Почему передачи обходятся дороже, чем любая отдельная слабость в цепочке

The short version: most digital work does not fail because one team is weak. It fails at the joints, in the silent gap between the strategist's insight, the designer's file, the developer's build, and the marketer's live page. Each handoff loses a little intent. String four together and you have quietly given away half the plan before a single visitor arrives.

Let me show you where that loss actually happens, because once you can see it, you can start closing it, whether or not you can put one mind across the whole thing.

Where does the value actually go missing?

Not in the work itself. Almost every team I have watched fail at this was individually good at their job. The strategist's positioning was sharp, the designer's file was clean, the developer shipped what the ticket asked for, and the growth team ran a legitimate test. Everyone did their part correctly, and the result still missed.

The loss happens in the gaps between those parts, at the moment one person's understanding has to become another person's instructions. That translation is never perfect. A brief can only carry so much context, a handoff meeting only runs so long, and the person receiving the work was not in the room when the reasoning took shape. So they fill the gap with a reasonable guess. Most of the time the guess is close enough. Sometimes it is not, and nobody finds out until the thing is live.

This is the old telephone game, played with a six-figure budget. Each person repeats what they heard as faithfully as they can. The distortion is never any one person's fault. It is a property of the chain itself.

What is intent decay, and why does it compound instead of just adding up?

I call the specific failure "intent decay": the gradual loss of the why behind a decision as that decision passes from the person who made it to the people who have to execute it. It rarely shows up as an error. It shows up as a thousand small, defensible choices that each drift slightly from what the original thinking actually needed.

Three realistic, hypothetical examples, the kind I have seen some version of at almost every studio and in-house team I have worked around:

• The strategist's insight that never reaches the developer. Say the positioning work concludes that speed is the one thing this brand can own, faster checkout, faster support, faster everything, because the category is full of bloated competitors. That finding lives in a strategy deck. Nobody tells the development team it is the differentiator, so they build a checkout with a polished five-step wizard, on schedule and on brief, and quietly bury the one thing the strategy was built around.
• The SEO plan the designer never saw. The search plan calls for a real, indexable heading and body copy answering a specific question, because that is the query the whole page exists to capture. The brief that reaches the designer talks about "hero impact" instead. The designer builds the answer into a full-bleed graphic with the text baked into the image: stronger than any competitor's page, and functionally invisible to search.
• The copy the brand team wrote that the growth team quietly overrides. Brand writes a restrained call to action, "Explore the collection," to protect a premium tone. Growth quietly tests "Only 3 left, order now" against it, because urgency reliably lifts clicks, and never loops back to ask what it costs the positioning. Repeat that across a year of tests, and the brand gets rewritten by whoever happened to win one, not by anyone deciding on purpose.

None of these people did anything wrong given what reached them. That is the uncomfortable part. Intent decay is not a competence problem. It is a transmission problem, and transmission problems compound.

Here is the compounding, reasoned from a simple model rather than any single study: imagine each handoff preserves ninety percent of the original intent, which is generous for most real handoffs I have seen. After one handoff you have ninety percent of the plan. After two, eighty-one percent. After three strategy to design to build, seventy-three percent. After a fourth, into growth and live optimization, about sixty-six percent. You have not lost everything at any single step. You have lost a third of the plan through ordinary, well-intentioned friction, and nobody along the way could point to the moment it happened.

Where exactly does the value leak? A handoff to leak map

It helps to name the leak at each boundary specifically, because "communication problems" is too vague to fix. Here is the pattern across a typical strategy, design, build, and growth chain.

Read down that middle column and a pattern jumps out: what survives a handoff is almost always the artifact (the deck, the file, the ticket, the number). What leaks is almost always the reasoning behind it. Documents travel. Judgment does not, unless a person carries it across the boundary themselves.

Why does a coordination tax hit every project, even the well-run ones?

This is not just a story about creative work. It shows up in the data on how knowledge work actually gets done. Research led by Rob Cross and collaborators, tracked across two decades and reported in Harvard Business Review, found that time spent by managers and employees in collaborative activities, meetings, messages, requests for input, has grown by fifty percent or more over that period, to the point that people at some companies spend as much as eighty percent of their time responding to others rather than doing their own work (Harvard Business Review, to re-verify).

PMI's own research puts a number on what that friction costs: of every one billion dollars spent on projects, roughly seventy-five million is at risk specifically from poor communication (PMI, to re-verify). A 2026 Forbes analysis of engineering organizations estimated that a thirty-person team can absorb fifteen to twenty percent of its capacity, nine hundred thousand to one and a half million dollars a year, in pure coordination waste, and that this overhead scales quadratically as team size grows, not in a straight line, because every new person has to synchronize with everyone already there (Forbes, to re-verify).

There is a cognitive cost underneath that too. Gloria Mark's research at UC Irvine found it takes about twenty three minutes to fully return to a task after an interruption (UC Irvine, to re-verify), and every handoff is an interruption for someone. Google's DORA research adds the flip side: teams with loosely coupled architecture and clear, autonomous ownership of their piece of the system are roughly three times more likely to be elite performers than teams without it (DORA, to re-verify). Fewer live dependencies between people is one of the strongest predictors of shipping well this research has found.

Why does a single weak link, owned by one mind, still beat a strong link that nobody connects?

Here is the part that is easy to miss if you are only grading each stage on its own merits. A chain is not judged by its strongest link. It is judged by whether the links actually hold together, and a specialist's excellence at their one stage says nothing about whether their output survives the trip to the next stage intact.

Picture two versions of the same project. In the first, four specialists, each meaningfully better at their craft than any generalist could be, hand work down the line with no one accountable for what survives the handoff. In the second, one competent person, not the best at any single stage but good enough at all four, carries the intent from the first decision to the live page. The individual stages in the first version are stronger. The finished result in the second is usually more coherent, because nothing had to survive a translation it might not survive.

This is not an argument against specialists. It is an argument for something connecting them. A weak link that is directly wired to everything around it will transmit almost all of what it has. A strong link sitting in isolation, brilliant on its own terms, transmits only what happens to survive the gap on either side of it, and as the model above shows, that gap eats real value even when everyone involved is good at their job.

What do you actually do about it, whether or not you can hire one owner?

You cannot always put one mind across strategy, design, build, and growth. Most teams cannot, and most projects are built by more than one person for good reason: real depth in four disciplines rarely lives in one head, and it should not have to. What you can do, in either case, is shrink the leak at every boundary you have got.

1. Reduce the number of handoffs before optimizing any single one. Every boundary is a chance to lose intent, so the fastest fix is often structural: fewer vendors, fewer approval layers, fewer people between the decision and the shipped page.
2. Give one person end-to-end accountability, even if they do not do all the work. Amazon's internal "single-threaded owner" practice is built on this: one person tracks the thread from decision to delivery across a team of specialists, so intent has somewhere to live besides a document.
3. Make the reasoning travel with the artifact, not just the artifact. A brief that states the decision without the "why" gets reinterpreted at every stage. State the non-negotiable in writing next to the deliverable: not "premium and fast," but "speed is the differentiator, protect it before polish."
4. Close the loop backward, not just forward. Most chains only pass work downstream. Build a real path for growth's findings and the developer's constraints to reach the strategist, so the original decision maker learns when reality disagreed with it.
5. Measure the leak directly. Ask the same question of both sides of a handoff: "what is the one thing this must protect." Mismatched answers show you exactly where the intent decayed.
6. Where you cannot remove a handoff, put a person on both sides of it. A designer in the strategy conversation, or a developer in design review, carries context no brief fully captures: not one person doing everything, but one person's judgment present at every seam.

I make the fuller case for why a single owner is worth building toward elsewhere on this site, in the journal. Here, the point is narrower and more useful regardless of your structure: name your handoffs, and treat each one as a place where value is actively at risk, not a formality on the way to the real work.

Your turn

Take the last project that underperformed and draw its chain: who decided what, and who received it next, at every step from idea to live page. Circle each handoff. For each one, ask what the receiving person actually knew about why the decision was made, not just what the decision was. You will usually find the leak in under ten minutes, and once you can see it, closing it is a much smaller job than the one you thought you had.