Two companies can post the exact same revenue growth this year. One of them is scaling. The other is just getting bigger, and paying more for every additional unit of it, in headcount, in overtime, in the ten small breakdowns nobody had time to fix.
From the outside, the two look identical for a while. Same revenue chart, same headline. The difference only shows up in what it cost to get there, and by the time it's visible on a P&L, it has usually been compounding quietly for a year or two.
This is the actual distinction two separate bodies of business research converged on independently: McKinsey's 2024 study of B2B SaaS companies that had passed $100 million in annual recurring revenue, and Harvard Business Review's 2023 research on what they call the extrapolation stage. Neither is about being a tech company. Both are about what has to exist before growth stops costing you proportionally more.
Key takeaways
- McKinsey's research names it directly: the businesses that actually scale build what they call the engine room first, talent, a growth plan, and structure, not headcount. That's the real difference, not effort or ambition (McKinsey, Jules, Markovitch, and Seiler, 2024).
- Harvard Business Review has a name for the stage most companies skip entirely: the extrapolation stage, the point where each new customer starts adding more revenue than cost. Without it, revenue keeps growing while cost drags along right behind it (HBR, Rayport, Sola, and Kupp, 2023).
- The clearest real example McKinsey cites is not a product change. Netflix redesigned how it ran meetings and cut total meeting volume by more than 60 percent, while 85 percent of employees rated the new format more effective. A process fix, not a headcount fix.
- The math makes the difference concrete: a $50 million company with 20 percent churn needs $60 million in new sales just to double, three times the prior year's total. Get bigger without fixing the engine room and you run that math forever, at a worsening rate.
01Why "growing" and "scaling" get treated as the same word
Most founders use them interchangeably, and most of the damage starts right there. Growing means revenue is going up. Scaling means revenue is going up while the cost of serving each new customer is going down, or at minimum, staying flat.
A ten-person consultancy that doubles its clients by hiring ten more consultants is growing. It has not scaled anything, it has just made a bigger version of the same constraint. The distinction is not academic. It decides whether your business gets easier to run at twice the size, or exactly as hard, times two.
McKinsey's research puts a name on the part that actually determines which one you're doing: the engine room, the team, tools, and systems that produce the work and grow the business, sitting underneath the parts everyone talks about, market fit, funding, and product.
02The stage Harvard Business Review says most companies skip
Traditional business theory splits a company's life into two modes: exploration, where you're testing whether anyone wants what you're building, and exploitation, where growth has slowed and you're fine-tuning your edge. Research published in Harvard Business Review argues there's a stage in between that most companies never manage deliberately, called extrapolation.
During extrapolation, a company is still exploring and still exploiting at the same time, and the entire point of the stage is making sure each new customer brings in additional revenue while costing only marginal amounts to serve. That single mechanic, in the researchers' own framing, is the secret to lasting, profitable growth.
Skip that stage, or run through it on instinct instead of on purpose, and a business keeps adding customers the same way it added its first ten: by adding a proportional amount of people, time, and manual attention behind each one. Nothing about that arrangement gets cheaper as it gets bigger. It just gets louder.
Scaling is not about growing faster. It is about making sure the hundredth customer costs less to serve than the tenth did.
03The five things McKinsey found inside every company that actually scaled
**What builds the engine room, in practice.** McKinsey's research identified five elements that showed up consistently across the companies that scaled well, and dispelled the myths attached to each one along the way.
First, overinvesting in talent early, not once revenue justifies it. Close to 70 percent of the founder CEOs in the research named talent development and culture as their actual competitive advantage, ahead of product. Second, adapting and pivoting continuously, not only when something breaks, the highest performers kept adjusting their market, pricing, and structure even while things were going well. Third, having a real growth plan while still small, the belief that strategy is only for large companies turned out to be the myth most correlated with stalling out. Fourth, organizing deliberately for scale, so the company's processes, structure, and culture don't quietly accumulate what the researchers call organizational debt, breaking down every phase because nobody rebuilt them for the new size. Fifth, running more than one growth engine at once instead of betting everything on a single channel or product line.
None of these five require a specific industry or a specific funding stage. What they require is treating them as infrastructure decisions, made before the business needs them, not repairs made after it breaks.
04The Netflix example nobody expects
When people picture a company scaling, they picture new product lines or new markets. One of the clearest examples in McKinsey's research is neither. It is Netflix redesigning how it ran meetings.
The company shortened meetings, made them more action oriented, and gave employees recordings plus a channel to ask questions that got answered within 24 hours. A cross functional team tracked whether the new format was actually working and adjusted it. The result: total meeting volume dropped by more than 60 percent, and 85 percent of employees rated the redesigned meetings more effective than the old ones.
Nothing about that is a product improvement, and nothing about it required more headcount. It is exactly the kind of unglamorous, process level change that determines whether a company's cost per employee, per decision, per customer goes down as it grows, or stays flat, or quietly climbs.
60%+
Reduction in Netflix's total meeting volume after a deliberate process redesign, with 85% of employees rating the new format more effective (cited in McKinsey, 2024, sourced to Harvard Business Review, 2014)
05What NOT to do
Do not treat headcount growth as a proxy for scaling. Adding people at the same rate revenue grows is the definition of just getting bigger, not evidence that anything is working.
Do not wait for 'we'll fix process once we're bigger.' McKinsey's research found the opposite pattern in the companies that scaled well, they built the growth plan and the structure while still small, before the business forced their hand.
And an honest limit on this research itself: McKinsey's study looked at B2B SaaS companies that had already passed $100 million in annual recurring revenue, a specific and successful population. The five elements transfer to other industries and other stages, the researchers say so directly, but a five-person or fifty-person company should treat this as a direction to build toward, not a checklist it should already have fully in place.
06Getting bigger vs. scaling, side by side
The distinction holds up better as a comparison than as a definition. Most businesses will recognize themselves somewhere on the left column, at least in a few rows.
| Getting bigger | Scaling |
|---|---|
| Headcount grows at roughly the same rate as revenue | Headcount grows slower than revenue, or in bursts tied to a specific new capability |
| Process gets fixed after it breaks | Process gets built ahead of the size that would break it |
| One growth channel carries almost all new revenue | Multiple growth engines run at once, so no single one has to carry everything |
| Strategy conversations start once things feel out of control | A growth plan exists while the company is still small |
| Each new customer costs roughly the same to serve as the last one | Each new customer costs less to serve than the one before it |
07Getting started: the two-question audit
1. Pull your last four quarters of revenue growth next to your headcount growth over the same period. If the two lines move almost exactly together, that is the getting-bigger pattern, not evidence of a healthy business.
2. Estimate what it actually costs, in people-hours and systems, to onboard and serve your most recent ten customers versus your first ten. If the number per customer has not come down, the engine room McKinsey describes has not been built yet.
3. Name your organizational debt honestly. Which process, team structure, or tool still reflects the company at half its current size? That is where the next breakdown is already scheduled.
4. Check how many real growth engines the business is running. One channel carrying all of it is fragile by definition, regardless of how well that one channel is performing right now.
Frequently asked questions
Growing means revenue is increasing. Scaling means revenue is increasing while the cost of serving each additional customer stays flat or falls. A business can grow for years without ever scaling, and the cost of that shows up as headcount, overtime, and process strain that increases in lockstep with revenue instead of falling behind it.
It is a term from Harvard Business Review research describing the developmental stage between exploring whether a product works and exploiting an established advantage. During extrapolation, a company deliberately works to make each new customer profitable at the margin, not just at scale eventually. Most companies pass through this stage on instinct instead of on purpose, which is exactly where the getting-bigger pattern takes hold.
Compare your headcount growth rate to your revenue growth rate over the last four quarters, and estimate the cost to onboard and serve a recent customer against one of your first ten. If headcount tracks revenue almost exactly and the per-customer cost has not fallen, the business is getting bigger, not scaling.
McKinsey's underlying study looked at B2B SaaS companies that had already passed $100 million in annual recurring revenue, but the researchers note directly that the five elements they found are not exclusive to SaaS and apply across industries. For a smaller company, the five elements are a direction to build toward deliberately, not a list that should already be fully checked off.