Power, Fit, and Agility

Why the AI era is putting a premium on fit and agility rather than power or moats

Hey! It’s Andreas from Monetisation Matters, the home of in-depth articles and actionable insights on Strategy and Monetisation. Built for Founders and leaders as they navigate their $1m to $100m growth journeys and beyond. This month explores why agility and fit is much more important than power or moats, and why sources of competitive advatnage can become burdens in a fast changing AI era.

---

---

Power in the new world, same as the old

Given the pace of progress, both in the foundational models, and the applications they are built on, investors are reflecting on which businesses will have durable moats in future vs. those that will be disrupted, commoditised or die. It’s tempting to think that because of AI the rules of the game have changed – that sources of market power and the nature of moats have changed. Sometimes I hear the moat in SaaS was building the software itself, and now with LLM driven code synthesis that is no longer true, but it was never true. It was never a source of market power as it was easily replicable with a bit of capital and some basic engineering talent.

Hamilton Helmer’s Seven Powers framework provides a rigorous and complete model of market power. It is just as true now as it was pre Open AI. I doubt LLMs, robotics, or any future innovations suddenly means there is an elusive eighth power that should be added, or that any of the seven are no longer relevant. If you are unfamiliar with the seven sources of market power they are:

• Network Economies

• Branding

• Process Power

• Counter-Positioning

• Switching Costs

• Cornered Resources

• Scale Economies

Time spent thinking about how market power is now different is time wasted. It’s looking for answers to the wrong question. If you abstract away from the specifics of LLMs, what we are experiencing is a combination of a productivity shock, creative destruction, and faster change rates. Markets, and the participants that constitute them are having to absorb a faster rate of change than they did a few years ago. But if you look back over the last 100 years, rates of change have been constantly increasing, particulalry in wartime. It just feels more disruptive for us because it’s what we are experiencing now and haven’t directly felt the rates of change over the last hundred years.

Hamilton’s framework provides an excellent explanation of why a particular business does or does not have market power, but it assumes static equilibrium positions: it doesn’t tell us anything about how a business acquired power, or whether it will still have it in future. The Seven Powers is a compass rather than a GPS, providing directional guidance when making strategic decisions, but we need additional perspectives to account for dynamics – how businesses gain, keep, and lose power.

---

Fit matters more than moats

Market power can’t exist without fit with the environment. A new invention that doesn’t solve a valuable problem for a meaningful set of customers will never generate power. Even disruptive startups that get the invention right face a harder task: AI has quickened the pace of change, and fast-moving threats are harder to respond to. Incumbents will see their power erode if they can’t absorb and respond to that change over time.

Yet our instinct is to defend rather than adapt. We use the word moat as a synonym for power. Build a moat so your business is hard to attack. The language assumes we are fighting a positional war, where the enemy can’t manoeuvre around and isolate you. When everything is changing, moats can become prisons. It is far more useful to think in terms of agility and fit than moats or power.

To maintain fit, both new entrants and incumbents need to keep up with the pace of change: that could mean actively abandoning the things that helped them gain market power in the first place. I like to think of fit as information. The product, the organisational structure, the processes and so on are all encodings of the outside world. Information is equivalent to fit, so when the environment changes, information is destroyed and fit declines. Some of those encodings are no longer useful: they describe a world that no longer exists.

If fit decays every time the world moves, what matters is how fast you can restore it. That is why we should care about speed. Two kinds matter. The first is operational: a product org with optimised cycle times ships faster than its competitors, a genuine edge for a product-led business. The second is structural: the ability to reconfigure those processes quickly when the market moves. The two pull against each other. A highly optimised process has fast cycle times but is slow to change, because the optimisation is precisely what makes it rigid.

Maintaining agility means we don’t want the business to be highly optimised overall. We want some slack and ‘inefficiency’, because it makes it easier to pivot and deploy spare capacity. So we are trading off speed at a micro-operational level for adaptability at the corporate level. Slack is one lever; information is the other. Decision-makers take the decisions they have the information to take. So we want our finger on the pulse of what is changing in the environment, and to incorporate that into our decision-making. With AI it’s much easier to do this cheaply, for example by creating agents to monitor the competitive landscape and report back.

Monitoring tells you what is changing now. Scenario planning prepares you for what might come. The larger the organisation, the more critical that second kind of foresight becomes for agility. The likes of Shell and Microsoft use it extensively to respond to threats and environmental change. I think startups get away with doing less of this, as they are far more agile. If a threat emerges that they didn’t anticipate, they can still redeploy resources and respond quickly. One of the most impressive recent responses to AI-driven change has come from Intercom. As soon as ChatGPT launched, the business began to pivot and redeploy resources to respond to what was both a major opportunity and a threat to its survival.

---

A masterclass by Intercom

Intercom faced an existential threat when ChatGPT launched in late 2022. The company was preparing for an IPO when the markets closed in November of that year, and the increasing usefulness of LLMs directly threatened a business model built on selling seats for human customer service representatives. Rather than treat the moment as incremental, Intercom built Fin, an AI product that now generates $100 million in revenue and is positioned as the future of the company. The pivot also flipped one of the company’s traditional assets into a liability: an established brand now carries “the burden of having an older brand”, explaining why the company has now been renamed to Fin. Intercom has been willing to destroy the old to make way for the new. It demonstrates how a source of power, in this case Brand, can become a liability, and must be actively abandoned to maintain fit with the changing environment.

The pivot has been accompanied by an explicit reversal of priorities. Where the SaaS era prized stability and predictability as preconditions for productivity, the AI era prizes agility and adaptability as preconditions for survival. AI now also sits at the core of how the business operates, driving rapid increases in productivity through smaller teams which require less coordination overhead. Fin leverages a dedicated R&D Services team as a sort of shock absorber to respond to commercially driven demands whilst protecting the core product roadmap. The shift in GTM has also become more future oriented, selling the vision rather than the current capabilities.

Operationally, the transition has required active demolition of accumulated processes rather than gentle evolution. Intercom recognised that the processes they have refined over the years are no longer relevant to the new world. It has had the courage to destroy what has come before to re-imagine what is needed to maintain fit. Intercom recognised that this wasn’t a choice, it could either lean in and be the agent of change or progressively lose fit over time.

The strategy has clearly worked given Fin’s reacceleration of ARR and its recent acquisition by Salesforce for $3.6B. Congratulations to the team.

---

I enjoy writing, but it is a tremendous investment. If you enjoyed this article, please consider sharing it so that it may grow.

Thank you.