Radical Innovation and R&D Investment

Kodak engineer Steve Sasson built the first digital camera in 1975. The invention was inside the company. But the company earned its money from film and did not know what to do with it. Radical innovation stories often start like this: there is no shortage of ideas. What is missing is a structure to carry the idea and a measure that encourages it.

Radical innovation does not mean doing your current business a bit better. It means trying something that questions the assumptions of your current business. The investment this takes is not only money; it is a different management logic.

When you need it

If your core business is still growing but the growth rate is slowing, this is the best time for radical investment, because you have both cash and time. Starting when the crisis arrives usually means starting too late.

Other signals: a technology is maturing in your industry that questions the basic assumptions of your product (a cost curve has collapsed, a new infrastructure has become possible), new “jobs” are emerging that customers no longer hire you for, or your best engineers are starting to leave because they are full of idle ideas.

It can also be too early. If your core business is not solid, your cash flow is fragile, or the board has no patience for a bet that shows no results for six quarters, radical R&D investment weakens the company. In that case it makes more sense to learn through small, bounded experiments first.

What most teams miss

An R&D budget does not produce radical innovation; radical innovation comes out of a portfolio managed with different rules. Applying the core business's measures (first-year revenue, margin, ROI, predictability) to a radical project kills it early, because a radical project brings neither revenue nor margin in its early years. What it brings is learning.

The Three Horizons framework organizes this well: horizon one is today's business, horizon two is growing new businesses, horizon three is discoveries that have not yet turned into revenue. Each horizon has to be managed with different measures. As a common rule of thumb (who first proposed it is debated, but it is often quoted), roughly 70 percent of resources go to the core, 20 percent to adjacent businesses and 10 percent to real exploration. Ratios vary by industry; what matters is that the share for horizon three is treated as a protected budget, not a line to be cut.

The story of AWS is an example of a radical bet born from a core capability. Amazon had developed a large infrastructure capability to scale its own business. Offering that capability to others as a service did not look like a retailer's business at the time, but a new line of business emerged. The lesson: radical innovation is often not an idea from zero but your existing capability meeting a new market.

Managing a radical project with the core business's measures kills it before it starts.

The playbook

1. Define radical bets as a separate portfolio

Set up a budget, leadership and measures separate from the core. This portfolio should have its own board discussion and its own reporting rhythm. The core's monthly results should not decide this portfolio's fate.

2. Separate risk by type

For each bet, write which uncertainty is largest: is the technology possible, does the customer want it, does the business model leave a profit, can the company execute it? Decide which of these questions must be answered first. Start with the biggest risk that can be answered most cheaply.

3. Fund in stages, set criteria

Instead of giving all the money up front, open the next tranche when a defined learning criterion is met at each stage. This is the “real options” logic: buy an option for a small price and increase investment as uncertainty falls. Criteria should be about learning, not revenue: for example, “three of five pilot customers are using it and paying with their own money.”

4. Write the kill criteria up front

For each bet, write the sentence “if this evidence does not arrive by this date, we stop.” Learning from a stopped project is not a failure, it is the portfolio's output. Projects without kill criteria turn into ghosts that drain resources from the core.

5. Plan the bridge to the core

If the radical bet succeeds, how will it connect to the core's customers, supply chain and sales team? When is it a separate business unit, when does it integrate into the core? Debate this before the project succeeds, not after; otherwise the core rejects it like an organ transplant.

What to measure

• Stage-gate pass rate: How many bets move to the next stage and how many are stopped? If none is ever stopped, the portfolio is not bold enough or not honest enough.
• Learning speed: The time and cost to resolve an uncertainty. A leading indicator.
• Early customer commitment: Pilot customers paying or committing to extensive use. A far more valuable signal than opinions.
• Share of revenue from horizon three (long term): A lagging measure that shows in a few years whether the portfolio works.

Pitfalls

ROI pressure. A finance process that computes first-year return is the earliest destroyer of a radical bet. A separate measure set is a must.

Doing everything at once. Resources spread across ten bets cannot feed any of them enough. Run a small number of bets with enough resources but clear kill criteria.

Innovation theater. Hackathons, showcase labs, press releases; but if at the end there is no business unit, customer or revenue, it is only a show. At the end of every initiative there should be a sponsor and a customer question.

Start this week

1. Split your company's R&D and new-business spending into three horizons on one page and see how much goes to horizon three.
2. Pick your three largest ongoing new projects and write for each the “biggest uncertainty” and the “kill criteria” sentences.
3. Ask the board what measure these projects are judged by; if the answer is “revenue and ROI,” prepare a proposal for a separate measure set.