All insights
operating-modelprocess-designoperational-excellenceautomation

The seam tax: why faster steps don't add up to a faster business

Most organisations buy tools to speed up steps, then wonder why the end-to-end result barely moves. The value doesn't leak inside the steps — it leaks in the seams between them. And because that loss compounds, the tax is far larger than any single step is slow.

Yasir Aheer20 July 20268 min read

You bought the software. You automated the invoice capture, the data entry, the report generation. Every step on the process map is now measurably faster. And yet the thing your customers and your board actually feel — the time from start to done — has barely moved.

This is the most expensive illusion in operations, and almost every organisation is paying for it. It shows up at scale: some 88% of organisations now use AI in at least one function, yet only 39% report any enterprise-level EBIT impact from it [1]. Near-universal adoption; rare end-to-end result.

The instinct is to blame the tools, or the change management, or the data. But the gap between near-universal adoption and rare end-to-end impact points somewhere more structural — to the parts of the process nobody owns.

Loss compounds; it doesn't add up

Here's why the seam tax is so easy to underestimate: we think about process loss additively — "we lose a little here, a little there" — when it actually behaves multiplicatively. A small amount of value lost at each seam compounds into a large loss end-to-end.

Set the model below to your own reality: how many steps does a typical piece of work pass through, and how much of its value — momentum, accuracy, context — survives each handoff?

The seam tax, compounded

With 6 steps and 92% of value surviving each of the 5 handoffs, end-to-end yield is 66 percent — a seam tax of 34 percent of the original value.

End-to-end value that survives

66%

The seam tax

34%

Removing a single handoff recovers about 6% of end-to-end value — because loss compounds across seams, that beats making any one step faster.

A model of compounding loss. Even 'good' seams — 92% value retained — become a heavy tax once work crosses enough of them.

The number that surprises people is not the per-seam loss; it's the total. At a comfortable-sounding 92% retention, a six-step process quietly sheds a third of its value before it finishes — and that loss is invisible on any single team's dashboard.

Why the tax is invisible

Every step is measured; the seams are not. Each team optimises its own box, hits its own SLA, and reports green. Meanwhile the work waits in a queue between two green boxes — and no one's dashboard shows the wait.

This is exactly why bolting tools onto existing steps disappoints. In finance, for instance, rule-based automation now handles invoice capture and reconciliations comfortably — but the close still stalls, because value leaks where terms, exceptions, and approvals are handed between systems and people [4]. The automation made the boxes faster. It did nothing about the seams.

The anatomy of a seam

"Seam" sounds abstract until you learn to see the four kinds. Each hides its loss somewhere different — and each has a tell you can look for tomorrow morning.

The Four Seams

Where the Tax Actually Hides

Tap a seam to see how to spot it

Select a seam above to explore it

Once you can name the seams, you start seeing them everywhere — and you notice that speeding up the boxes on either side does nothing to the wait between them.

The tell: tool-first vs seam-first operations

There are two ways to run an operation, and they produce very different economics. The difference isn't effort or talent — it's where you point them. Switch between the two and watch the same dimensions change.

Two ways to run an operation

Tool-first

Optimise the steps

Point effort and budget at making each step faster. Intuitive, easy to buy — and it plateaus, because the loss lives between the steps.

Unit of improvement
The task / step
What gets measured
Step speed, team SLAs
Where automation goes
Onto existing steps
How cost behaves
Adds up linearly, then plateaus
Typical result
Faster steps, same end-to-end time

Every box turns green while the end-to-end result barely moves.

Seam-first

The higher-order move

Optimise the flow

Point effort at the gaps between steps — removing handoffs, waits, and re-keying. Harder to see, far more valuable, because it stops the compounding loss.

Unit of improvement
The end-to-end flow
What gets measured
Total lead time, handoffs removed
Where automation goes
Across / into the seam
How cost behaves
Falls as seams are removed
Typical result
Fewer handoffs, compounding gains

Lead time drops because the tax between the boxes is collected back.

The evidence favours the seam-first posture unusually clearly. In McKinsey's data, the organisations actually capturing enterprise value — about 6% of the total — are nearly three times as likely to have fundamentally redesigned their workflows rather than layering technology onto the old ones; that redesign is one of the strongest predictors of real business impact of any factor tested [1]. Tools sit inside the steps. Redesign is what happens to the seams.

How to collect the tax back

Collecting the seam tax back is not a technology project. It's an operating-model move, and it runs in a deliberate order — because automating a broken seam only makes you lose value faster.

The Collection Sequence

How to Collect the Tax Back — In Order

Order matters: automating a broken seam only loses value faster

  1. Follow one real unit of work from trigger to done. Measure wait time separately from touch time.

    Why it matters

    In most processes the work spends far longer waiting than being worked on — and the wait is exactly what no dashboard shows.

    In practice

    Draw the flow as a timeline, not an org chart. Mark every point where the work sits idle. Those idle bars are the seam tax, quantified.

This is what "reinventing the operating model" actually cashes out to. When organisations redesign around end-to-end outcomes rather than functional boxes, the reported results aren't incremental: costs down by up to half, execution speed doubling or tripling, and decision cycles cut by as much as 70% [3]. Those gains don't come from faster steps. They come from removing the seams — and the layers and handoffs that created them [2].

The question that matters

The tool-first question is: Which step can we make faster? It feels productive, it's easy to buy, and it will keep disappointing you — because it optimises the boxes while the tax runs in the gaps.

The seam-first question is harder and far more valuable: Which seam is taxing the entire flow? Answer that, and you stop paying for speed you never feel.

Faster steps don't add up to a faster business. Fewer seams do.

Sources

  1. McKinsey & Company (QuantumBlack). The State of AI: Global Survey 2025. November 2025.88% of organisations report regular AI use in at least one function, yet only 39% report any enterprise-level EBIT impact; the ~6% of 'AI high performers' are nearly three times as likely to have fundamentally redesigned workflows.View source
  2. Boston Consulting Group. Scaling AI Requires New Processes, Not Just New Tools. 2026.Capturing value depends on reworking processes, spans of control, and organisational layers — not on adding tools to existing workflows.View source
  3. Boston Consulting Group. To Thrive in the AI Era, Tech Leaders Must Reinvent Organization and Operating Models. 2026.Companies that redesign operating models around end-to-end outcomes report cost reductions of up to 50%, execution speed doubling or tripling, and decision cycles cut by up to 70%.View source
  4. McKinsey & Company. AI in finance: How finance teams are putting AI to work today. 2025.Distinguishes rule-based automation (invoices, reconciliations) from agentic systems that orchestrate whole workflows such as the close — and describes value leaking where terms and steps are handed between systems.View source
Y
Yasir Aheer· Founder, OpsTeam

Yasir Aheer is the founder of OpsTeam. He writes about engineered operations, operating-model design, and the business and organisational implications of running People + Engineered Platforms + Production AI as one integrated system.

Ready to put this thinking into practice?

We design and operate integrated operating models for organisations ready to compound efficiency. Let's discuss yours.

Schedule Discovery Call