tl;dr: I’m writing a seven-part series on common misunderstandings about strategy. A bit of background here, followed by links to Parts 1 and 2 of the series.
Some weeks ago, I was doing a whiteboard thought session with a pal when he asked me to explain how my view of strategy is distinctive and special. Fortunately, I’ve been thinking about, doing, and teaching strategy for so long — more than 15 years — that I could step up to the board and immediately dash off seven pairs of words (you can see the originals at the very end of this post).
Over the intervening weeks, I nudged these seven pairs of words into a bit more clarity. You can see them below, though they’re almost guaranteed to change as I unpack each one in turn.
Seven strategy tensions leading to misunderstandings about what strategy is and how to do it.
I soon realised that presenting these seven concept pairs without more context leaves the situation murky. What do these concept pairs even stand for? My answer for now is that each pair represents a tension in how we could think of strategy, and misunderstandings arise when we choose the wrong thing to emphasise in each pair.
The common misunderstandings are that strategy and planning are the same thing, that strategy is abstracted, predominantly cognitive, and goals-focused, that strategic actions should be clear, big, legible, singular, and monolithic. This view of strategy as a legible, clear thing also happens to be comforting and easily explainable. Unfortunately, it is also based in deep misunderstandings about strategy.
And these misunderstandings crop up frequently.
So I’m going to unpack each of these tensions to explain why strategy is not the same as planning, that good strategy needs to be embedded, affective, and tradeoffs-oriented, and that good strategic actions are probably profuse, distributed, amorphous, small, and illegible.
And if you’d like to have a chat about applying these insights on strategy in your organisation, just reply to this email (or get in touch with me some other way).
Over the new year, I was in upstate New York, the DC area, and Cambridge. There was a bit of work, but it was largely an excuse to escape Singapore’s heat and humidity and reflect on 2024. One thing I re-realised: Each year passes slowly in the moment but fast as a whole, and remembering even major events is nearly impossible without resorting to the calendar. What doesn’t get recorded somewhere and actively recalled usually gets forgotten. I wasn’t planning to be gone for three months, but here we are.
Sam’s Point, Ellenville, NY (31 December, 2024).
I went quiet because relocating back to Singapore (yes, I’m back; Singapore friends and neighbors, let’s catch up) has taken up more headspace than I wanted it to. Handling the numerous uncertainties that swirl around aging parents in declining health is hard to schedule around.
The uncertainty of early 2025 has a different flavour from the Covid-19 uncertainty in early 2020, or the first phases of the Global Financial Crisis in 2008. The world order seems to be on the cusp of a dramatic dislocation. (The dislocation has probably already happened, with what remains of the appearance of normalcy being preserved by the inertia and momentum of very large systems.)
In all this, I’m keeping my eye on how not-knowing can be generative and create space for opportunity, even as it disorients and creates panic and distress. All my work for the last 15 years has been about helping organizations and systems understand uncertainty and what to do about it.
I’ve been working with the UN Development Programme for over a year, building and deploying concretely applicable management tools to help teams that have to operate in an increasingly uncertain and rapidly changing development environment.
Most of January I spent preparing and delivering workshops for UNDP offices in Tunisia and Istanbul. These were workshops to help teams articulate tradeoffs more transparently and straightforwardly than they had before — and to teach them how to do this tradeoff articulation on their own. Seeing how quickly these workshops had concrete effect reinforced my belief that talking about tradeoffs is the only way to be pragmatic and creative about shared strategy and strategic goals, especially when resources and constraints are uncertain.
The Bosphorus (30 January, 2025).
Knowing there would be a language barrier, I refactored my usual workshop to make the process much more granular and even clearer than before. The feedback was better than I’d hoped—one participant said the process was “exceptionally straightforward, concrete, and immediately applicable for a tool that was newly introduced to a team.”
So now I’m building again—this time a tool to solve a problem I’ve been wrestling with for over a decade. I’ll share more when I have a junky prototype. (The last time I built a tool was in 2021, and the result was idk, which is the first training tool for productive discomfort. You should check it out.)
See you in a couple weeks,
VT
Somewhere in the west of Singapore (21 February, 2025).
tl;dr: Small-TAM tech companies get no love. But small-TAM tech companies that build software for mid-size clients in traditional industries fill a big and underserved niche, have a counterintuitive competitive moat, and have a more sustainable mechanism for growing big if they want to.
An unsolicited endorsement: For your less uncomfortable gift-giving needs, you should consider KIOSK. It is run by good friends with strange and impeccable taste, who spend a lot of their time doing the irreplaceable and fundamentally human work of meaningmaking that I keep on writing about.
Artificial intelligence (AI) has exploded in the public consciousness. It now attracts huge amounts of investment and attention. The focus has mostly been on building and improving underlying technologies and building out infrastructure. However, as we marvel at all these new AI-enabled capabilities, there remains a huge but often overlooked gap: Applications of AI to solving existing real-world business problems — especially in traditional industries like manufacturing, logistics, and construction.
While venture capital-backed startups chase massive but still not-yet-existent markets, there’s a missed opportunity for small, cashflow-focused tech companies to provide industry-specific AI solutions for medium-sized traditional businesses.
Since September, I’ve been nosing around this opportunity by talking to people in the AI industry, private equity investors, and operators in traditional industries. This issue unpacks some counterintuitive answers to their questions around the desirability and feasibility of a “technology Mittelstand” — a robust ecosystem of tech companies that build software for mid-size clients in traditional industries and which thus have small total addressable markets (TAMs).
Plage des Catalans (27 November, 2024).
Why focus on the medium-sized firm in a traditional industry?
Because the medium-sized traditional firm is the excluded middle.
Large firms often have both the resources and the business-case for a fully customized AI solution. A bank like JPMorgan Chase or a construction giant like Bechtel can afford to develop in-house customized AI tools (or engage a top-tier consulting firm to build these custom applications).
Small firms usually lack the resources to adopt sophisticated AI applications. They might benefit from off-the-shelf tools that improve operational efficiency, such as payroll automation or CRM systems. However, for these businesses, the investment required for retrofitting a customized AI application into their existing workflows is too high compared to the potential benefits.
In my conversations with both PE funds and mid-sized companies, the consensus appears to be that the perfect target profile for lightly customized AI tools is a traditional company that is big enough to have a Chief Procurement Officer (CPO) or Chief Operations Officer (COO) but without a dedicated technology executive (like a CIO or CTO) or an in-house technology practice beyond basic IT support. Businesses of this size operate at enough scale to benefit from tooling up but are simultaneously too small for a full custom solution to be economic, underserved by standard off-the-shelf solutions, and don’t have the capacity to build the tools internally.
Medium-sized firms are both the excluded middle and the sweet spot.
Why do these tools for mid-sized businesses need to be industry-specific?
Solutions best suited for these medium-sized clients are highly specialized, industry-specific applications. General AI tools with large, cross-industry total addressable markets (TAMs) might sound appealing, but to get to enormous TAMs you often need to build a product that spans multiple industries and so by definition must omit the understanding of existing industry-specific business process context that medium-sized firms always need.
For instance, a voice-to-text-to-prompt application that could serve many sectors would attract both VC funding and tremendous competition due to its potentially broad applicability and enormous TAM — but mid-size firms would have to adjust their business processes to make it work for them, and such firms just won’t have the internal capacity to do that kind of reorganisation just to use a new tool.
A tool designed specifically for one industry — such as an AI tool for construction companies to automate their safety compliance checks based on the regulations in one specific country — has an inherently smaller TAM and also requires more specialized, industry-specific knowledge to build. But once built, its utility to firms inside the industry becomes rapidly apparent.
Beams is an example of a small-TAM tech company. It provides the airline industry with AI tools for analyzing safety risk and now works with over 20 airlines including the Delta Airlines Group, Lufthansa Group, and Icelandair.
Valuable tools in traditional industries must be tailored to already-existing workflows, business processes, and regulations. All these are likely to be industry-specific rather than industry-spanning.
Aren’t small TAMs strategically bad?
Products with small TAMs are overlooked by VCs (and by product people looking for VC funding) because they aren’t scalable products that can achieve unicorn status. However, the overlooked advantage of the small TAM is that it discourages large tech companies from entering these niches. Small TAMs become a barrier to entry, defending tech SMEs from competition from companies seeking big TAMs.
Each of these industry-specific small-TAM tech companies would end up occupying a distinct tech ecological niche, much as a tropical rainforest has a huge number of species coexisting, each in its own niche that other species have no particular interest in trying to invade. (The need for industry-specific knowledge is another barrier to entry.)
And small TAMs aren’t even that small. SMEs in traditional industries such as construction, logistics, retail, manufacturing, finance, and real estate always make up a huge portion (usually well over 80%) of most countries’ GDPs.
Small TAMs act as an effective moat for attractive markets that aren’t as small as they superficially appear.
Will small-TAM tech companies never get big?
Designing products with (relatively) small TAMs doesn’t preclude a small-TAM tech company from becoming big — but if they get big, they do so starting from a different position.
VC-funded tech startups begin from envisioning a product for an enormous market. Neither the product nor the market exist yet, and often the product relies on technology that also doesn’t exist yet. (Some of these products also rely on the enormous market to exist before the products “work.” All the multisided platforms fall into this category.) So VC-funded startups have to build the technology, the product, and the market, then successfully sell the product into the market. All this is very hard, which is why most VC-funded tech startups fail along the way. (We aggressively celebrate and envy the ones which made it all the way to an IPO and forget that they are the exception, not the rule.)
The small-TAM tech company starts from a different place: a real need faced by actual companies working in an existing industry. The market already exists and the value of the product can be established. The need is frequently addressable without developing fundamentally new technology. The small-TAM tech company’s challenge faces is not like what VC-funded startups face in terms of developing new underlying technology or building not-yet-existent markets. Instead, the challenge is to understand what real clients actually need and are willing to buy. This takes yet more industry-specific knowledge about how real clients’ existing legacy business processes work and how to build a product that can integrate non-disruptively with those legacy processes with only light customization.
Acquiring deep industry-specific knowledge leads to identifying other needs in the same industry. For small-TAM tech companies that want to pursue it, this ends up being the path to bigness.
ServiceTitan shows how a small-TAM tech company can get big. It was founded in 2007 to build software for independent contractors to manage paperwork. Over time, it grew by adding more industry-specific tools, becoming the back-office software solution for small and mid-sized contracting businesses. Last year, ServiceTitan’s 8000+ contractor clients ran $6.2 billion in revenues through the company’s suite of services. The company has raised over $1.5 billion in funding so far, and filed for an IPO a few weeks ago.
By initially focusing on a clearly defined product that fills a need for an existing market in a specific industry, small-TAM tech companies create a strong foundation for sustainable growth if they want it. And they don’t need to pursue the high-failure-rate business models of VC-funded startups to do it.
Expanding how we think about tech entrepreneurship.
The intersection of technology deployment and traditional industries is an area overdue for innovation — especially in expanding what we valorise and support in technology entrepreneurship.
VC funding is crucial for driving highly uncertain new technology development. But we already valorise unicorns and have extensive infrastructures for funding and supporting VC-type tech entrepreneurship.
What’s underserved is the traditional industries that are the backbone of every economy, and especially the SMEs in those industries. Supporting these traditional businesses strengthens the whole economy. Apart from encouraging VC funding models and business models, we also need to fund and focus on the sweet spot in the missing middle: Developing a dense layer of small-TAM technology companies to service these traditional mid-size firms.
There’s literally no other way to equip these traditional industries to create jobs and foster long-term economic growth.
Miðbakki Port (17 October, 2024).
Next time here, I’ll probably write about the fundamentally different approach to product management needed to build a small-TAM tech company.
For Noë, the answer is that thinking is about “resistance,” the internally intentioned acts that prevent us from being completely dominated by external conditions — and intention is uniquely human. He concludes that “Our values are always problematic. We are not merely word-generators. We are makers of meaning. We can’t help doing this; no computer can do this.” (The emphasis is mine.)
I agree. For the last few years I’ve been puzzling out what it means to make meaning and why it separates humans from machines. For me, meaningmaking is what we do in the presence of uncertainty about what things are worth — more precisely, not-knowing about relative value.
There are in fact at least four conceptually distinct kinds of meaningmaking, each with different practical implications for how we act:
4 types of meaningmaking.
I grow ever more convinced that meaningmaking is a crucial but nearly entirely overlooked lens when thinking about AI policy, how AI research should be oriented, what kinds of products to build, how the present and the future of work should be designed, and what it means to be human. I’m glad that a philosopher with Noë’s reach has injected this idea into the noisy discourse on AI and I really hope it sticks.
So I’ve put together the 6 essays on meaningmaking I’ve written over the last 2 years — if you find them insightful or merely thought-provoking, please share them widely. There are more to come, but these are some of the working components of a broader exploration of not-knowings: understanding the different types of true uncertainty we face — none of which are what colloquially we call “risk” — and figuring out how we can respond to them generatively. I thought this would be a book, but maybe another format (or multiple formats) makes more sense.
So … what should I be writing about and in what forms should it find expression? I’d love your comments and suggestions.
tl;dr: AI investment is missing a huge opportunity in the form of industry-specific AI apps designed to integrate with legacy business processes in medium-sized firms in trad industries. This investment should go into industry-specific incubators that bridge the capability and capital gaps that prevent startups from building these real-world AI solutions that drive immediate impact.
So much time and money is flowing into AI today, but it’s nearly all focused on building foundational technology: massive models or compute infrastructure. The promise driving this investment is that AI will transform the real world: AI will make healthcare (or logistics, or law enforcement, or copywriting, or whatever) cheaper or better (or faster, or in more languages, or available to more people, or whatever).
This transformation will take much more than just good models and huge compute. These will have to be translated into actual applications that real-world people and organisations can use. Applications that fit into existing business processes in existing industries. Without an ecosystem of small tech companies doing this translation work, AI’s promise of transformation will go unfulfilled.
So this is the problem: Most startups are trying to become the next unicorn, chasing enormous markets that promise outsized returns. Too few startups are focused on building translational AI applications that make a tangible, immediate impact on the existing processes of medium-sized firms. This is a major market opportunity that almost no one is systematically exploiting.
I’ll unpack some of the reasoning below.
The shipyards along the Avenida da Beiramar, Vigo (2 August, 2018).
Why AI apps for traditional industries?
Because traditional industries are an enormous market in any country. Traditional industries — like construction, logistics, and insurance — form the backbone of every national economy. They don’t have the razzmatazz of digital and other new industries, but they nonetheless produce the majority (always more than 70%) of GDP. These industries predate widespread modern digitalisation, so they usually operate using legacy processes that could be transformed by AI applications.
The kinds of applications I have in mind are not general services that multiple industries can use, like project management tools (e.g., Asana or Slack). Instead, I’m thinking of applications that work only in one particular industry, that address the issues and processes specific to that industry.
An example from the construction industry: An application for construction site operations that processes drone, GPS, and mobile device data to automate building progress updates and material input/output management. Medium-sized construction firms using this application improve project management efficiency by up to 20%. These savings come from things like increasing the accuracy of just-in-time materials delivery to construction sites. Cumulatively, this reduces costs from project delays and increases overall project profitability. (BTW, this is an application built by a company called Unearth, which was acquired by Procore in 2023.)
Why medium-sized firms as clients?
Medium-sized firms in traditional industries are the sweet spot for startups building industry-specific AI applications. They’re large enough to see significant benefits from business process improvements using new technology and be able to pay for lightly customised products.
Why not target small or big firms? Big firms don’t need these applications; they’re large enough to staff in-house tools teams (or hire consultants) to build applications that are heavily customised to their own firm-specific processes. At the other end of the spectrum, small firms in traditional industries can’t afford lightly customised applications; they often lack the resources to buy them and the internal capacity to use them effectively.
The capability and capital gaps.
Medium-sized businesses in traditional industries are high-value targets for startups, but few startups are going after them. Why? Two reasons: a capability gap and a capital gap.
The capability gap.
It’s hard to build AI products that integrate into legacy business processes. This takes a different kind of product management approach, one that focuses on building products for small actually-existing markets, not large potential ones.
The problem is that most startups use a product management approach that is implicitly optimized for venture capital (VC) funding. The VC approach designs products to have enormous total addressable markets (TAMs). Enormous TAMs necessitate general utility with speculative business models and execution plans. Those lead to high failure rates in pursuit of unicorn status.
This VC approach to product management doesn’t work for startups trying to build AI applications for medium-sized firms in traditional industries.
The TAM for such an AI application is necessarily much smaller because it’s built to address the particular needs of a specific industry. This makes it valuable for firms in the industry, but less valuable and harder to sell to firms in other industries. (Though the TAM is small, finding product-market fit is easier, getting to positive cashflow is faster, and the small TAM naturally deters competition.)
Addressing industry-specific needs means deeply understanding and integrating with existing business processes. To do this requires a different approach to product management centred on an awareness of what I call meaningmaking.
Unfortunately, a meaningmaking-based approach to product management is not yet systematised or commonplace — this is the capability gap which needs to be filled.
The capital gap.
Beyond the capability gap, startups building AI applications for medium-sized firms often fall between the cracks when it comes to funding. They’re a poor fit for VC funding because their TAMs are too small (see above). They’re also a poor fit for private equity, because they’re not yet cashflow positive.
These startups would become cashflow positive more quickly if they had access not just to capital, but to industry expertise and connections. Specifically, industry insiders who know which business processes are most in need of AI support and the decision-makers who control procurement budgets.
The capital gap I see is for finance, industry-specific expertise, and access to procurement.
Filling both gaps with industry-specific incubators.
The answer to both the capability and capital gap is an incubator focused on building AI applications for medium-sized firms traditional industries. Such an incubator would do two things:
Develop product management capabilities: Train startups in a product management approach optimized for small but high-value markets. This means focusing on building products that integrate seamlessly into existing business processes, prioritizing cashflow-positive business models, and understanding the specific needs of traditional industries (like construction, logistics, insurance, brick and mortar retail).
Provide structured access to industry leadership: Give startups access to industry operations leaders who can help them identify high-potential processes for automation and efficiency improvements, as well as purchasing decision-makers who can shorten sales cycles.
The tech Mittelstand: A new vision for AI startups
The rewards are huge if we can successfully build this ecosystem. What will result is a a technology Mittelstand: The Mittelstand (“the middle estate”) is the large population of SMEs in Germany, known for their innovativeness, long-term stability, prosperity, and contribution to Germany’s industrial success. Instead of only trying to build tech unicorns, we should be trying to create conditions that enable a large number of startups to emerge, which mature quickly into a large number of small-headcount, stable-cashflow technology companies which provide a highly diverse range of industry-specific AI applications to serve medium-sized firms in traditional industries.
Medium-sized firms are under-optimized, not because the technology doesn’t exist, but because the right kinds of startups don’t exist in sufficient numbers. By addressing the capability and capital gaps, we can unlock a large, underserved market—one that could increase efficiency and profitability for thousands of businesses, ultimately strengthening the broader economy.
Why now?
AI has the potential to transform how businesses operate, but the current focus is skewed toward the development of underlying technology. AI’s missing middle is the application of that technology in ways that drive immediate value for real businesses. Medium-sized firms — too large for off-the-shelf solutions, too small for custom-built AI from large tech providers — are the ideal target. And the tech startups that will serve these mid-size firms will necessarily be small and scrappy, and there will have to be many of them.
So we need to rethink both how AI products are developed and how they are funded. By focusing on specificity, cashflow, and real-world application, we can create a profusion of AI companies that are individually small but collectively crucial to the future of work and the economy.
