Why do so many organizations invest so much in innovation — yet struggle to translate that investment into consistent results?
In his forthcoming book, Innovation Portfolio Management: Linking Strategy to Execution, Noel Sobelman says the secret to addressing that problem lies in building a disciplined system for translating the C-suite’s stated ambitions into action. The book features case studies of companies such as Becton Dickinson, Carrier, Tenneco, Novartis, Microsoft, and Garmin.
In this excerpt, Sobelman explains why using the right criteria to evaluate projects is so important. (Innovation Portfolio Management is available for pre-order now; it will be released on Sept 22nd, 2026.)
•••
Project evaluation criteria…determine how innovation projects are assessed, compared, and prioritized. The right evaluation criteria create clarity and bring objectivity to the decision process. And when the stakes are high, they prevent gut feel from masquerading as sound judgment.
In simple terms, project evaluation criteria are the standards by which investment-worthy projects are judged. They answer questions like: What makes one project more valuable than another? What level of business case confidence do we require before committing resources? How do we fairly compare vastly different opportunities? The answers define how the organization allocates its scarcest resources: time, talent, and capital.
Evaluation criteria are more than valuation metrics. They reflect strategy. When chosen well, they reinforce what the business wants more of. When poorly defined or inconsistently applied, they distort the portfolio, favoring familiar and safe projects over those with real growth potential.
In most companies, financial metrics are the default language for decision-making on core business innovation projects. Senior managers are proficient in thinking in terms of revenue, ROI, NPV, IRR, and payback periods. Financial criteria bring analytical rigor to what can otherwise be subjective debates. They force teams to connect their ideas to business outcomes and to grapple with the assumptions behind their forecasts. In the core business, financial models offer a reasonable picture of value and risk.
However, in early-stage innovation, especially for transformative, highly uncertain opportunities, those same models can be misleading. Projecting ROI when entering brand-new territory with no history to draw on is, at best, guesswork. The assumptions are often built on little more than hope. Teams lack reliable data on customer adoption, pricing, cost structure, and even technical feasibility. Scott Cook from Intuit summed it up well: “For every one of our failures, we had a spreadsheet that looked awesome!”
Over-reliance on financial estimates creates a subtle but powerful bias. It encourages the portfolio to tilt toward small, low-risk projects where the math looks good, even if the strategic upside is limited.
Over-reliance on financial estimates creates a subtle but powerful bias. It encourages the portfolio to tilt toward small, low-risk projects where the math looks good, even if the strategic upside is limited. Meanwhile, bolder bets with uncertain financials but a strong strategic rationale struggle to gain traction.
To correct this, leading organizations incorporate qualitative evaluation criteria that balance financial discipline with strategic intent and risk-adjusted insights. They evaluate projects along several dimensions, such as strategic alignment, commercial potential, technical feasibility, operational leverage, and market risk. These categories may be scored using a simple ordinal scale, with weightings tailored to each project’s level of uncertainty. For example, a project in early feasibility might be judged primarily on its strategic importance and an order-of-magnitude estimate of opportunity size, while a project nearing the full-scale design phase of development will be evaluated more heavily on financial return and execution risk.
Scoring models, when applied thoughtfully, support structured decisions without pretending to deliver precision. Their value lies in comparison, calibration, and discussion. They work best when they are simple, intuitive, and transparent, and used to guide judgment, not replace it. In some companies, scoring is done by the cross-functional portfolio analysis team and then confirmed by a portfolio governance body. In others, product line and functional leaders assign scores independently, then reconcile differences in a facilitated session. While the criteria need to be clearly defined and consistently applied, the score is a starting point for a conversation. There are no magic algorithms that can prioritize projects for the organization.
In some organizations, these qualitative criteria are unified by an overarching decision lens that sits above individual metrics. At Edwards Lifesciences, a global medical technology company focused on treating structural heart disease, that lens is purpose. The company’s credo, patients first, is not a slogan but the operating system of the business, shaping its strategy, investments, and portfolio decisions. As Tim Rumbaugh, Vice President of Program Management for Edwards’ Transcatheter Heart Valve business, described it, projects are not simply ranked by ROI or payback period. They are filtered through a higher-order lens: Does this initiative improve or extend patients’ lives? That question fundamentally changes how decisions are made.
Tim recalls the company’s decision to pursue a valve project for a very specific unmet need for patients, where the commercial rationale was not yet clearly defined across the broader market before Edwards got involved.
“We had a valve project that no one else would touch because the business case for it was not well understood in the market. But Edwards said, ‘We’re doing it because these patients need it.’ That kind of decision makes you realize this place really is different.”
For most organizations, such a project would never clear the hurdle rate. But at Edwards, it advanced because it aligned perfectly with the company’s purpose. Tim’s reflection underscores a broader point for leaders elsewhere: when purpose is clear and shared, it simplifies investment allocation decisions. Trade-offs remain hard, but the criteria for making them are widely understood and broadly supported.
Different types of projects often require different evaluation criteria and approaches. Cost reduction initiatives, sustaining engineering, and compliance mandates all play different roles in the portfolio and cannot be judged by the same yardstick as new products, new platforms, and transformational projects. Each strategic bucket, as defined in the portfolio structure, will likely have its own unique set of criteria. Similarly, continuous-delivery software projects, where value is delivered incrementally throughout a product’s lifecycle, are better assessed using behavioral metrics such as user engagement, retention, and customer satisfaction, which can be translated into financial impact through correlation modeling or cohort analysis.
Regent Education’s CEO, Jim Hermens, has firsthand experience with the complexity of valuing software in a continuous delivery model. Regent, which provides a cloud-based financial aid management platform for higher education, operates in a heavily regulated space with constantly evolving requirements.
In such an environment, the traditional upfront ROI model doesn’t always apply, especially when compliance-driven updates can dominate the roadmap. “We have a release model in our software development lifecycle (SDLC) that delivers updates four times a year by design,” Hermens explained. “Two of those are oriented toward compliance, and we have two other major releases and a series of minor maintenance releases.”
Rather than forcing long-range ROI projections for every initiative, Regent aligns evaluation to more immediate strategy and behavior-based outcomes. “We evaluate everything against cash, retention, and new sales.”
Retention, in particular, functions as a proxy for value creation. For Hermens, regulatory-driven releases are seen as both required maintenance and strategic levers to preserve customer loyalty: “From our standpoint, we’re going to try to get client benefit by arguing that we’re really good at managing these compliance problems and constant compliance changes. That clearly falls into a retention bucket, and it also supports new sales.”
…The company experimented with using Net Promoter Score (NPS) as a leading indicator but found it lacked signal clarity…
Hermens also shared how the company experimented with using Net Promoter Score (NPS) as a leading indicator but found it lacked signal clarity: “We ran Net Promoter Score for a couple of years, and we just kind of concluded it wasn’t giving us actionable insight. Institutions that loved us, renewed for multiple years, renewed with significant price increases, and bought more product were still giving us fives and sixes.”
His experience reinforces a broader point: while customer satisfaction metrics like NPS can be helpful, they’re not always reliable predictors of value. Measurable behavioral data, such as renewal rates, upsell success, or actual product usage, often provide a more grounded signal of the value delivered by continuous updates.1
The Regent Education example reinforces a key point: evaluation methods must match the context. Simpler, business-aligned metrics work well in some environments, while others require more complex approaches to account for risk (see Figure 6.5).
In those cases, the sophistication of valuation methods increases accordingly. Traditional financial metrics like ROI, margin, payback, and scoring models are relatively easy to apply and are familiar to leadership. More advanced methods attempt to account for uncertainty and staged decision-making. These include:
- Expected Commercial Value (ECV): Builds on discounted cash flow and layers in probabilities of commercial and technical success.
- Monte Carlo Simulations: Generates a range of outcomes by simulating variable inputs, producing probability distributions rather than point estimates.
- Decision Trees: Map contingent decisions and future events, back-calculating value along different paths.
- Real Options Analysis: The most complex approach, incorporating market volatility and valuing flexibility in future choices, similar to financial options.
While these sophisticated methods offer deeper insight into uncertainty and decision flexibility, they require more data, more modeling expertise, and more time. In industries like pharmaceuticals or aerospace, where billion-dollar investments are at stake, the effort is worthwhile. In others, the added complexity may not justify the cost. Still, organizations that make frequent, high-risk, high-reward bets increasingly incorporate these methods into their toolkit.
Author Noel Sobelman is Vice President of Business Development and Innovation at TDK Ventures. He was previously a Partner and Director at Accel Management Group, Change Logic, and Kalypso. His new book is Innovation Portfolio Management: Linking Strategy to Execution.
1 Interview with Jim Hermens, CEO, Regent Education, September 2025
You must be logged in to post a comment.