Skip to main content

Delivery vs Capability: Designing Engineering Organisations Across Time Horizons

· 5 min read

In my most recent posts, I’ve explored how AI is starting to reshape both the structure of engineering teams and the shape of individual capability.

  • Teams shifting from pyramids to more diamond-like structures
  • Individuals moving from T-shaped profiles towards multi-pillar depth

There’s a third dimension that sits across both: Time.

Many of the tensions we’re seeing are not just structural. They are temporal.

The hidden axis: time horizon​

Most organisations operate across at least two overlapping horizons:

  • Short-term (tactical) — delivery, execution, output
  • Long-term (strategic) — capability, resilience, sustainability

In theory, both matter.

In practice, one tends to dominate.

Why the short term wins​

The short term is measurable.

  • Features shipped
  • Deadlines met
  • Revenue delivered
  • Incidents resolved

It is visible, immediate, and tied directly to performance.

The long term is not.

  • Capability development
  • Team resilience
  • Knowledge depth
  • Future leadership

These are slower, harder to quantify, and often only visible in hindsight.

So when organisations come under pressure, they default to optimising for delivery.

That's rational, but it's also where the problem begins.

A second tension: speed of change vs speed of planning​

There is another layer of friction that is becoming more pronounced.

AI and tooling are evolving on a cadence measured in weeks and months.

Most organisations plan on a cadence measured in quarters and years.

That mismatch matters.

  • Roadmaps are set based on assumptions that may already be outdated
  • Capability decisions are made for roles that are actively changing
  • Investments are locked in before the landscape stabilises

Which creates a structural lag:

By the time many organisations adapt, the environment has already moved on.

This is not just about being slow.

It is about operating with planning cycles that are no longer aligned to the rate of change.

AI amplifies both tensions​

AI sits at the intersection of these two dynamics.

It increases short-term delivery capability:

  • Fewer people can do more work
  • Senior engineers gain significant leverage
  • Output becomes easier to scale

At the same time, it accelerates environmental change:

  • New tools emerge rapidly
  • Best practices shift continuously
  • The definition of roles evolves in real time

This creates a compounding effect:

  • Organisations optimise for short-term delivery
  • While the long-term target is moving faster than expected

The capability deficit​

If you consistently prioritise delivery over development in a fast-moving environment, the effects compound more quickly:

  • Fewer entry points into the system
  • Less experiential learning
  • Narrower progression pathways
  • Greater reliance on a small number of high performers

But now there is an additional risk:

  • The skills you are building may not match the environment you will face

You are not just under-investing in capability, you may be investing in the wrong shape of capability.

Three interacting systems​

It’s useful to think about this as three interacting systems:

1. Delivery system (short-term)​

  • Focus: output, efficiency, predictability
  • Structure: tends towards lean, senior-heavy teams
  • Optimisation: speed and throughput

2. Capability system (long-term)​

  • Focus: skill development, knowledge transfer, progression
  • Structure: requires space for learning and growth
  • Optimisation: future effectiveness

3. Adaptation system (rate of change)​

  • Focus: sensing and responding to external change
  • Structure: requires feedback loops and flexibility
  • Optimisation: alignment with a moving environment

Most organisations are relatively mature in the first.

Some invest in the second.

Very few explicitly design for the third.

The adaptation gap​

This is where many of the current challenges sit.

Organisations are trying to:

  • Deliver at speed
  • Build long-term capability

But without a system that allows them to adapt at a comparable rate to the tools they are adopting.

The result is a gap:

  • Between what teams are optimised for
  • And what the environment actually demands

This shows up in subtle ways:

  • Tooling is adopted, but not fully integrated
  • Processes lag behind capability
  • Individuals are expected to adapt faster than the organisation around them

Designing across three horizons​

The challenge is no longer just balancing short-term and long-term.

It is operating across three interacting horizons:

  • Now — what do we need to deliver?
  • Next — what capability do we need to build?
  • Changing — how is the environment shifting under us?

Each requires a different kind of thinking, and each pulls in a different direction.

What this means in practice​

A few implications follow from this:

  • Planning needs to become more fluid
    Fixed roadmaps struggle in rapidly evolving environments

  • Capability development needs to be more targeted
    Not just more training, but training aligned to where the environment is heading

  • Teams need slack for adaptation
    Without space to experiment and learn, adaptation does not happen

  • Senior judgement becomes more critical
    Deciding what to adopt, what to ignore, and when to change direction

This is less about adopting AI tools.

It is more about building organisations that can evolve alongside them.

Where individuals fit​

This connects back to the changing shape of the individual.

A multi-pillar capability model only works if individuals have:

  • Exposure to different domains
  • Time to integrate learning
  • The ability to operate in ambiguous, shifting contexts

These are precisely the conditions that are hardest to maintain in delivery-optimised environments.

A transition, not a choice​

This is not a binary decision.

You don’t choose between delivery, capability, and adaptation.

You are operating across all three, whether you design for it or not.

The risk is not that organisations focus on delivery.

The risk is that they do so using planning and capability models that assume a slower-moving world, because the environment is no longer waiting for those models to catch up.

The organisations that recognise that earliest will not just deliver more, they will adapt faster.