Beyond the Balance Sheet
Asking a manager how much they intend to invest in innovation often sounds like a routine accounting query, yet it is fundamentally a question of character. The answer reveals the company's true stance on the future, its tolerance for uncertainty, and the extent to which staff are empowered to make mistakes. A figure stated in isolation, devoid of this context, tells us very little.
The flaw in the question, as it is typically posed, lies in both the verb and the noun. The verb, invest, implies a predictable return, whereas innovation is largely the act of purchasing information about the unknown. The noun, innovation, is a term so broad that it encompasses everything from painting a boardroom to reinventing an entire business model. Those who decide without distinguishing between these two ends up squandering resources without understanding what they have actually acquired.
Rethinking the Mechanism
This perspective proposes reversing the usual order. Instead of starting with the budget, we must start with the mechanism: what do we wish to learn, what bet are we willing to lose, and who is overseeing the experiment? Only then does it make sense to discuss the funding.
Routine expenditure exists to maintain what is already functional. Innovation expenditure exists to discover if something might one day become functional. These are distinct logics requiring different evaluation criteria. Measuring an experiment by immediate returns is akin to judging a seed by the harvest of the following week.
Consider a family-run distribution firm that suspects its younger clientele shops differently. They might commission an extensive study, launch a small pilot at a single location, or do nothing and wait for the suspicion to either be confirmed or fade away. Each path carries a cost, including the last one, whose price is invisible and therefore dangerous: the cost of inertia never appears on an invoice.
The Tolerance Threshold
The rule of thumb is straightforward. Before approving any funds, write down in a single sentence the question the experiment is intended to answer. If the sentence cannot fit on one line, the idea is not yet mature enough to receive funding. And if the only possible answer is yes, the experiment is merely decoration, not learning.
There is a useful way to think about the value to set aside: not as an ambition, but as a tolerance. What is the loss the company can sustain without compromising its core activities, salaries, or banking relationships? This is the real ceiling, far more honest than any figure copied from textbooks or competitors.
This leads to an uncomfortable consequence. Companies with narrow margins have less room for error and should therefore bet on smaller, more frequent experiments rather than one grand gamble. Companies with financial leeway face the opposite risk: spending far more than they learn because abundance often discourages rigour.
Attention as the Scarcest Resource
It is also vital to distinguish between money and attention. Management attention is the scarcest resource and rarely enters the calculation. A cheap experiment that consumes half of the managing director's schedule may be more expensive than a financially heavy project that runs independently. Decision-makers must ask not only what it costs, but who will be left unable to perform their duties while the experiment takes place.
Most innovation initiatives fail for organisational, not technical, reasons. They fail because the proposer lacks the authority to decide, because the decision-maker is disconnected from the field, or because no one knows when an experiment should conclude. The money is present, but the mandate is absent.
Defining Success and Failure
A useful rule is to define three things before starting: who is responsible, what would be a clear signal that the idea should continue, and what would be a clear signal that it should stop. The latter is the most difficult, as it requires accepting in advance that one might be wrong. Without a termination criterion, weak projects survive out of sympathy and consume the space meant for the strong.
There is also a cultural dimension. If the team perceives that failing in an authorised experiment carries the same reputational cost as failing in routine tasks, they will quickly learn not to experiment. Innovation requires a clear distinction between negligence and exploration, and that distinction only exists if leadership practices it, rather than merely proclaiming it.
The Collective Laboratory
The decision to innovate rarely remains within the walls of the firm. When a company learns something new, suppliers adapt, workers acquire skills, and other local businesses observe and imitate. In a territory with few companies willing to experiment, each one that does acts as a laboratory for the rest.
This changes the perspective for those managing a municipality, a business association, or a region. The question shifts from how much companies are investing to what conditions facilitate experimentation: access to qualified people, networks where failures are shared, and a local market willing to test novelties. A territory that punishes error, through bureaucracy or distrust, pushes companies toward safe imitation and stagnation.
In geographies with close economic ties, learning also circulates through cultural and linguistic proximity. A solution tested in one place can be adapted in another with lower costs than starting from scratch, provided there are channels for the conversation to occur.
Ultimately, the question that should accompany any budget remains: if your company were to lose, in its entirety, the budget reserved for innovation, what would you have learned regardless, and what would remain unknown? The answer to this question is worth more than the figure that precedes it.