasets (one exemplar, one average, one problem case).
Score them independently with the rubric; refine thresholds until scores align with expert intuition.

Document how to score someone else’s case
- For external assessments (e.g. vendor data, acquisition due diligence), specify:
  - What evidence you expect (sample data profiles, contracts, case studies).
  - Which dimensions you can score directly vs which need proxies (e.g. you may see adoption, but not internal EVI).

Assume EVI is the weighted sum of four dimensions:

Question: How much incremental revenue is attributable to using this data?

Score	Descriptor
1 – 2 None or negligible	No demonstrated revenue impact. No experiments, or A/B tests show no material uplift.
3 – 4 Minor	Small, local uplift (<0.5% of relevant revenue stream or <0.1% of total revenue). Benefits not yet stable or only in pilots.
5 – 6 Moderate	Clear, measured uplift (0.5–2% of relevant revenue, or 0.1–0.5% of total revenue) in at least one scaled use case.
7 – 8 Significant	Strong, recurring uplift (2–5% of relevant revenue, or 0.5–1% of total revenue) across multiple use cases or regions.
9-10 Major driver	Material revenue driver (>5% of relevant revenue, or >1% of total revenue). Data is essential to core revenue‑generating engines (pricing, cross‑sell, product success).

Question: How much cost does this data save or help avoid?

Score	Descriptor
1 – 2 None or negligible	No observable reduction in operating costs, losses, or manual work.
3 – 4 Minor	Some localised savings (e.g. a small team’s workload), but <0.5% of relevant cost base. Benefits anecdotal or not tracked systematically.
5 - 6 Moderate	Measurable savings (0.5–2% of relevant cost pool: e.g. claims, operations, servicing), verified by finance or operations.
7 - 8 Significant	Ongoing cost reduction (2–5% of a major cost pool, or 0.5–1% of total operating expense), sustained over time.
9 - 10 Major driver	Structural cost advantage (>5% of a major cost pool or >1% of total opex). Data is key to automation, capacity reduction, or structural efficiency.

Question: How does this data improve risk management, capital use, or regulatory outcomes?

Score	Descriptor
1 – 2 None or unclear	No identified impact on risk metrics, capital requirements, or regulatory fines.
3 - 4 Incidental	Some qualitative sense that risk is better managed (fewer surprises, better monitoring), but no quantified capital or loss impact.
5 – 6 Measured reduction	Quantified reduction in losses, provisions, or capital (e.g. fewer defaults, better fraud detection) equivalent to 0.1–0.3% of RWA or similar risk base.
7 – 8 Significant improvement	Clear reductions in losses/capital (0.3–1% of RWA or equivalent), or materially lower regulatory incident/penalty rates.
9 – 10 Strategic risk asset	Data is core to risk and capital strategy: supports models or processes that materially lower capital charges, avoid large fines, or enable entry into risk‑sensitive markets.

Question: What is the net economic contribution once we include CVI (all relevant data costs)?

Score	Descriptor
1 – 2 Negative	EVI (revenue + savings + risk benefit) < CVI. Net value is negative or uncertain; no clear business case.
3 – 4 Marginal	EVI only slightly exceeds CVI (EVI ≈ 0–0.5× CVI). Payback is long or fragile; the project is vulnerable to small shocks.
5 – 6 Acceptable	EVI is 0.5–2× CVI. Positive business case with reasonable payback, but not a standout investment.
7 – 8 Strong	EVI is 2–5× CVI. Clear, robust ROI; data is a strong economic asset.
9 – 10 Exceptional	EVI > 5× CVI, or EVI is large and highly resilient. This is a star asset in the portfolio.

For each dataset/use case, assign 1–5 for each dimension based on the descriptors.
Compute a weighted average:

𝐸𝑉𝐼=0.30⋅𝑅𝑢+0.30⋅𝐶+0.20⋅𝑅𝑘+0.20⋅𝑁

Where:

We can then rank assets by EVI and by EVI/CVI to guide capital allocation.