Choose the 80 Before You Design the Enterprise

• A concentration curve is a diagnosis of the past. Strategy is a forward choice about where to play and how to win. Most enterprises have the first and assume it substitutes for the second.
• Front-to-Back 80/20 is the discipline of connecting that choice through the whole enterprise — segment, offer, price and mix, operations, supply, service, support cost and cash — rather than running it as an internal cost-reduction program.
• The evidence that a choice was actually made is a stop / defer / de-invest decision with reallocated capacity behind it. A ranked priority list is not evidence. Nothing was chosen until something was released.
• Channel economics are usually the least examined part of the picture and often the most expensive: the same product sold through the wrong channel carries a different cost to serve, a different working-capital load and a different margin.
• Do this before organization design. A clean sheet cannot be drawn against an unstated 80 — the structure would simply re-encode today's spread.

The step most leadership teams skip

The recognition arrives quietly and usually correctly: the design that produced the current business is no longer producing the results the next stage requires. Effort is up, throughput is flat, decisions take longer than they used to, and every function has a defensible reason for the load it carries.

What happens next is predictable. The team reaches for structure — a reorganization, new roles, a re-drawn chart — or for cost, in the form of a percentage taken out of overhead. Both are architecture decisions made before the strategic decision that should govern them. Structure without a stated choice re-encodes the spread that created the constraint. Cost without a stated choice removes capacity indiscriminately, most often from the parts of the business with the least political weight rather than the least economic value.

The missing step is a decision about concentration. Not an analysis of where value came from, but a commitment about where the enterprise will put the next unit of capital, capacity and leadership attention — and where it will stop.

Structure without a stated choice re-encodes the spread that created the constraint.

A concentration curve is not a strategy

Most enterprises can produce the curve. Twenty per cent of customers, products or geographies carry the majority of the economic value, and a long tail consumes a disproportionate share of capacity. Reading that curve honestly is real work, and it is where the discipline begins.

But the curve describes the business the enterprise already has. It says nothing about the business it intends to have. A segment can be small today and be exactly where the next stage lives. A segment can be large today and be structurally in decline. Ranking history and calling it strategy is how a company concentrates itself into a shrinking position with excellent margin discipline.

So the curve is the input and the choice is the output. The choice has two halves, and both must be written down. Where will we play — which markets, customer types, product families and channels are we building the enterprise to win? And how will we win there — on what basis will a customer in that segment choose us over the alternative they actually have?

The second half is where most statements fail. "Quality and service" is not a basis for winning; it is a description of not losing. A basis for winning is specific enough that it implies a cost structure, a lead time, an engineering standard, a channel and a set of capabilities the enterprise must be able to fund.

The four choices, in order

The choice is made across four connected dimensions. Taking them out of order is how a plan comes apart in execution: a channel commitment that the product portfolio cannot supply, or a customer promise the operation was never designed to keep.

1. Markets

Where the next stage of demand actually is

Which markets are structurally growing, which are consolidating, and which have moved past what the enterprise is built to serve? The test is not attractiveness in the abstract but fit: does this market reward something this enterprise can be genuinely better at, and can it be served without a second operating model bolted onto the first? A market that requires a parallel cost structure is not an adjacency; it is a second company.

2. Customers

Who the enterprise is built to serve fully

Named customer types, not revenue bands. For each, the requirement set is written out: lead time, engineering involvement, order pattern, terms, service intensity, quality regime. Two customers with identical revenue can carry completely different demands on the enterprise. The choice is which requirement set the operating design is going to be optimized for — because it cannot be optimized for all of them at once, and pretending otherwise is what produces the exception load.

3. Products and offers

The portfolio the chosen customers require

Products follow customers, not history. The register lists what is core to the chosen segments, what is a strategic enabler that strengthens the system solution even at modest standalone margin, what is opportunistic and maintained without further investment, and what is a complexity trap consuming engineering, tooling and floor capacity out of all proportion to its contribution. Retaining a low-margin enabler is a legitimate decision. Retaining a complexity trap because no one will say so is not.

4. Channels

The most expensive unexamined decision

The same product through a different route to market is a different business: different cost to serve, different working-capital load, different price integrity, different information about the customer. Direct, distribution, rep, OEM, integrator and digital each carry an economics the P&L rarely separates. The choice names which channels carry the chosen segments, which are being deliberately narrowed, and what the enterprise will stop paying to support in the channels it is leaving.

Allocation is the decision — a list is not

A choice that changes nothing about where resources go was not made. This is the point at which most 80/20 work quietly becomes a communication exercise: the deck names the vital few, and the following quarter every function still funds what it funded before, because nothing was taken off anyone's plate.

The evidence of a real decision has three parts. First, a stop / defer / de-invest ledger — the named work, accounts, products, channels or initiatives the enterprise is releasing, with a date and an owner. Second, a capacity read: what those decisions actually free, in the units that constrain the business — engineering hours, floor time, senior leadership attention, cash. Third, the reallocation itself: where that freed capacity is now committed, and who is accountable for the outcome it is meant to produce.

This is also the honest test of whether the enterprise can execute at all. An organization that cannot name one meaningful thing it will stop doing does not have a capacity problem it can solve with better prioritization. It has a governance problem, and the next stage will be funded by exhaustion.

Nothing was chosen until something was released.

Front to back: connecting the choice through the enterprise

Front-to-Back 80/20 is not the internal half of the same idea. It is the discipline of carrying one choice through every link in the chain so the enterprise is allocated to the same few things end to end.

The chain runs in one direction and each link is a design decision, not a report. Chosen segments set the offer and the promise. The offer sets price and mix discipline — including where the enterprise declines work rather than absorbing it at a price that funds someone else's complexity. Price and mix set what operations must be designed to do repeatably, which sets the supply base, inventory posture and lead-time commitment. Those set the service and support model, which is where cost to serve is actually incurred. And all of it lands in cash: working capital, capital expenditure, and the funding available for the next stage.

Run this way, the exercise produces margin the same revenue was never producing before, because the cost of serving the chosen 80 falls while the cost of subsidizing the rest is either priced, standardized or released. Run as a cost program, it produces a temporary saving and a slower business, because capacity is removed without any change in what the enterprise is trying to be.

Then, and only then, design the organization

With the four choices written down, the stop ledger signed and the front-to-back chain connected, organization design becomes a tractable question instead of a negotiation. The structure has something to be right about: the roles, decision rights, spans and hand-offs the chosen strategy requires.

Attempted in the other order, a clean sheet inherits the ambiguity. Every role gets defined to cover every segment, decision rights are written to avoid offence rather than to place authority, and the new chart reproduces the old spread with new titles. The design was never wrong; it was answering a question no one had asked.

This is why the sequence matters more than the sophistication of any single step. Recognize that the architecture is the constraint. Choose the 80 and connect it front to back. Design the organization the choice requires. Only then treat acquisition, new capacity or new geography as mechanisms inside a coherent strategy — rather than as the strategy itself.

The choice register

One page, filled in by the executive team together. If a row cannot be completed, that row is the work — not a reason to move on.

The fourth column is the one that makes the first three real.

Where to look first

• Whether a written statement exists of the markets, customers, products and channels the enterprise intends to win — dated, and agreed by the executive team.
• Cost to serve and working-capital load separated by channel, not just by product or customer.
• One meaningful thing the enterprise stopped doing in the last two quarters, with the capacity it freed and where that capacity went.
• Requirement sets the operation is absorbing as exceptions rather than pricing or declining.
• Whether any organization-design work underway is anchored to a stated 80 or to the existing chart.

Every enterprise is already allocated to something. The only question is whether it was chosen deliberately, front to back — or inherited one reasonable decision at a time.

SIMPLIFY™ names the vital few and Front-to-Back 80/20™ carries that choice through the enterprise; GROW™ is what the freed capacity is committed to.