Creating Value

The best-run team I have worked with built something nobody wanted.

They were genuinely good. Ownership was unmistakable, the boundary was drawn around a real capability, and they had the business judgement embedded rather than represented — the arrangement the whole of Part II argues for. They had cut their own lead time by more than half in a year, without being asked to. Nothing about them was broken.

Eighteen months later most of what they had built was switched off. Not because it failed. Because the thing it did turned out not to matter much to anyone, and the organization had been so pleased with how fast it was arriving that nobody had asked.

That is the failure this chapter exists for, and it is the one the previous chapter cannot prevent. Reducing friction makes an organization produce more. It has nothing to say about what.

Both halves are cultivated

There is a tempting and wrong way to hold this: friction is the thing you work on, and value is the thing that arrives afterwards if you did the first part properly. Remove the obstacles, and the good outcomes take care of themselves.

They don’t. An organization can be cultivated toward greatness on both halves, and the two require different attention.

Friction is cultivated by subtraction — a boundary settled, a decision moved closer to the information, an approval removed because ownership made it redundant. You are taking things out of the way of something that already wants to happen.

Value is cultivated by contact — with the person on the other end, with the outcome rather than the output, with evidence rather than the assumption written down at funding. You are not removing an obstacle. You are shortening a distance.

Neither substitutes for the other, and an organization that only ever does the first will get very fast at something, eventually at almost anything, including the wrong thing. The compliment I have heard most often about high-performing teams — they just ship — is a description of one half of a system, delivered as though it were the whole.

Why the second half gets postponed

Value is only legible from outside the organization. That single fact explains almost everything about why it goes unattended.

Output can be counted on a Thursday by someone in the building. Value requires asking someone who does not work for you, waiting for an answer that arrives late, and accepting a number that refuses to attach itself cleanly to any one team’s effort. Given a choice between a measurement you can produce this week and one you can’t, the week wins, in every organization I have worked in.

So the question of whether any of it produced anything gets answered once, at funding, in the one moment when no evidence exists yet — and then never again, because by the time evidence exists the decision is a year old and nobody wants to reopen it.

Four things that actually work

None of these are programmes. Each one is small enough to do without permission, which is deliberate.

Put someone in the room who experiences the outcome. Worth being precise about who that is, because it is easy to believe you have already done it.

Internal customers are real customers. When one capability hands to the next — underwriting to policy administration, recruitment to onboarding, claims to payments — the receiving capability genuinely consumes what the first one produced, and it deserves the same interface, the same documentation and the same attention you would give somebody outside who was paying. Nothing in this paper argues otherwise.

What it does argue against is a different arrangement that dresses itself in the same language: “the business” requests and IT delivers. That is not a customer relationship, because those two are not two capabilities. They are two halves of one, split into a requester who owns the outcome without the means and a supplier who owns the means without the outcome — which is why the vocabulary of service never quite fixes it. The Team That Has to Ask sets out what that split costs.

So there are two questions here rather than one, and an organization can be excellent at the first while completely blind on the second. Who consumes what we produce, and are we treating them as well as we would treat a stranger. And then: how many hops sit between us and somebody outside the organization altogether. In a platform or an ecosystem the people you talk to are usually the next team along, and they will ask you about fields, rate limits and edge cases. That conversation is valuable and it is not this one, because nobody in that room has met the person at the far end either.

With that said, this is the single highest-return move in the paper, and it appears twice for a reason — once as a friction fix in The Cost of Value Never Created, and again here. When the finance controller sat next to the team, the printouts went away; that was the friction half. The part I did not expect was that they started building things the controller asked for that nobody had specified, because the conversation was cheap enough to have. Proximity removes handovers and it generates value, and it is the same intervention both times.

The same thing worked running the other way, under a name I have kept: guest of reality. Developers spent time out where the work actually happens — beside somebody using what they had built, in the operation rather than in a workshop about the operation.

What came back was not the strategic insight anybody had hoped for. It was a pile of small things. A field re-entered three times. A report exported and reformatted by hand every Monday. A screen that made sense only if you already knew the answer. A good number were fixed within the week, because the person who could fix them had now seen them — and none of them would ever have been written down and submitted. They were each too small to be worth anybody’s request form, and together they were most of somebody’s week.

It also settled the estimating, in both directions, which I had not expected. Things the business had assumed were enormous took an afternoon. Things assumed trivial turned out to touch four systems and a contract nobody had read recently. Neither side could have known that on their own, and what resolved it was almost always somebody saying wait, say that again — why does it have to work that way.

Which is the part worth keeping. The fixes were good, and the conversation was the point. Cross-competence is not produced by putting the competences on the same org chart. It comes from them having enough of those conversations to stop guessing at each other’s half.

None of that happens on schedule, though, and it is worth being honest about what it actually depends on. You can put the finance controller next to the team and get nothing from it, if nobody involved particularly wants to know what the other person’s week is like. Wait, say that again — why does it have to work that way is an empathetic question before it is a useful one — it only gets asked by someone who has already decided the other person’s experience is worth understanding. And it only gets followed up by someone curious enough to want the actual answer rather than the convenient one.

Growing What Cannot Be Installed makes the case that empathy and curiosity are conditions to be cultivated rather than traits to hope for. This is where the absence shows up on the value side rather than the friction side: proximity without curiosity produces people sitting in the same room, still building the wrong thing, now with better attendance.

Fund outcomes for a period, not outputs on a date. Most organizations fund a scope and then measure whether the scope was delivered, which guarantees the answer is about delivery. Fund a team against an outcome for two quarters, and the conversation at the end is about whether anything changed for anyone. It is a harder conversation. It is also the only one that ever produced a decision to stop.

Make someone own the outcome, not the delivery. Ownership of a delivery ends when the thing ships. Ownership of an outcome does not end at all, which is precisely why organizations avoid assigning it. But an owner who is still holding the question six months later is the only mechanism I know of that reliably kills work that stopped being worth doing.

Close the loop, and let the closing be uncomfortable. Learning is the return leg. If the outcome never travels back to the people who chose the work, they will choose the same way again, confidently, because nothing contradicted them. The test is the one from the Learning chapter: after the last significant thing that did not land, what is now decided differently?

The trap, which is easy to fall into

Value work fails in a characteristic way, and it is worth naming because it looks like success.

An organization decides to focus on outcomes, discovers that outcomes are hard to measure, and adopts a proxy — engagement, adoption, usage, satisfaction, a score out of ten. The proxy is countable, so it gets counted, so it gets reported, so it gets targeted. Within a year the organization is optimizing the proxy with the same efficiency it previously applied to output, and the distance to the actual person has not shortened by a metre.

The defence is not a better proxy. It is keeping at least one channel open that cannot be optimized: someone in the organization who talks to real users often enough that the numbers can be contradicted by a specific human being. That channel is cheap and it is the first thing cut when the quarter gets tight.

What it looks like when both halves are cultivated

Work arrives faster and less of it is wrong. Fewer things get built, and more of what does get built is still running two years later. The organization’s arguments move upstream — from why is this taking so long to is this the right thing — which is a considerably more useful argument to be having, and one that only becomes available once the first question stops dominating the room.

And the two halves compound. Removing friction shortens the distance between deciding something and finding out whether it was right, which makes the learning loop tighter, which improves the next decision about what is worth doing at all. Cultivating value gives the recovered capacity somewhere worth going. Neither one on its own gets you there.

An organization that only removes friction becomes efficient. An organization that only chases value becomes well-intentioned and slow. The ones worth working in have been patient enough to grow both.

Which leaves the practical problem of finding out where you currently stand on either one, in a room full of people who have every reason to tell you it is fine.

This cultivates: Everything Was Green, in Part II.