Prioritizing product features for a multi-product portfolio requires a two-level decision-making structure: portfolio-level resource allocation across products and product-level feature prioritization within each product — with explicit rules for when platform investments and cross-product features override individual product roadmaps.
According to Lenny Rachitsky on Lenny's Podcast, multi-product companies often fail not because their products are bad, but because the resource allocation between products is driven by politics rather than strategic priority — and the strongest internal advocate wins the budget, not the highest-impact investment.
According to Gibson Biddle on Lenny's Podcast, Netflix's portfolio management principle was clear: maximize the happiness of the subscriber base, not the growth of any individual product team. When teams optimized for their own metrics, the overall subscriber experience suffered.
According to Chandra Janakiraman on Lenny's Podcast, the biggest risk in a multi-product portfolio is premature platform investment — building shared infrastructure before the individual products have proven their own product-market fit separately.
The Two Levels of Multi-Product Portfolio Prioritization
Multi-Product Portfolio: A collection of distinct products under one company umbrella, each with its own customer segment, value proposition, and success metrics — but sharing engineering resources, brand equity, and platform infrastructure.
Level 1: Portfolio-Level Resource Allocation
This is the CEO/CPO decision: how do you allocate engineering headcount across products?
The BCG Matrix Adapted for Product Portfolios:
- Stars (high growth, high strategic fit): Overinvest — these are the future
- Cash Cows (low growth, high revenue): Maintain with minimal investment — don't starve them but don't grow them
- Question Marks (high growth potential, unproven): Time-box investment — set a 2-quarter test, then promote or sunset
- Dogs (low growth, low strategic fit): Deprioritize or sunset — resource drain with no future
Level 2: Product-Level Feature Prioritization
Within each product, use standard frameworks (RICE, WSJF) but constrain by the portfolio allocation:
- Star products get uncapped prioritization discretion — build boldly
- Cash cow products apply a stricter cost-of-change filter — only ship high-confidence improvements
- Question mark products focus exclusively on the core hypothesis test — no feature sprawl
Handling Cross-Product Feature Requests
The most contentious decisions in multi-product portfolios are features that span products. A three-question test:
Q1: Is this a platform capability or a product capability? Platform capabilities (authentication, billing, notifications, data pipelines) belong in shared infrastructure. Product capabilities (workflow-specific features, domain logic) belong in individual products.
Q2: Does this accelerate all products or just one? If only one product benefits, it goes on that product's roadmap, not the platform's.
Q3: Would customers pay for this as a standalone? If yes, it might be a new product, not a feature of existing products.
The Platform Investment Decision Framework
Platform investments are high-effort, high-risk bets. Only build shared platform capabilities when:
- At least 2 products need the same capability (duplication cost justification)
- The capability is not a product differentiator (if it's differentiating, keep it product-specific)
- Both products have proven product-market fit (don't platform-build before PMF)
- There's a dedicated platform team (platform work done by product teams creates organizational debt)
Resource Allocation Models
Fixed allocation: Each product gets a fixed % of engineering headcount per year. Predictable but inflexible.
Dynamic allocation: Resource allocation is reviewed quarterly based on portfolio performance. More accurate but creates uncertainty within teams.
Investment-return model: Each product receives resources proportional to its projected return on investment. Most rational but requires rigorous financial modeling.
Common Pitfalls to Avoid
- Treating every product as a star: Resources spread equally produce no stars — portfolio thinking requires accepting that some products get less
- Letting product teams negotiate for resources individually: This creates a political allocation process; resource decisions should be made at the CPO/CEO level with a clear framework
- Premature platformization: Building shared infrastructure before products have PMF creates abandonment risk — you're building platform for products that may be sunsetted
- No sunset process: Multi-product portfolios accumulate zombie products. Without a clear sunset process, dog products drain resources indefinitely
Success Metrics for Multi-Product Portfolio Prioritization
- Star products grow faster than the market
- Cash cow products maintain revenue with <15% engineering allocation
- Platform investments serve >2 products within 12 months of completion
- No more than 15% of engineering is allocated to dog products
For more, visit PM interview prep and daily PM challenges.
Learn portfolio strategy from Lenny's Newsletter.
Frequently Asked Questions
How do you prioritize features across a multi-product portfolio?
Use a two-level framework: portfolio-level resource allocation (BCG matrix adapted for products) and product-level RICE/WSJF scoring within each product's allocation. Portfolio allocation is made at CPO/CEO level; feature prioritization is delegated to product teams.
When should a multi-product company invest in a shared platform?
Only when at least 2 products need the same capability, it's not a product differentiator, both products have proven product-market fit, and there's a dedicated platform team to own the investment.
How do you handle feature requests that span multiple products?
Apply the three-question test: Is it platform or product? Does it accelerate all products or just one? Would customers pay for it standalone? Platform capabilities go to the shared roadmap; single-product capabilities stay on that product's roadmap.
What is the BCG matrix and how does it apply to product portfolios?
The BCG matrix categorizes business units by growth rate and market share. Adapted for product portfolios: Stars (high growth, high strategic fit), Cash Cows (mature, high revenue), Question Marks (unproven high-growth), and Dogs (low growth, low fit). Each category gets different resource allocation.
How do you sunset a product in a multi-product portfolio?
Define sunset criteria before launch: if the product doesn't reach X metric by Y date, sunset it. Communicate to customers 90 days in advance. Migrate data, offer refunds, and document learnings. Freeing resources from dog products is how portfolio companies fund their next star.