You lead people through the presentation while simultaneously reading the reaction of the room. Here’s why it fits in well with current business strategy. Here’s proof of adoption via a small internal dogfood. Did the VP Engineering just nod her head? Awesome! Wait, did the VP Marketing just put his index finger to his lips and write something on a Post-it? What does that mean? Overall though, you’re feeling pretty good as you wrap up the demo and prepare to take questions.

The CEO leans back in her chair, looks straight at you and delivers a grim verdict, “I don’t know, I just don’t see this as a billion dollar business. And you know at our scale we need billion dollar businesses in order to grow.”

Discussions of innovation challenges at large companies usually focus on people, politics and culture. Trouble retaining talent. Too many middle managers. A culture which starts to reward steady hands as opposed to risk taking. 

Indeed those attributes might be found at some large companies and are corrosive to innovation, but there’s another drag on innovation which is equally insidious: the law of large numbers which causes managers to try and size market opportunities prematurely. This tendency comes from a reasonable place, namely an opportunity cost issue: if you are going to dedicate resources to a new project area you’ll want to know that it’s likely to “move the needle” in terms of overall company success, otherwise you can always get a higher nearterm ROI from putting those same resources on an existing business.

But market sizing forces you to understand the world as it exists today as opposed to how it might look tomorrow. And forecasts into the future are exercises in assumptions contingent upon any number of compounding hypotheses. I’m not suggesting market sizing is a useless gambit, just that it’s actually more important to be directionally correct than truly accurate. Come away from the work understanding what you believe to be true about the future and just get within an order of magnitude.

My proposed cure to premature asking “is this a billion dollar business” is improved incubation and project lifecycle management within large companies. Don’t prematurely try to size a market – start building a product and see if it works. Then you figure out what market you’re actually in.

1) Overall Resource Allocation framework
Just as Google traditionally allocated a percentage of their resources to core, extension and “totally new” areas, larger companies focused on innovation need to protect a percentage of resources to push into new business lines. This is different than R&D (or at least I think of it differently. I guess there could be way to have both ‘core R&D’ and ‘new venture R&D’). It’s about realizing that while the next dollar, heck the next billion dollars, may come from businesses you’re already in, that the incremental billion dollars will not. And the current business unit leads who are being judged on today’s opportunities will almost always put new resources on growing current business – it’s the bird in the hand. Accordingly, it’s really a CEO/upper management job to manage this overall distribution.

2) Progress Review framework
Ok, so once senior leadership blesses the overall allocation, you need a way to solve for which ideas get resources. There are a number of approaches – top down, bottom up. Acquihires vs internal developments – but as this isn’t a Harvard Business Review article, let’s hold off too detailed a discussion. Google’s “20% time” approach is one well-documented example, but has never been the only way the company makes expansive bets.

It’s helpful here to understand that projects have distinct stages and that your company should come up with a taxonomy which works for your own culture and business. It can be as simple as a single classification which says “these are new projects until they’re not” (ie they become formally funded). Or you can segment more granular into something like “early stage,” “product dogfood,” etc. Main goal is to allow CEO/senior leaders to have an air traffic controller map of incubating projects and understand their stage of development.

One note though – really important here to avoid too much management supervision – provide quick help or escalation paths but not executive hovering. There will be many people who want to be “helpful” but the whole idea here is to let small teams working quickly to prove out new ideas. Don’t overcrowd them. Set up milestone check-ins and then largely leave them alone. Use a champion or sponsor model but too many VPs turn into design sessions, business reviews, powerpoint decks, etc. The mantra here should be “demo or die” not “discuss to death.”

*Bonus* Reality Check: Don’t pretend these early stage internal projects are “start-ups”
Calling internally funded projects “startups” is the equivalent of pretending dry humping is sex (yes, I just went there). Don’t try to replicate market forces by creating strange valuation mechanisms. Don’t ban them from the company cafeteria in hopes of inspiring a “ramen profitable” mindset. Not only is it unnecessary complexity within the company, it also won’t hold water externally. Very tough to get the press and potential partners to see these projects as separate entities – and for good reason, they’re not. Similarly, BigCo cannot take the brand or legal risks of an entity they fully own. Increasingly in cases where the new venture needs to be separated in order to execute, I think we’ll see examples of parent BigCo investing in the entity as a financial equity owner rather than keeping in-house. [I obviously think the relationship between Google & Google Ventures is a strong example of how to do this – with examples like Craig Walker’s funding by GV]

Concluding, large companies often get themselves into trouble by trying to assess the market potential of new projects prematurely. Instead ensure a percentage of company resources are working on projects that may have great potential, and be rigorous about funding or killing decisions, rather than over-analyzing before anything has been built.