Analysis · How to digitise
Linear and scalable growth are the two ends of one spectrum. Four questions show where your business sits and how automation moves it to the right.
In short
Updated September 2026 · 10 min
If you want to double your revenue, do you need to double your team? The answer says where your business sits on the spectrum between linear and scalable growth; on the platform we hear it from every client who starts with one automated process. This analysis describes the two ends, the trap between them, the four questions that show your position and the steps that move you to the right.
In a linear business, doubling revenue means doubling the team; in a scalable one, revenue grows faster than costs, and most businesses sit in between. Linear growth adds resources in proportion to output: twice the customers means twice the people and twice the costs, with the margin unchanged. Scalable growth has leverage in the model, a product, an automated process or an asset that serves more customers without proportional effort.
Economists call that leverage economies of scale: the fixed cost is spread over more units, and the cost per unit falls as output grows (Corporate Finance Institute, 2019). The extreme end is the one Reid Hoffman described in the “Blitzscaling” interview in Harvard Business Review (2016): technology companies that grow much faster than the team building them. Few SMBs will get there, and they do not need to.
Between the ends sit four examples. A traditional law firm is on the left: every billed hour requires an hour of work, and you cannot bill a hundred hours with fifty. A marketing agency with templates and standard processes is centre-left: part of the work is reused, but delivery stays manual.
A manufacturer with partial automation is centre-right: some stages scale, such as the automated line, others do not, such as custom assembly. A software-as-a-service product, SaaS, is on the right: the same server serves a hundred or a hundred thousand users, and the cost of one more customer tends towards zero.
The position matters even if you do not move: a linear business grows only as fast as it hires and trains people, and its margin does not improve with volume. The owner stays stuck in operations, and a buyer pays less at exit. Linear businesses can be stable and profitable; it is just that every step of growth costs them as much as the previous one.
| Linear growth | Scalable growth | |
|---|---|---|
| Cost of a new customer | The same as the previous one | Lower than the previous one |
| Margin when you double volume | Unchanged or lower | Grows |
| Growth ceiling | How fast you hire and train people | How fast you sell |
| Owner's time | In operations, proportional to the number of customers | In decisions, independent of volume |
| Valuation at exit | Depends on people, low multiple | Depends on the system, higher multiple |
Automation is one of the few tools that move a traditional business towards the scalable end: it separates the volume of work from headcount. When you automate a process, the same people handle more volume, new customers no longer require proportional hiring and the margin improves as you grow. It is not the only tool, but it is the one that does not ask you to change what your business does.
The best candidates are the processes that look the same whether you have ten or a hundred customers. Invoicing, quoting, reporting and starting a new project have the same steps every time, with different data. That is exactly why they grow linearly when done by hand: every extra customer adds the same hours.
After automation, the person stays in the process, but on the exceptions. The bookkeeper no longer enters invoices but checks what did not match; the engineer no longer writes quotes but checks them; the director no longer collects figures but reads them. The difference is not that they work less, but that their workload no longer grows with the number of customers.
Before
After automation
There is a worse position than linear: sub-linear growth, in which costs grow faster than revenue, usually in a business that is growing fast. It happens when growth is rapid but chaotic, when processes do not keep up with volume and when you hire reactively, at every crisis, instead of systematically. Quality problems add rework and commercial discounts, and every new customer costs more than the previous one.
The trap is that it looks like success. Revenue grows and is celebrated in the meeting, while the margin erodes quietly, in a number nobody tracks monthly. The warning sign is personal before it is financial: growing revenue, a more stressed owner and less profit at year end.
Businesses usually get here after a good year. They took on more customers than they could serve with the old processes, hired quickly, and the new people inherited the manual way of working at double the volume. The way out is the same move to the right, only it starts from lower down and with less time.
Four questions give your position: repeatable work, what breaks under volume, where expensive people spend their time and the cost of a new customer. None of them needs a consultant or a report; they need an hour with the management team and honest answers. The answers change over time, so the questions are asked once a year.
The first question is what share of the work is repeatable and what share is custom; the more repeatable, the more potential for scaling. The second is what would break if you got ten new customers tomorrow; whatever breaks first is your scalability bottleneck. The third is what your most expensive people spend their time on; if it is repetitive work, that is the first automation opportunity.
The fourth question is the decisive one: does every new customer cost less to serve than the previous one? If yes, you already have elements of scalability. If the cost is constant, you are linear; if it grows, you are sub-linear, and the next question is no longer about growth but about processes.
Where you sit on the spectrum
Moving right is incremental: you identify the processes where human time grows with volume, prioritise them by impact, automate one and measure what changed. The goal is not to become fully scalable, but more scalable than you are today and than your competitors. One process per quarter, measured, beats a three-year transformation plan.
The first step is the map of linear traps: the processes in which people's time grows directly with volume, so the places where growth requires proportional hiring. The second is prioritisation, because not every linear process is worth automating. Worth it are the high-volume ones, daily or weekly, with clear rules and no complex judgement on every case, and those that are already the growth bottleneck.
The third step is automating a single process, taken all the way and with its value proven before the next one. The fourth is measurement, with four metrics: revenue per employee, gross margin, the onboarding time of a new customer and capacity utilisation. Without step four, the second process is chosen on impressions, and impressions always favour what hurts today, not what scales.
Four steps to the right
Map the processes and mark those in which people's time grows directly with volume.
Pick the high-volume processes, those with clear rules and those that are already the growth bottleneck.
A single process, taken all the way, with its value proven before the next one.
Revenue per employee, gross margin, onboarding time of a new customer, capacity utilisation.
An MEP (mechanical, electrical and plumbing) firm with 35 employees automated two of four linear processes and moved three metrics in six months, without hiring. The firm is fictitious; the scenario is the one we see at technical service firms with twenty to fifty people. At the start, four processes grew linearly with the number of projects:
| Process | How it was | What was automated | Metric, month 0 → month 6 |
|---|---|---|---|
| Quoting | An engineer wrote every quote from scratch, 3–4 hours | The quote is generated from the product library; the engineer checks it | Quotes per week, per engineer: 4 → 11 |
| Invoicing | The bookkeeper re-entered every approved estimate in the invoicing software | The invoice is issued from the approved estimate | Days from handover to invoice: 9 → 2 |
| Reporting | The director collected Friday figures from three Excel files | The Friday report generates itself | Reporting hours per week: 6 → 0.5 |
| Starting a project | Project files and folders created by hand | Untouched; it is the next process | Days to the first complete file: 2 → 2 |
Fictitious data. The figures are not promises but the order of magnitude we see at firms of this size after the first process is taken all the way.
On the five lines of the comparison between linear and scalable growth, the firm started with every answer on the left. After six months, two had moved to the right. The cost of a new customer fell, because the quote and the invoice no longer require engineer and bookkeeper hours, and the owner's time was freed from the Friday report.
The margin and the growth ceiling have not moved yet, and that is normal: they move when volume grows with the same team, not in the month of automation. Revenue per employee grew slightly, with the same 35 people and three more projects in progress. The next process is project start-up, chosen because it became the bottleneck on those three new projects.
The spectrum serves a business that sells an outcome, not a person's hour; in three situations moving right makes no sense or is not possible.
Put the four questions to the management team in a one-hour meeting, with the answers written down, and pick a single process from the map of linear traps. If the chosen process needs an automation plan, the business process automation guide shows how the first process is chosen, started and measured.
A business in which revenue grows faster than costs: a new customer costs less to serve than the previous one, and the margin improves as volume grows. It is not a category but one end of a spectrum; almost no business is fully scalable, but any business can be more scalable than it is today.
Not fully, and it does not need to. A service business sells people's time, but around that time sit processes that do not have to grow with it: quoting, invoicing, reporting, onboarding. When those are automated, the same people serve more customers, and the business moves to the right without changing what it does.
Revenue per employee is the main one: if it grows from one year to the next, the business is moving to the right. Complete it with gross margin, the onboarding time of a new customer and capacity utilisation. Four numbers tracked quarterly say more than any diagnosis done once.
The first automated process shows in its own metric within two or three months: reporting hours, days to invoice, quotes per week. Revenue per employee moves more slowly, because it depends on sales, not only on capacity; a fair comparison is year on year, same season. Do not wait for the annual accounts to continue.
One workflow goes live on a real project, with a success criterion set together.
One email a month, only when we publish. Nothing else.
Corning, Volvo and Cox Automotive automated their data flows to decide on today's numbers. A 50–100 employee company has the same problem, at a different scale.
Read →AI automates well only what it understands, and it understands only what is clean. Digitisation → context → automation, explained on one real case.
Read →Performance drops in the first weeks after go-live and climbs above the old level from month three. The stages, the critical moment and seven strategies.
Read →