Hard things are now easy

I have sat in a lot of strategy sessions over the past year, in government departments, in banks, in ASX 200 boardrooms. The decks differ. The shape rarely does. Horizon charts, a competitor grid, an operating model, and then, somewhere near the end, one slide about AI: an efficiency programme, a risk register entry, a savings number. The room nods at that slide and goes back to the real strategy.

I have nodded along with them, more than once, because it is comfortable and it lets everyone keep the plan they walked in with. And yet that slide is a question about whether the rest of the deck still stands.

Every strategy is a bet about difficulty. Underneath the frameworks and the horizon charts, a strategy is a set of assumptions about what is hard: hard to build, hard to copy, hard to learn, hard to do at quality, hard to do fast. You place your organisation on the right side of those difficulties and you win.

A large class of those hard things is now easy. Repriced, all at once. Producing a competent first draft of almost any knowledge artefact, an analysis, some code, the copy, a contract, a curriculum, a campaign, has gone from days of skilled labour to seconds of machine time. Difficulty is the ground a strategy stands on. When the ground moves, everything standing on it moves too, whether the document admits it or not.

An audit of quiet invalidations

Take them one at a time, and ask each one the same question: was this priced when things were hard?

Moats made of accumulated expertise. A firm that spent twenty years developing its own methodology, a way of doing due diligence, of structuring deals, of running assessments, has been holding an asset whose value depended on that methodology being hard to copy. When a capable model and a good practitioner can approximate much of it in weeks, the moat has not vanished, but it has narrowed from a river to a creek. What stays defensible is the part machines genuinely cannot do: the relationships, the judgement under ambiguity, the accountability someone will pay for. A strategy that does not separate the automatable core from that residue is defending the wrong asset.

The talent pyramid. Professional services, and most corporate functions, are shaped like pyramids because the economics of hard work demanded it. Many juniors do the laborious groundwork, fewer seniors apply judgement, and the margin lives in the gap. When the groundwork gets easy, the base of the pyramid stops being a profit centre and starts being a cost question. And yet the pyramid was never only an org chart. It was the training system. Juniors became seniors by doing the hard, boring work that is now automated. Any strategy that assumes a continuing supply of seasoned judgement, while removing the work that seasons it, has a contradiction in it with a ten-year fuse.

Validate before you build. A generation of product strategy rests on the idea that building is expensive, so you test the idea first: minimum viable products, staged funding, fail fast. When a working version of most software ideas takes days, the scarce resource inverts. Experiments become cheap. What is expensive is attention, distribution, and the judgement to know what is worth trying at all. Fail fast was a rationing rule for costly construction, and rationing something that is now abundant is not discipline, it is a habit. The organisations that update will run dozens of cheap, real experiments while their competitors are still writing the business case for one.

Agile ceremony. Agile was also an adaptation to difficulty, specifically the difficulty of specifying software up front when building was slow and change was costly. Sprints, ceremonies and story points are machinery for rationing scarce developer hours against uncertain requirements. When implementation is fast and cheap, the bottleneck moves upstream, to deciding what to build and to recognising whether it is any good. Much of the ceremony carries on anyway, because process outlives the constraint that justified it. Which of our processes exist to manage a difficulty that no longer exists? That is one of the most valuable questions a leadership team can spend an afternoon on.

Headcount as the measure of capacity and status. Budgets, spans of control, band levels, the whole grammar of organisational size assumes that output scales with people. When a team of five with agents outperforms a team of fifty without, that grammar breaks, and the incentive structures break quietly along with it. If your leaders' status still scales with the size of their empires, you are paying people to resist the exact restructuring your strategy needs. Most large organisations I sit with are doing precisely that, and none of them set out to.

The grace period before competitors respond. This is the least examined assumption of the lot. Strategies assume that when you move, competitors need time to see it, decide, and copy it. That interval was set by how hard copying was, and it is collapsing. Anything visible and automatable will be copied faster than your planning cycle assumed. Which forces a sharper question than most strategy sessions ask: of everything we do, what specifically stays hard?

What stays hard

Some things are, because the repricing is uneven. Knowing which of your assumptions still hold, and which have quietly been repriced, is the skill that matters now.

Trust stays hard. Institutional permission, being the organisation a government, a bank or a regulator will let do the thing, is accumulated slowly and no machine shortens it. Distribution stays hard: attention, relationships, and the people who already buy from you. Accountability stays hard, and it may be appreciating. When anyone can produce the analysis, the question becomes whose name stands behind it, and a name is not something a machine can offer. Taste and problem selection stay hard, because when production is abundant the chooser becomes the scarce role, not the producer. And getting hundreds of humans, and now thousands of agents, pointed at the same intent may be the hardest thing left on the list.

The deliverables got easy. What stays hard is everything around the deliverable: the judgement before it, the trust beneath it, the accountability after it. That residue is what a strategy should now be built on.

The meeting to have

Take your current strategy, the real one that resourcing decisions follow, and run a difficulty audit on it. List the assumptions holding it up: what we believe is hard for us, hard for competitors, hard for anyone new arriving. Then ask two questions of each. Was this priced before 2023? And if this got ten times easier, does the strategy still stand?

In my experience the exercise produces three lists. The assumptions that survive, which are genuinely still hard and should now be defended on purpose. The assumptions that fail, where the strategy needs redesigning rather than adjusting. And the third list, which is the one people go quiet over: the things your organisation always wanted to do and never could, because they were too hard, sitting there now, easy and unclaimed.

Every organisation I work with has that third list. Almost none of them has looked at it, because the strategy process is still staffed, scheduled and framed as though difficulty sat where it did five years ago. The teams that redraw the map first, and are honest enough to act on what it shows, including the parts that gore a sacred cow or two, will win the next cycle. The rest will execute a good plan built on ground that has already moved.

Our strategic foresight roundtables exist to run exactly this examination with a leadership team. If your strategy has not been audited against the new landscape, start the conversation.

Attribution

Written by Andrew Ramsden. AI tools assisted research and drafting; all outputs verified.

Accountable

Andrew Ramsden.

Limitations

No external sources.

References

No external sources. Claim labels are in the SOURCED sidecar.