# Summary: 2026-07-26_16-05-39Z_AIStrategy_HowtoChooseWhatAIProducttoImplement.md
Saved: 2026-07-27 21:29
Source: 2026-07-26_16-05-39Z_AIStrategy_HowtoChooseWhatAIProducttoImplement.md
Model: None

---

## Summary  
The paper introduces the eROI (expected ROI) framework to help firms decide which AI products to implement, breaking down each potential project into three independent components: Value if Successful, Likelihood of Success, and Investment Required. By separating these elements, the authors resolve a classic chicken‑egg problem—teams cannot reliably estimate ROI until they know whether a project will work, yet they need that knowledge to evaluate it—and instead recommend evaluating a pool of ideas before ranking them to build a diversified portfolio rather than funding only the single top‑ranked project. The framework is illustrated with Compass’s two AI initiatives: a “Likely‑to‑Sell” recommendation system and a “Time‑on‑Market” pricing tool, where one was deemed high‑value/high‑likelihood/low‑cost while the other was low‑value despite higher investment.

## Key Contributions  
- [Finding 1] eROI decomposes AI project decisions into three distinct components—Value if Successful, Likelihood of Success, and Investment Required.  
- [Finding 2] The framework breaks the circularity between ROI estimation and success probability assessment by treating each component independently.  
- [Finding 3] It recommends evaluating a pool of ideas before ranking them to construct a portfolio of bets rather than investing solely in the highest‑ranked single project.

## Methodology  
The authors applied eROI to Compass, a residential real‑estate brokerage, by gathering executive judgments on each AI product’s value if it succeeded, its probability of success, and its implementation cost. These qualitative ratings were aggregated into an overall score for each project, allowing the team to compare them objectively without needing precise financial forecasts.

## Results  
The “Likely‑to‑Sell” recommendation system received high scores across all three dimensions: strong value potential (generating nine figures in annual gross commission revenue), a high likelihood of success given its data‑driven approach, and modest investment cost. In contrast, the “Time‑on‑Market” pricing tool scored low on value despite higher investment, leading to its shelving. The eROI ranking thus guided a clear strategic decision: invest in the recommendation system while discarding the pricing tool.

## Significance  
eROI offers a practical, non‑financial decision‑making tool that aligns AI investments with business outcomes, reduces overcommitment to uncertain projects, and supports portfolio diversification. By providing a structured way to compare ideas on value, likelihood, and cost, it helps organizations allocate limited resources more effectively in the rapidly evolving AI landscape.

## Related Concepts  
- ROI (Return on Investment)  
- Expected Value (EV)  
- Portfolio Theory  
- AI Project Evaluation  
- Value‑Likelihood‑Investment Tradeoff