What Innovation Teams Should Focus On Before the Year Ends

At the start of the new season at Novum, members shared their priorities and the challenges they are facing until the end of the year. Here's what five experts suggest they focus on.

One team keeps watching well-evidenced ideas die late in the funnel. Another is building portfolio governance from scratch, starting with data that currently lives in slide decks and people's heads. Other members are trying to establish foresight as a credible, ongoing practice rather than a one-off exercise. A one-person innovation team is trying to get the rest of the organization excited about adopting new technology. And several members are wrestling with the same question underneath all of it: how to move innovation from a separate function into the way the organization actually works.

Five experts working across strategy, foresight, portfolio management, R&D, and business building responded with the actions they suggest the teams make until the year ends. And as that often goes, we start with a conversation with leadership.

Kyle Edmonds, Engagement Manager at Disruptive Edge, has watched the same scenario play out many times: a strong idea with good evidence moves far down the funnel and then dies, because the company's strategy changed along the way. Priorities shift, or leadership's view of 2030 changes, and the idea no longer fits.

To prevent that, he recommends taking three questions directly to company or business unit leadership, since they own the strategy the ideas will be judged against: where the team will play within the organization, how winning will be measured, and what is out of scope for the team. Kyle considers the third the most important. A team can have strong answers to the first two and still lose the idea, because leadership had already ruled that area out. Knowing those boundaries before investing months of work keeps good ideas from dying late in the funnel.

He suggests also to spend less time on ideation, which AI now handles well, and more on the decisions in the middle and back of the funnel. The capability most teams need now is to prioritize and execute as fast as they can generate ideas.

Longer term, Kyle expects the role of the innovation team itself to change. As AI reshapes how organizations research, test and launch, the opportunity is for innovation teams to build and orchestrate the workflows and guardrails that let the rest of the organization innovate, along with the learning loops that show each business unit what worked and what didn't.

Most portfolio conversations start in the middle. Teams talk about individual projects, the funnel, and the process that moves ideas through it, but spend far less time on what the full set of projects should look like at the end.

When nobody has defined that end state, time pressure shapes the portfolio instead. Small, incremental ideas are faster to deliver, and AI tools make them faster still. Doug Williams, Associate Director at SmartOrg, sees real value in that work: it saves time and money. The risk comes when quick wins become the whole plan. He recently met a company with exactly that problem. It had plenty of incremental work and an execution plan to deliver it, but no process in the middle to develop the bigger ideas that would close its growth gap.

The fix is to start with the portfolio and work back to the process. The most important conversation in this step is with leadership, agreeing on the company's growth goals. That target determines which processes and data the team needs. Quick wins keep a defined share of the portfolio, and the bigger bets get protected room. The same logic applies to individual ideas. Doug encourages teams to start with the big vision, meaning how this could be incredibly successful, and scale back as they learn.

Doug also sees a risk in how much teams now lean on AI to move faster. The more routine it gets, the easier it is to hand over more of the thinking than intended, which has been called cognitive surrender. His advice is to be deliberate about the balance: which AI outputs the team can trust as they are, which ones it needs to validate itself, and keeping the decisions with people.

Mark Putnam, Chief Strategy Officer at NewEdge, sees most foresight efforts lose their purpose along the way. Teams collect trend reports, turn them into slides, and lose track of why they're doing it. The purpose is to create urgency and get people aligned on where the world and the industry are going. You pick your clothes and video background for a call based on who'll be on it. Organizations make better choices the same way, when everyone expects the same future. When each person expects a different one, people pull in different directions.

That alignment matters most in planning. Too often, Mark says, strategy gets built from the bottom up: start with last year, pick the opportunities, then find trends to justify them. Foresight should come first, and it should look beyond the industry to adjacent shifts. For a healthcare company, that could mean how homes are changing to house people differently, or how energy costs affect the industry.

To keep foresight alive between planning cycles, Mark makes it a habit. Many meetings open with a safety moment, and he proposes adding a trends moment: five minutes at the start where someone shares what they're paying attention to.

Foresight also shouldn't be a few people going off on their own. A week at CES is great, Mark says, but it has more impact when supply chain, manufacturing and the business units take part, and when the innovation team turns their time into an output that helps them think more strategically about their own business.

AI implementation came up repeatedly in the members' challenges, from scaling beyond a first pilot to getting the wider organization on board. Suzanne Allers, Co-founder and Partner at Untapped Innovation, sees many of these efforts start from a top-down request rather than a user need: people are told to use AI or put an AI system in place, and get credit for showing where they've used it. Her advice is to push back and ask leadership to define exactly what they expect.

From there, she applies a product development approach. Treat the organization as the user base, define the job AI needs to do for those users, and design the rollout around that job. What she'd spend less time on is new systems and processes, which people resist when they can't see what's in it for them.

Getting people on board depends more on shared language than on process. Suzanne often sees teams believe they've communicated something that the other side understood differently. One framework she uses is a simple narrative: who you're serving, what frustrates them today, what's getting in the way, what the change would do, and how you'll measure the difference. Once everyone agrees on that story, less time goes into debating the basics and less work gets redone. That matters, because rework costs more than time. Every revisited decision makes stakeholders a little less convinced the opportunity is real.

Innovation teams are increasingly asked to take the company beyond its current business and operating model, whether the goal is scaling, new business, or AI. The default way to fund that work is the one-shot pitch. A team presents the whole new business at once (the customer, the problem, the business model, the value it delivers) and asks for full funding.

Frank Mattes, Founder and CEO of Lean Scaleup, recommends stopping that practice. In territory this uncertain, a more reliable route starts with a list of what must be true for the plan to work. Each statement then becomes a step in a sequence of smaller commitments, such as the next 100k in funding or the next salesperson the team can bring in, and the team answers the relevant "what must be true" questions before moving to the next stage. Large companies make decisions slowly by design, because they're built for reliability and predictability, and they're already overloaded with change. Bite-sized commitments work with both of those realities.

Each commitment also depends on the right kind of stakeholder support, and Frank distinguishes three levels.

Buy-in is the minimum, but on its own it rarely moves anything.

Sponsorship goes further, with a stakeholder co-funding parts of the work or lending a couple of people from their team. The level he aims for is skin in the game, where the innovation solves a problem on the stakeholder's own agenda, one that's mission-critical for their function or tied to their bonus. At that point, a small test becomes a story they want to hear: here's what we did, and here's how it solves your problem.

A successful pilot creates a similar trap, and Frank's second recommendation is to stop treating one as proof that scaling is solved. Pilots often run under conditions the organization can't reproduce at scale. Management pays close attention, a dedicated team works on the project, and specialists join calls to solve problems as they come up. Frank's suggestion is to treat scaling the same way: as its own list of must-be-true statements, tested during the rollout rather than assumed from the pilot.

Ask yourself one question: which of your current challenges would make the biggest difference if you solved it before the year ends?

Start there, and give it your full attention. If you’d like to compare notes, members of the Novum Society work through their current challenges with peers in the same boat.