Google 70–20–10 innovation model

1. What Is the Google 70–20–10 Innovation Model?

The Google 70–20–10 model is a simple portfolio rule of thumb for managing innovation. It says: allocate roughly 70% of resources to your core businesses, 20% to adjacent opportunities that extend the core, and 10% to transformational bets that could create new businesses. The goal is balance—continuously improve today’s engine while deliberately investing in tomorrow’s engines.

Within Agile, Innovation & Networked‑Organization frameworks, 70–20–10 provides a practical top‑down guardrail for enterprise portfolios. It complements team‑level methods (Design Thinking, Lean Startup, Scrum/Kanban) by ensuring the portfolio has the right mix of exploit (core) and explore (adjacent/transformational) work—and that each tier is governed appropriately for its uncertainty.

In plain terms: don’t let the urgent crowd out the important. Protect capacity for near‑term improvements, medium‑term extensions, and long‑term bets in a deliberate ratio—and adjust as you learn.

2. Origin and Background

Popularized by Google leadership in the mid‑2000s, the 70–20–10 split was used to describe how the company distributed resources across core services (e.g., search/ads), adjacent expansions, and new ventures. It is often discussed alongside—but is distinct from—Google’s “20% time” concept for individual employees. Over time, 70–20–10 spread beyond technology into consumer, industrial, and public‑sector organizations as a portfolio heuristic for balancing exploitation and exploration.

Origin note: The precise originator is not definitively attributed in public sources; the model has been referenced in talks and interviews by Google executives and in business press. Its core logic aligns with broader portfolio thinking (e.g., McKinsey’s Three Horizons; Innovation Ambition Matrix), but 70–20–10 was the memorable, operational shorthand that many executives adopted.

Why it emerged: organizations tend to over‑invest in low‑risk, incremental work because it has near‑term ROI and clear owners. 70–20–10 forces explicit capacity protection for adjacencies and transformational options that are essential for long‑run growth and resilience.

3. How the 70–20–10 Model Works

Think of 70–20–10 as a guardrail, not a law. The value is in the discipline it imposes on classification, governance, and resource allocation.

Define the Three Buckets

• 70% Core (Exploit): Enhancements to existing products, services, processes, and channels for current customers. Examples: feature improvements, cost reduction, reliability, incremental automation, pricing tweaks.
• 20% Adjacent: “Next to” the core—new customers with current offerings, or new offerings with current capabilities, often new channels or regions. Examples: moving from enterprise to SMB; bundling services for an existing customer base; adding a data product to a hardware install base.
• 10% Transformational: New to the company (often new to market) offers or business models that could create new S‑curves. Examples: platform plays, marketplaces, new categories, disruptive business models.

Allocate Capacity, Not Only Budget

• Apply the split to capacity (teams, talent, time) and funding. Protect headcount/time for 20% and 10% work, or the core will consume everything.
• Use ranges (e.g., Core 60–75%, Adjacent 15–25%, Transformational 5–15%) that reflect context and strategy.

Tailor Governance and Metrics

• Core: ROI/payback, margin uplift, cycle‑time reduction, reliability (SLOs). Quarterly reviews, standard business cases acceptable.
• Adjacent: Customer validation, willingness to pay, pilot‑to‑scale conversion, early unit economics (LTV/CAC proxies). Stage‑gated with learning milestones.
• Transformational: Venture‑style tranches tied to learning (desirability → viability → scalability), partner traction, option value. Govern through a venture board, not annual budget cycles.

Refresh the Mix Regularly

• Rebalance the portfolio quarterly based on evidence (outcomes, flow metrics, learning velocity) and the strategic context (competition, regulation, technology shifts).
• Shift capacity as bets graduate (10% → 20% → 70%) or are retired (“good kills” return talent to the pool).

4. When to Use the 70–20–10 Model

Most helpful when:

• Your portfolio is dominated by incremental work and long‑term growth options are thin.
• “Innovation” exists but runs through core ROI gates, dying early or drifting indefinitely.
• You need a board‑ready, communicable way to balance exploit and explore investments.
• You are moving to a product/value‑stream operating model and need portfolio guardrails across streams.

Especially powerful: In mid‑to‑large enterprises with multiple value streams and a strong core franchise at risk of disruption; in scale‑ups nearing S‑curve flattening; and in public entities balancing mission reliability with modernization.

Less suitable or potentially misleading:

• As dogma. Market context, strategy, and capability should shape the split; 70–20–10 is a starting point.
• As a poster without changing governance and talent models; numbers alone don’t protect exploration.
• When used to justify pet projects re‑labeled as “adjacent” or “transformational” without evidence.

5. How to Apply 70–20–10: Step‑by‑Step

1. Anchor on strategy and guardrails.

   Define your 2–5 year growth story (defend/extend core; enter specific adjacencies; build new platforms). Set initial allocation ranges (e.g., 65–75% core; 15–25% adjacent; 5–15% transformational) aligned to strategy and risk appetite.

2. Define classification criteria.

   Create a short rubric (score 1–5) on:

   • Customer/segment novelty.
   • Offering/business model novelty.
   • Capability/technology novelty.
   • Market/competitive uncertainty.
   • Time to impact.

   Agree cutoffs for each bucket (core/adjacent/transformational) and calibrate with a cross‑functional panel.

3. Inventory and classify current work.

   List inflight and proposed initiatives; normalize scope and investment; classify using the rubric. Expect to find BAU items masquerading as “innovation”—remove them from the portfolio counts.

4. Set targets and reallocate capacity.

   Compare the current mix to targets; reallocate budget and people accordingly. Reduce low‑value core backlog; fund a few high‑quality adjacent/transformational bets with clear learning plans. Protect exploration capacity (teams, time).

5. Install tier‑specific governance.

   Document how decisions are made for each bucket:

   • Core: Quarterly ROI gates; product/GM ownership; operational KPIs.
   • Adjacent: Stage gates tied to demand validation and early unit economics; product leadership with finance support.
   • Transformational: Venture board; tranche funding (6–12 weeks) tied to learning milestones; option‑value logic.
6. Pair with the right methods and metrics.

   Core: continuous improvement, DevOps, A/B testing; metrics—margin, cycle time, reliability. Adjacent: Double Diamond + Build–Measure–Learn; metrics—problem/solution fit, pilot conversion, willingness to pay. Transformational: business model experimentation, ecosystem partnerships; metrics—validated learning, partner traction, early retention.

7. Wire into funding and planning cycles.

   Move from annual, project‑based funding to rolling, outcome‑based capacity allocation (Lean Portfolio Management). Use OKRs to express targets per bucket; rebalance capacity quarterly.

8. Build enabling structures and talent.

   Create an incubation “foundry” or venture studio for the 10%; enable adjacent work in value streams; keep stable product teams for core. Define rotational pathways; recognize learning velocity (explore) and reliability (exploit) differently.

9. Manage interfaces and cannibalization.

   Set sandbox rules (brand, channel, pricing) for exploration. Pre‑agree cannibalization thresholds and incentives tied to enterprise outcomes, not just core P&L, to avoid antibody reactions.

10. Review and refine quarterly.

    Hold a portfolio review: start/stop/scale; check mix vs. targets; publish outcomes and learning. Evolve the split with context (e.g., increase transformational during platform shifts; dial it down if cash‑constrained short‑term).

6. Example: 70–20–10 in Action

Context: A 4,500‑employee global logistics firm with a strong parcel business faced margin pressure and new digital competitors. The portfolio skewed toward core ops projects; ambitious digital services languished. Leadership adopted 70–20–10 to rebalance and communicate the growth plan.

Application:

• Guardrails: Target mix set at Core 70%, Adjacent 20%, Transformational 10% for the next 12 months. Strategic goals: +200 bps margin in core; $50M new revenue from services in 24 months.
• Classification: 112 initiatives inventoried; 86% of spend was core. BAU items (fleet maintenance, compliance updates) were removed from the innovation tally.
• Reallocation: Pruned 18% of low‑impact core items. Funded adjacencies: SMB self‑serve shipping bundle, cross‑border customs concierge, and a partner API product. Transformational bets: on‑demand urban micro‑fulfillment (pilot in two cities) and a marketplace for spare capacity with third‑party carriers.
• Governance: Venture board for 10% (8–12 week tranches tied to demand signals and unit economic proxies); stage‑gates for 20% (willingness‑to‑pay, pilot conversion); quarterly ROI gates for 70%. OKRs set per bucket; dedicated 10% capacity (50 FTEs) ring‑fenced.
• Enablement: Launched a lightweight internal developer platform for APIs; created a small incubation team (product, design, engineering, legal) to accelerate transformational pilots; established sandbox rules for brand and pricing.

Outcomes (9–12 months): Portfolio mix shifted to Core 68%, Adjacent 22%, Transformational 10% by capacity. Core margin improved 160 bps (route optimization, automation); SMB bundle reached $14M ARR run‑rate with sub‑12 month payback; customs concierge conversion 27% in pilots; micro‑fulfillment showed 2.1× faster delivery in pilots with promising unit economics; capacity marketplace was killed after weak liquidity despite incentives—talent redeployed to the API product. The board gained confidence from transparent “stop/scale” decisions and the visible balance across horizons.

7. Strengths and Limitations

Strengths

• Clarity and balance: A simple, shared language to avoid over‑indexing on the core or on moonshots.
• Protection for exploration: Capacity guardrails prevent the core from consuming everything.
• Fit‑for‑purpose governance: Different metrics and cadences improve decision quality and throughput.
• Communication power: Board‑ and employee‑friendly story about where growth comes from and how it’s funded.

Limitations

• Coarse heuristic: 70–20–10 is a starting point; context should drive the mix.
• Label gaming risk: Without a rubric, teams may re‑label initiatives to access funds.
• Execution dependency: Requires changes in funding, governance, and talent models; ratios alone don’t move outcomes.
• Short‑term constraints: Cash or regulatory shocks may force temporary deviations—communicate and restore balance deliberately.

8. Common Pitfalls (and How to Avoid Them)

• Treating 70–20–10 as dogma.
  What goes wrong: Inflexible targets misfit strategy or context.
  Avoid by: Using ranges and revisiting quarterly; align to strategy, risk, and capability.
• Applying core ROI gates to transformational bets.
  What goes wrong: Early kills; option value lost.
  Avoid by: Venture‑style tranches and learning milestones; shift to economics as uncertainty falls.
• Counting only budget, not capacity.
  What goes wrong: People time is still consumed by the core; explore starves.
  Avoid by: Ring‑fencing teams and time for adjacencies/transformational work.
• BAU disguised as innovation.
  What goes wrong: Portfolio appears balanced; impact is not.
  Avoid by: Excluding BAU from the counts; enforce a clear classification rubric.
• “Pet projects” in the 10% bucket.
  What goes wrong: Low‑quality bets crowd out real options.
  Avoid by: Requiring hypotheses, learning plans, and venture board approval with kill/scale rules.
• Conflating 70–20–10 with “20% time.”
  What goes wrong: Confusion between portfolio mix and individual time allocation.
  Avoid by: Communicating clearly: 70–20–10 is about portfolio resources, not necessarily personal free time.
• No path to scale.
  What goes wrong: Pilots succeed but die at hand‑off.
  Avoid by: Defining scale owners, budgets, and KPIs from the start; design migration paths (10% → 20% → 70%).

9. How 70–20–10 Relates to Other Frameworks

• Innovation Ambition Matrix (Core–Adjacent–Transformational): 70–20–10 is a practical allocation overlay. Use the matrix for classification and governance; use 70–20–10 to set portfolio capacity targets.
• Three Horizons (H1–H3): Time‑based lens for growth. Map 70–20–10 to horizons (e.g., 70% H1, 20% H2, 10% H3) to communicate pacing and ambition.
• Ambidextrous Portfolio (Explore–Exploit): 70% aligns to exploit; 20%/10% to explore. Use explore governance (learning milestones) for the 20%/10% buckets.
• Lean Portfolio Management (LPM): The operating mechanism for quarterly rebalancing, capacity shifts, and OKR alignment; 70–20–10 sets the guardrails LPM implements.
• OKRs: Express outcomes for each bucket; review quarterly to adjust the mix and funding.
• Ten Types of Innovation: Ensure adjacent/transformational bets use multiple types (profit model, network, channel, service), not just features, to increase defensibility.
• Design Thinking / Lean Startup: Methods for the 20%/10% buckets to validate desirability, viability, and feasibility quickly.

10. Key Takeaways

• 70–20–10 is a portfolio guardrail: ~70% core, ~20% adjacent, ~10% transformational—adjusted to strategy and context.
• Apply it to capacity and funding; protect exploration with dedicated teams, venture‑style governance, and learning milestones.
• Use a clear rubric to classify; exclude BAU; rebalance quarterly based on outcomes and evidence.
• Pair with Lean Portfolio Management, OKRs, and explore/exploit governance to turn ratios into results.
• Communicate clearly: 70–20–10 is not “20% time”; it’s how the enterprise balances today’s engine with tomorrow’s.

11. FAQs About the Google 70–20–10 Innovation Model

Is 70–20–10 still relevant today?
Yes—as a starting point and a communication device. In fast‑moving or disruptive contexts, you may temporarily tilt more to adjacencies/transformational work; in cash‑constrained periods, you may lean into the core. The value is in deliberate balance and quarterly rebalancing, not the exact numbers.

How is this different from Google’s “20% time” for employees?
70–20–10 is a portfolio allocation across core/adjacent/transformational work. “20% time” refers to an individual practice allowing employees to spend a portion of time on self‑directed projects. Some companies do both; they are independent decisions.

Should we enforce the same split in every business unit?
No. Use enterprise‑level targets and allow value‑stream or BU‑level variation based on maturity and strategy. For example, a nascent product might run 50–30–20, while a regulated, stable business might run 80–15–5.

How do we measure ROI for the 10% transformational bucket?
Early on, use learning milestones and option‑value logic (validated demand, willingness to pay, partner traction). As uncertainty declines, shift to unit economics and NPV. Avoid forcing precise ROI too soon—it biases against valuable options.

Do we apply 70–20–10 to budget or people?
Both. Apply to funding and capacity (teams, time). If you only move budget, the core will still consume people’s time. Ring‑fence teams for adjacent/transformational work and protect their time.

How often should we rebalance?
Quarterly is typical: review outcomes and evidence, make start/stop/scale decisions, and adjust capacity. In dynamic markets, do monthly light reviews with a deeper quarterly reset.

What capabilities are prerequisites?
Lean portfolio practices, OKRs, and basic product operating model discipline help. For the 10%, you’ll benefit from an incubation function (venture board, experimentation capability) and enabling platforms (billing, data, DevOps) to test quickly.

Can small or early‑stage companies use 70–20–10?
Yes—lightly. With a handful of bets, classify them and set a simple mix (e.g., 80–15–5). Protect at least some capacity for the “10%” to explore the next S‑curve. Review monthly; keep governance lightweight.

How do we prevent label gaming?
Use a written rubric, cross‑functional calibration, and an “evidence score” per initiative. Require hypotheses and learning plans for 20%/10% items; exclude BAU from counts; publish the portfolio map for transparency.

What if cash is tight this year?
You can temporarily bias toward the core, but make it explicit and time‑bound. Maintain a minimal transformational pipeline (small tranches) to preserve option value, and plan to restore balance when conditions improve.