Breadth arrives one defensible decision at a time and departs only by deliberate act, which is why almost every portfolio is wider than anyone chose.
Most executive teams can name their largest line by revenue within seconds. Fewer can name the line that consumes the most executive attention. Fewer again can say whether those are the same line, and the ones who can are usually uncomfortable with the answer.
That gap matters because breadth is never decided; it accumulates. A customer asks for something adjacent and refusing seems churlish. A capable executive wants a new venture and the cost of saying no is a resignation. An acquisition brings three lines when the intention was one. Each decision is defensible alone, none is reviewed collectively, and the portfolio reaches its present shape without anyone having chosen it.
The question in the title is not rhetorical, and it is not an argument for minimalism. It asks for an honest accounting. What is breadth costing in unit cost, in what customers can recall about you, and in the attention of the few people whose judgement the organisation depends on? And if that accounting were done, which lines would survive it?
The Strategic Context
Breadth imposes costs through three channels, and it persists because no single management report shows all three together.
The first is unit cost. More lines mean shorter runs, more changeovers, more variants held, more exceptions and more shared-service complexity. Much of that cost never lands on the line that caused it; it settles into overhead recoveries spread evenly, which makes the marginal line look profitable and the core line look worse than it is.
The second is what customers can recall, which determines whether you are considered at all when a need arises. An organisation known for one thing is retrieved from memory when that thing is needed. One known for seven is retrieved for none, and must buy its way into consideration through selling effort a concentrated competitor does not have to fund.
The third is executive attention, the scarcest input in any enterprise and the only one with no market price. It is allocated by escalation rather than economics: trouble is loud, compounding growth is quiet. Michael Porter's account of generic strategies — cost leadership, differentiation and focus — warned that an organisation failing to commit to one ends up competing effectively on none. The attention channel is a large part of why.
Why Breadth Feels Safer Than It Is
The first misreading treats revenue diversification as risk diversification. Seven lines drawing on the same customers, the same channel, the same balance sheet and the same executive team are one risk wearing seven names, and they will fail together when the shared dependency fails.
The second holds that marginal lines are cheap because they use spare capacity. Genuine spare capacity is rarer than it looks; what is usually consumed is priority — position in the queue, the attention of the best engineer, the slot in the shared system. The marginal line pays an overhead rate and consumes a scarce one.
The third judges lines on contribution margin alone. Margin is measurable and attention is not, so the measurable thing wins the argument. A line returning a modest margin while absorbing a fifth of the executive agenda and a disproportionate share of exceptions in shared functions is not the contributor its margin suggests.
The fourth concerns where the pruning decision sits. It is never made well inside the unit that owns the line, because no leader assesses their own line for removal with detachment and none should be asked to. Pruning is a portfolio act requiring a body with portfolio authority. Where that body does not exist, breadth is permanent by default.
Reframing the Issue
The productive reframe replaces "which lines are profitable?" with "which lines have earned the right to our scarcest inputs — capital, capability and attention — measured against the best alternative use of each?"
That question is uncomfortable because it introduces opportunity cost, which most reporting suppresses. A line can be profitable and still be the wrong holder of a capability that would compound faster elsewhere. Profitability says a line is not destroying value; it does not say the line deserves what it consumes.
This is a portfolio question, distinct from the stage-gate machinery that governs individual initiatives through delivery. Gates decide whether a thing proceeds. Pruning decides whether the organisation should be in it at all, and that question is asked far less often.
Four Tests for What Deserves Concentration
Not every line is a candidate for concentration, and the difference is structural rather than a matter of ambition. Four tests separate what compounds from what merely occupies.
| Test | Diagnostic question | What a failing answer implies |
|---|---|---|
| Is it numerical? | Can the output be expressed in a standard, verifiable unit? | Quality is a matter of opinion, so it cannot be standardised, compared or centralised |
| Is it scalable? | Does serving twice the volume cost materially less than twice as much? | Growth adds cost proportionally; scale buys nothing |
| Is it cash-generating? | Does it convert to cash quickly and predictably? | Growth consumes cash, so success creates a funding problem |
| Is it process-proof? | Does the result depend on the judgement of a scarce individual? | It is a practice, not a platform, and it will not survive their departure |
A line passing all four rewards concentration: investment compounds, quality holds as volume rises, and advantage accumulates rather than resetting with each engagement. A line failing several absorbs investment without compounding, however attractive its margin.
Two qualifications matter. Failing a test is not a verdict of abolition. A line may fail on scalability and still be worth holding as an anchor to a relationship, a proving ground for a capability, or an obligation attached to a licence to operate. The tests say where concentration compounds, not what to abolish, and confusing the two is where portfolio reviews go wrong.
The second is that these tests examine supply, not demand. A line can pass all four and still serve a market too small or too indifferent to matter — a separate enquiry into who has the problem and what they would pay to be rid of it. [Related article: The People Who Have the Problem, and the People Who Want Never to Have It] takes up that demand-side question.
The Attention Ledger
The most useful diagnostic available to an executive team is also the least used. Log where senior attention actually goes — standing meetings, escalations, decisions taken, board items, informal time — attribute each to a line, a customer or a contract, and set the result against economic contribution.
The result is reliably uncomfortable. Attention clusters on the smallest and most difficult customer, the newest venture that has not yet earned its place, and the legacy line nobody will end. Consider a hypothetical illustration: a business with six lines finds that one producing under a tenth of contribution occupies close to a third of the executive agenda. No management pack reveals this, because attention is not a reported quantity. It is visible only when deliberately counted.
Two consequences follow. Attention consumed by a difficult line is a withdrawal from whatever the leadership would otherwise have done, and that opportunity cost falls hardest on the work only senior leaders can do — capital allocation, key appointments, market entry, the decisions that set the trajectory.
The second is structural rather than personal. If growth work happens only in the time left after operations, it will not happen, because operations expand to fill what is available. Protecting a genuine block of leadership time is an organisational design choice about who absorbs operational load, not a matter of individual discipline. Where no one can absorb that load, the answer is a different structure, not a better calendar.
Centralised for Control, Dispersed for Cost — Usually Backwards
Most organisations centralise what is politically sensitive — approvals, spending authority, communications, brand — and leave dispersed what is operationally repetitive: processing, testing, fulfilment, administration. That arrangement is inverted on both counts.
What deserves centralisation is work with high fixed costs, a standardisable method and a quality standard that ought not vary — the back-of-house process. Where every outlet, branch or unit runs its own version of the same activity, the organisation buys the same capability many times over and gets a different answer from each instance. Cost multiplies, quality disperses, and neither effect is visible in any single unit's accounts because each instance is small.
What deserves to stay local is work depending on relationship, language, judgement about a particular customer and speed of response — the customer-facing edge. The test is simple: does the customer care where this is done, and does variation in the output help anyone? Where both answers are no, centralise it.
Two cautions keep this from becoming dogma. A hub is a single point of failure and a queue, and it needs capacity headroom and a genuine service commitment to the units it serves. Without those, the units quietly rebuild local capability and the organisation pays for both — centralisation is usually defeated by the hub's own service failures rather than by resistance.
The second is that a hub built for internal volume often has capacity beyond it, which turns a cost decision into a commercial one. [Related article: When Sharing Infrastructure With a Competitor Is the Right Call] takes up what to do with that surplus. And where the edge is operated by partners rather than employees, their margin determines whether they sell at all; designing that payoff is a discipline in itself, and [Related article: Restructure the Payoff and Let Self-Interest Do the Enforcing] treats it.
Decision Framework
A pruning test is a sequence of questions asked of every line on a fixed cycle, with review rather than continuation as the default. Five carry most of the weight.
Would we enter this line today, given the capital and attention it absorbs and what else we could do with them? A line that would not be started is being continued by inertia, and inertia is not a strategy.
What would we do with the released capacity? If the honest answer is nothing, pruning saves less than it appears, because most of the cost will not leave with the line. Where the released resource is a scarce capability or senior attention, the saving is real.
What does this line carry for others? Shared overhead absorption, channel access, a relationship spanning lines, a regulatory permission. These are legitimate reasons to hold a marginal line, but they must be stated explicitly, because unstated they justify everything.
What is the exit cost? Some lines stop in a quarter; others carry contractual tails, customer obligations and reputational consequences lasting years. Reversibility should set the sequence in which pruning is attempted.
And who loses? Where the people who lose are the people assessing it, the assessment will not happen, and the governance design is itself the finding.
From Strategy to Execution
The immediate work is measurement rather than decision. Build the attention ledger for one quarter and rank lines by contribution per unit of executive attention and per unit of capital employed. That ranking will differ from the revenue ranking, and the difference is the agenda.
The medium-term capability is the hub. Where the four tests indicate that concentration compounds, build it properly: the process, the service commitment to the units, the internal pricing that makes it legible, and the capacity headroom that keeps it trusted. A hub that is cheaper but slower will be abandoned and rebuilt locally at greater cost.
The long-term positioning is recall, which takes years to build and is lost quickly. It is made by consistent excellence in a narrow domain and eroded by every adjacent thing added without conviction. Organisations that hold it treat adding a line as seriously as removing one — the discipline breadth quietly suspends.
Signals to Monitor
Count the lines that received a substantive portfolio review in the past year against the number you have. The gap is the size of the decision nobody is making.
Watch the complexity metrics that precede cost: variants held, changeovers, one-off approvals, exceptions processed by shared functions. These rise before margin falls, and they rise first in the units furthest from the executive team.
Measure unprompted recall in your category rather than asserting it. What customers say you do, without being offered a list, is the honest position.
Track the share of executive agenda time consumed by the bottom quartile of contribution. Watch hub utilisation alongside the satisfaction of the units it serves, and for local capability being quietly rebuilt. And watch where your strongest people ask to be posted — the most accurate internal forecast of where the organisation is going.
Questions for the Leadership Team
- Which of our lines would we enter today, given the capital and attention they absorb, and what would we do with the capacity released by the ones we would not?
- If we logged executive attention for a quarter and set it against contribution, which line or customer would embarrass us most?
- Which lines pass all four tests — numerical, scalable, cash-generating, process-proof — and are those the lines receiving our investment?
- What are we running separately in every unit that ought to be run once, and what is the duplication costing in cost and in variation?
- Which body has the authority to end a line, and when did it last do so?
- What are we known for, in the words customers would use, and is that what we intend?
Closing Perspective
The choice in the title is a false binary read as a rule and a real one read as a discipline. Some organisations should hold seven lines, because their capital, capability and attention genuinely support seven. Very few have tested whether that is true of them.
What separates the two is not the number but whether the number was chosen. An organisation that can name what it would stop, and under what conditions, has a portfolio. One that cannot has an accumulation, and it will discover the difference when capital tightens and every line argues at once that it is essential.
The most revealing question a board can ask is not which line performs best. It is which line the executive team has been unable to end, and why nobody has been asked to explain that.
About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.