Innovation Portfolio Management for Manufacturing: Why standard portfolio advice fails on the factory floor

Innovation portfolio management for manufacturing requires a fundamentally different approach than what most guides describe. I have managed innovation portfolios at more than 40 industrial companies, and the gap between standard portfolio advice and manufacturing reality is enormous.

The typical portfolio management guide assumes you can run 20 explore experiments for €500.000. In manufacturing, that budget might cover one. It assumes you can validate a business model in six weeks. In manufacturing, a single prototype can take six months and €200.000 before you have anything to test with a customer. It assumes that killing a project means deleting a repository. In manufacturing, it means writing off tooling, reassigning specialized engineers, and explaining to the board why €800.000 produced learning instead of revenue.

These are not minor adjustments. They change the portfolio math, the governance cadence, and the kill criteria. Get them wrong, and you end up with one of two outcomes I see constantly: a portfolio stuffed with safe incremental projects that leadership calls “innovation,” or three ambitious explore bets that drain budget for years without anyone having the governance mechanism to stop them.

For the full portfolio management methodology, read the Innovation Portfolio Management practitioner’s guide. This article focuses on what changes for manufacturing and industrial B2B.

Why manufacturing breaks standard portfolio management

Four structural realities of manufacturing make generic portfolio advice dangerous.

The cost per explore bet is 10 to 100 times higher. A software company can test a new business model with a landing page, some ads, and a prototype built in two weeks. Total cost: €10.000 to €50.000. A packaging machinery company exploring a subscription-based maintenance model needs sensor integration, a data platform, field trials with existing customers, and regulatory compliance review. Minimum viable exploration: €500.000 to €1.500.000. This means your portfolio has fewer bets, each one matters more, and the cost of backing the wrong project is measured in capital equipment, not cloud credits.

Validation cycles are 12 to 18 months, not 2 to 6 weeks. Physical products cannot be A/B tested. Building a prototype for customer testing involves procurement, tooling, assembly, certification, and field deployment. An agricultural equipment manufacturer I worked with needed 14 months from concept to a prototype that customers could actually evaluate. During those 14 months, the explore project consumed budget and engineering capacity with no market evidence. The portfolio governance system needs to account for these timelines without confusing “slow progress” with “no progress.”

Engineering capacity is the real constraint, not budget. In most manufacturing companies, the binding constraint on innovation is not money. It is the 15 senior engineers who understand the core technology. Every explore project that pulls an engineer from the core product line creates a visible, measurable impact on exploit performance. The portfolio is not just managing financial allocation. It is managing the scarcest resource the company has: the people who can actually build things.

Regulatory approval adds 6 to 18 months after validation. For manufacturers in food, pharma, medical devices, chemicals, or automotive, an explore project that achieves customer validation is still 6 to 18 months from market. Regulatory submissions, certification processes, and compliance testing extend the timeline past what most portfolio frameworks account for. Your portfolio needs a “regulatory runway” layer that software companies never think about.

These four realities do not make innovation portfolio management less important for manufacturing. They make it essential. Without deliberate portfolio governance, the default outcome is predictable: 95% of resources go to core improvements, explore starves, and three years later leadership asks why the company has no new growth engines.

What innovation portfolio management for manufacturing actually looks like

I will not repeat the frameworks. The practitioner’s guide covers the Business Portfolio Map, Three Horizons, Innovation Ambition Matrix, and when each applies. Here I focus on how manufacturing context changes the implementation.

Allocation: the 70-20-10 rule does not apply as written

Nagji and Tuff’s 70-20-10 framework is the most cited allocation model in innovation. 70% core, 20% adjacent, 10% transformational. Their research found that companies following roughly this ratio outperformed peers.

The problem for manufacturing: when your total product development budget is €20.000.000 and a single transformational explore project costs €1.500.000, the “10% transformational” allocation of €2.000.000 gives you exactly one project with a thin buffer. One. If that project fails (and most explore projects should fail, that is the nature of exploration), your entire transformational portfolio is empty for the year.

What I see working in manufacturing:

Separate the allocation conversation from the percentage conversation. Instead of “what percentage goes to explore,” start with “how many explore bets can we run simultaneously, and what is the minimum viable investment for each?” An industrial automation company I worked with determined they needed at least three active explore projects to have a reasonable chance of one succeeding. At €600.000 minimum per project, that required a ring-fenced explore budget of €1.800.000 per year. That happened to be about 8% of their product development budget, but the number came from the bottom up, not from a percentage rule.

Protect the explore budget at CFO level. In manufacturing, quarterly pressure from exploit is intense. The production line needs an upgrade. A key customer needs a customization. The cost reduction target is at risk. Every one of those requests is urgent and justified. Without a CFO-level commitment that the explore budget is protected, it will be raided within two quarters. I have seen this happen at six different companies. The explore budget needs the same protection as the capital expenditure plan: a specific amount, approved by the board, with explicit re-approval required to reallocate.

Count engineering hours, not just euros. A manufacturer can have €2.000.000 ring-fenced for explore and still have zero capacity if all senior engineers are committed to exploit projects. I track portfolio allocation on two axes: financial investment and engineering capacity. If the financial allocation says 10% explore but the engineering allocation is 2%, you have a paper portfolio with no execution capacity.

Governance: slower cadence, higher stakes

Standard innovation governance suggests weekly or bi-weekly experiment reviews. For manufacturing explore projects, that cadence produces meetings where the answer is “we are still waiting for the prototype” for months in a row.

Monthly evidence reviews work better. Not progress reviews. Evidence reviews. The question is not “what did you do last month?” but “what did you learn last month?” Even during a 14-month prototype development cycle, there is evidence to generate: customer interviews, supplier feasibility assessments, regulatory pre-consultations, competitive analysis, pricing validation through letters of intent. A team that has no new evidence after a month is a warning signal, regardless of where the prototype stands.

Quarterly portfolio-level decisions. Every quarter, the governance board reviews the full portfolio against allocation targets and makes three types of decisions: advance (this project has enough evidence to move to the next stage), pivot (the evidence suggests a different direction), or kill (the evidence does not support continued investment). In manufacturing, “kill” decisions are harder because of sunk costs. A project with €600.000 invested feels different from one with €20.000 invested, even though the governance logic should be the same. Pre-set kill criteria, defined before any money was spent, remove the emotional weight.

Annual strategic rebalancing. Once a year, step back from individual projects and assess whether the portfolio allocation still matches the competitive environment. Has a new competitor entered your market? Has a technology shift changed the relevance of your explore bets? Have raw material prices shifted the economics of your core business? In the manufacturing companies I work with, annual rebalancing typically results in adding one or two new explore directions and retiring one that has become less strategically relevant.

Kill criteria: set them before the first euro

This is the single most important governance practice for manufacturing innovation portfolios, and the one most companies skip.

Every explore project should enter the portfolio with explicit conditions under which it will be stopped. Not “if it does not work” but specific, measurable thresholds:

• “If fewer than 4 out of 12 target customers agree to a paid field trial within 10 months, we stop.”
• “If the prototype cannot achieve target production cost within two iterations and €400.000 total tooling investment, we stop.”
• “If regulatory pre-assessment identifies a blocker that adds more than 24 months to market entry, we reassess viability.”

I worked with a chemical company that set kill criteria for an explore project developing a new coating process. The criteria: achieve target adhesion properties in lab testing within 9 months and €180.000. At month 7, the lab results were 30% below target with no clear path to improvement. The kill criteria made the decision straightforward. Without them, I am certain that project would have continued for another year on “we’re close, we just need more time.” That is the pattern I see at companies without pre-set criteria: explore projects that never succeed and never get killed, consuming budget and attention indefinitely.

For the full guide on kill decisions, see How to Kill Innovation Projects Without Killing Innovation.

The five patterns I see in every manufacturing portfolio

After 40+ portfolio reviews in industrial companies, the same patterns repeat. If you manage a manufacturing innovation portfolio, at least three of these will sound familiar.

Pattern 1: the “innovation” budget that is not innovation

A manufacturer reports that 15% of the product development budget goes to innovation. Sounds healthy. But when I map the projects in that budget, 12% funds cost reduction initiatives, product line extensions, and customer-requested customizations that engineering was planning anyway. The actual explore allocation, projects testing genuinely new business models, new markets, or new technologies, is 3%.

This is not dishonesty. It is classification confusion. Core improvements are important. But calling them “innovation” in the portfolio creates the illusion of balance where none exists. The fix: separate the portfolio into three clear categories (core, adjacent, transformational) with strict definitions. A project is only “transformational” if the target customer, the business model, or the core technology is new. Everything else is core or adjacent.

Pattern 2: the hero project that cannot be killed

Every manufacturing portfolio I have reviewed has at least one: an explore project championed by a senior leader, running for two or more years, consuming significant budget, with no validated customer demand. The project has survived multiple reviews because the sponsor is influential, the technology is genuinely interesting, and nobody wants to be the person who killed the VP’s initiative.

This pattern is more damaging than it appears. The hero project does not just consume budget. It blocks better ideas from entering the portfolio because “we do not have capacity for more explore projects.” It creates cynicism among teams whose ideas get rejected while the untested hero project continues. And it makes innovation governance look like theater.

The solution is structural: kill criteria set at portfolio entry, applied equally to every project regardless of sponsor seniority. When the governance system, not a person, makes the kill decision, the political cost disappears.

Pattern 3: explore reviewed like exploit

Portfolio review meetings where explore projects are evaluated using the same metrics as exploit projects. Revenue projections. Market share targets. Three-year financial forecasts. An explore project that is six months into a 14-month prototype development cycle has none of these. Reviewing it alongside an exploit project that delivered €3M in margin improvement makes the explore team look incompetent.

Separate governance for explore and exploit is not optional in manufacturing. The ambidextrous organization research by O’Reilly and Tushman showed that structural separation outperforms contextual ambidexterity. This extends to governance: explore projects need different metrics (learning velocity, evidence strength, assumption validation rate), different timelines (milestones measured in months, not weeks), and different reviewers (people who understand exploration, not just exploitation).

Pattern 4: engineering capacity invisible in portfolio decisions

The portfolio map shows five explore projects with adequate budget. But three of them share the same two senior engineers who also support the core product line. In practice, those engineers spend 80% of their time on exploit fires and the explore projects progress at a fraction of the planned pace.

I now include an engineering capacity overlay in every manufacturing portfolio review. Each project gets a “capacity score”: the percentage of required engineering time that is actually available and protected. A project with full funding but 20% capacity allocation is a project that will take five times longer than planned. Making this visible forces honest conversations about what the portfolio can actually execute.

Pattern 5: no connection between portfolio and individual business models

The portfolio map sits at leadership level. The individual business model work, Business Model Canvas sessions applied to manufacturing, Value Proposition Canvas work adapted for manufacturing, testing experiments, happens at project level. The two rarely connect. Portfolio decisions are made without seeing the underlying business model evidence. Project teams run experiments without knowing how their results affect portfolio-level decisions.

The fix: every explore project on the portfolio map should link to a Business Model Canvas that shows the model being tested, with clear markings of which assumptions are validated, which are invalidated, and which are untested. Portfolio reviews should reference this evidence directly. “Project A has validated 6 of 9 critical assumptions and needs €300.000 for the remaining three” is a portfolio decision based on evidence. “Project A is making good progress” is a portfolio decision based on nothing.

How to start: three steps this quarter

If you manage an innovation portfolio in a manufacturing company and recognize these patterns, here is where to start. Not a six-month transformation. Three concrete steps you can take this quarter.

Step 1: audit the real allocation. Map every project in your innovation budget. Categorize each as core, adjacent, or transformational using strict definitions. Calculate the actual percentages. Share the result with your leadership team without commentary. The numbers will speak for themselves.

Step 2: set kill criteria for every active explore project. Gather the project teams. For each explore project, define the specific conditions under which it will be stopped. Write them down. Get sponsor sign-off. This is the single highest-leverage action you can take because it converts a political decision into a governance mechanism. If your teams need a structured approach to generating the evidence that feeds these criteria, the Testing Business Ideas methodology provides the experiment toolkit.

Step 3: separate the next portfolio review. Run separate review sessions for explore and exploit projects, with different metrics and different expectations. Explore projects present evidence: assumptions tested, experiments run, results observed. Exploit projects present performance: revenue, margin, timeline adherence. Same governance board, different sessions, different criteria.

To assess whether your organization has the leadership support, organizational design, and innovation practice needed to run this kind of systematic portfolio management, start with the Innovation Readiness assessment. For the assessment adapted to industrial realities, see Innovation Readiness for manufacturing. Portfolio governance does not operate in a vacuum. It needs the organizational foundations to support it.

For the broader B2B context of managing innovation portfolios, including multi-stakeholder dynamics and long sales cycles, see Innovation Portfolio Management for B2B Companies.

For common failure patterns in portfolio management, see Innovation Portfolio Mistakes: Why Most Companies Fund the Wrong Projects.

For the strategic tension underlying every portfolio decision, see Explore vs Exploit: Balancing Innovation and Core Business.

Frequently Asked Questions

Does innovation portfolio management work for manufacturing companies?

Yes, but it requires significant adaptation. Standard portfolio management assumes cheap experiments and fast pivots. Manufacturing reality includes explore projects costing €500.000 or more in tooling, validation cycles of 12 to 18 months, and regulatory approval adding 6 to 18 months. The core principles, balance explore and exploit, set kill criteria, govern with evidence, apply directly. The implementation needs to match manufacturing economics: fewer explore bets with higher individual investment, longer milestone timelines, engineering capacity tracking alongside financial allocation, and a regulatory runway layer that software-focused guides never mention.

How should a manufacturing company allocate its innovation portfolio?

Start from the bottom up, not the top down. Instead of applying the 70-20-10 rule (from Nagji and Tuff’s HBR research), determine how many simultaneous explore projects you need for a reasonable success rate and what each one costs at minimum. That calculation gives you a concrete number to ring-fence, rather than a percentage that may not translate to a viable number of projects. Most capital-intensive manufacturers I work with end up between 5% and 10% for genuine transformational exploration, but the number matters less than the commitment: protected budget, CFO sign-off, board-level approval required to reallocate.

How do you set kill criteria for manufacturing innovation projects?

Before any explore project receives funding, define three types of stopping conditions. Market evidence: a minimum number of target customers who commit to a paid pilot or letter of intent within a defined timeframe. Technical feasibility: target performance specifications achievable within a defined number of prototype iterations and total tooling budget. Financial threshold: maximum cumulative investment before requiring board re-approval. Set these criteria when everyone is objective, before emotional and financial attachment develops. At each portfolio review, compare actual results against criteria. The decision framework was built when it was easy. Applying it should be mechanical.

What is the biggest innovation portfolio mistake in manufacturing?

Counting incremental improvements as innovation in the portfolio. When a manufacturer reports 15% innovation allocation but 12% of that funds cost reductions and product line extensions, the actual explore allocation is 3%. This classification confusion creates a false sense of portfolio balance. The fix is strict categorization: a project is only “transformational” if the target customer, the business model, or the core technology is genuinely new. Everything else is core or adjacent, regardless of how the project was labelled in the budget request.

How often should a manufacturing company review its innovation portfolio?

Three cadences. Monthly evidence reviews for active explore projects: what did you learn, not what did you do. Quarterly portfolio-level decisions: advance, pivot, or kill each project based on evidence against pre-set criteria, and assess overall portfolio balance. Annual strategic rebalancing: reassess allocation targets based on competitive shifts, technology changes, and what the portfolio has learned over the past year. This rhythm accounts for manufacturing’s longer validation cycles while maintaining the governance pressure that prevents explore projects from drifting without accountability.