Why a Growth Machine Has Become Essential to Scale an Industrial SMB

In most industrial SMBs and mid-market B2B companies I meet, growth rests on three things.

The founder’s network. Two or three tenured salespeople. And a volume of referrals that goes up and down quarter to quarter without anyone really knowing why.

When it works, we call it “organic.” When it slows down, no one knows which lever to pull.

That’s exactly the symptom of a company that doesn’t have a machine — only people. And that’s exactly why it doesn’t scale.

The diagnosis I make in nine companies out of ten

The pattern is almost always the same. The company is between ten and thirty years old. It has reached a plateau — often between €500k and €5M in revenue. It remains profitable, sometimes comfortably. But growth is no longer systematic.

When you look at where new customers come from, you always find the same sources: the founder’s rolodex, referrals from existing customers, two or three salespeople hunting old-school style, and a trade show presence that produces an irregular pipeline.

This isn’t a sales problem. It’s a systems problem. The company doesn’t have a machine — it only has people.

Why person-dependent growth always hits its ceiling

A model that relies on individuals has four structural limits that everyone eventually encounters.

It’s not predictable. You can’t tell your board or your shareholder which quarter the two big deals will close. Forecasts are disguised bets.

It doesn’t survive turnover. The day your tenured salesperson leaves — to a competitor, to retirement, to a personal project — a significant portion of your pipeline leaves with them. In the companies I work with, that’s regularly 30 to 50%.

It exhausts those who carry it. The founder who continues to close deals themselves is capping their company at the threshold of their own energy.

It severely penalizes your valuation. A private equity fund, a strategic acquirer, an industrial partner — all discount a business whose revenue is attached to individuals. What they’re buying is a revenue machine: a system that produces pipeline independently of its individuals. At comparable EBITDA, this system structurally commands a higher price than a rolodex that will leave with its owner. The gap is far from anecdotal — it’s measured in multiples, not points.

As long as these four points aren’t addressed, the company remains a good business — but a capped one.

What a growth machine really is

The term is misused. Most executives hear “machine” and think “we hired two SDRs” or “we launched paid.” That’s not it.

A growth machine is a structured system that combines five elements:

A precise definition of who you’re targeting, segment by segment, with measurable criteria.
Repeatable and documented plays to enter into conversation with these targets — outbound, inbound, content, paid, partnerships — independent of the individuals executing them.
Consistent qualification criteria applied by everyone, from SDR to Account Executive.
A pipeline measured at each stage, with known conversion rates by segment.
A decided level of investment, not a residual budget.

Above all of this, one non-negotiable characteristic: the machine produces opportunities continuously, independently of any particular person. The day your best salesperson leaves, your pipeline should drop by 10%, not 40%.

If that’s not the case, you don’t have a machine. You have people doing sales.

The method comes down to four movements

You don’t build a growth machine by buying a tool or hiring. You install it in a precise order, and it’s the order that makes all the difference:

Data before ambition — define the ICP from what you already serve, not what you dream of serving.
The 80/20 focus — decide which segments you industrialize, and which ones you simply serve.
The GTM matrix — one go-to-market per segment, not a single go-to-market applied to everyone.
The unit economics filter — let LTV and CAC decide where capital goes.

The rest of this article unfolds these four movements.

Movement 1 — Data before ambition

The mistake I see most often in established SMBs is the deep conviction that the solution addresses a broad market.

After fifteen or twenty years of existence, the company has accumulated customers in very different sectors, very different sizes, with very different use cases. Each of these customers was obtained by opportunity — a salesperson who had a contact, a trade show that went well, an existing customer who referred.

The result: a fragmented portfolio, and an executive convinced that “our solution can apply everywhere.”

That’s true technically. It’s false strategically.

What got you to €1M won’t get you to €10M. Scaling requires a focus that the previous phase didn’t demand. Wanting to address everyone always produces the same thing: dispersion, complexity, and operational efficiency that collapses.

The ICP — Ideal Customer Profile — is therefore not defined in a strategic seminar. It’s defined in an Excel export.

Concretely: you pull out the list of your customers from the last 36 months. You segment by industry, size, use case, acquisition channel. For each segment, you calculate:

Average basket size
Gross margin
Sales cycle duration
24-month retention
Expansion rate (upsell, cross-sell)
Actual acquisition cost when calculable

The segments that come out on top on these axes are your real ICP. Not the ones you talked about at your last offsite. Not the ones you’d like to serve. The ones you already serve, successfully, profitably.

This distinction is central. Defining the ICP from ambitions rather than data is almost always a guarantee you’ll invest in segments that won’t convert.

Movement 2 — The 80/20 focus

In a company that’s been around for ten years, you can easily count ten, twenty, sometimes thirty different ICPs — because each product, each geography, each tenured salesperson has built their own niche over time.

Taken individually, each of these segments has generated value. Taken collectively, they create a dispersion that paralyzes the rest of the company:

Marketing tries to address everyone and reaches no one strongly
Sales can no longer identify a “good” lead
Onboarding is custom every time
The product is pulled in twenty directions
Pricing becomes incoherent

As long as this fragmentation exists, the machine can’t be built — because there’s no production unit to industrialize.

The Pareto law applies here with almost mechanical regularity. In most B2B portfolios I analyze, about 20% of customers generate the bulk of revenue, an even smaller subset concentrates margin, and a handful concentrates nearly all expansion and long-term retention.

The work consists of identifying this subset — not to abandon the rest, but to decide where acquisition effort is invested:

On your top 20% segments: you build the machine. Outbound, marketing, content, sales, onboarding — everything is sized to scale these segments.
On the remaining 80%: you maintain, you serve well, you take what comes, you cross-sell what’s relevant — but you stop investing in acquisition.

This decision frees up often considerable capital of attention and budget. It’s this capital that we reinject into the machine. And this focus work isn’t a marketing job: it’s a strategic executive job.

Movement 3 — The GTM matrix, segment by segment

Once priority segments are identified, you structure a GTM matrix segment by segment. For each, you explicitly define:

A memorable name — e.g., Newbie, Rockstar, Diva, etc.
The sales motion: product-led, sales-led, or hybrid
The team structure: AI Agent + Human ratio (SDR / Account Executive / Account Manager)
The marketing mix: paid, content, events, partnerships, in-app notifications, etc.
Onboarding complexity and resulting cost — e.g., dedicated session, self-serve, etc.
Post-sales follow-up: self-serve, FAQ, chatbot, Account Manager, Key Account, etc.
Average cycle duration and number of touches required
Expansion potential at 12, 24, 36 months

Some segments will naturally be product-led, self-serve, with very little human intervention — typically small accounts or simple use cases.

Others will require complex onboarding, personalized demos, long cycles, and dense human support — typically enterprise accounts or regulated use cases.

You can’t apply the same GTM to a self-serve segment with €5,000 average basket and an enterprise account at €250,000. Yet most companies try — because they haven’t done this matrix work upstream.

Movement 4 — The unit economics filter

For a growth machine to be useful, it must produce profitable revenue. Not pipeline.

For each segment, you must be able to calculate:

LTV (Lifetime Value) — the value generated by a customer over their total lifetime
Complete CAC (Customer Acquisition Cost) — including sales, marketing, and onboarding costs
Payback period — the number of months to recover CAC
Gross margin per segment
Net Revenue Retention by cohort

These metrics are exactly the ones tech leaders and SaaS scale-ups use to drive their GTM investments. They’re not reserved for SaaS. An industrial SMB that calculates them makes infinitely better decisions than an SMB that pilots by feeling.

The fundamental point: a segment with excellent product fit but failing unit economics is a trap. You’ll burn capital scaling something that doesn’t generate intrinsic value.

Conversely, a segment with healthy unit economics but imperfect product fit deserves investment in product correction. Because the economic engine is viable.

The rule of 3 — and the real ambition at 4 or 5

A simple rule, used just about everywhere in B2B:

A healthy business at scale aims for an LTV/CAC ratio ≥ 3.

In other words: the value generated by a customer over time must be at minimum three times greater than the complete acquisition cost.

Below 3: you’re burning capital to grow. This is viable for a moment with fundraising, never long-term.
At 3: you have a viable machine. You can scale cleanly.
At 4 or 5: you have a defensible economic moat. This is typically the ratio of companies that get the best multiples in M&A or LBO.

Most established SMBs have never calculated this ratio — often because their growth is too person-dependent for it to be properly calculable.

The day you can calculate this ratio segment by segment, you’ve started building a real machine. Before that day, everything is intuition.

Where to start this week

Three concrete movements, in this order, to start in the next 14 days:

One — Export your customer base from the last 36 months. For each, three numbers: cumulative revenue, gross margin, active/inactive status. Group by segment. Sort by descending value. The real picture of your portfolio, in one afternoon.

Two — Take your top 20% segment. Write in one sentence: who is the buyer, what is the trigger, what is the value proposition that made them sign. If you can’t write it in one clear sentence, you don’t have an ICP yet — you have an ambition.

Three — Choose a single channel to test in the next 60 days for this segment. Just one. Not three. Define what success looks like: number of qualified meetings, expected conversion rate, pipeline created. Build the machine on this channel before adding a second one.

The discipline of advancing channel by channel, segment by segment, is exactly what distinguishes companies that scale from those that remain at the ceiling.

The short conclusion

The transition from person-dependent growth to system-driven growth is the most important transition for an industrial SMB or mid-market B2B company.

It’s also the most uncomfortable. It requires admitting that the model that got you to your current revenue is not the one that will triple it. It requires choosing which customers you stop chasing. It requires building an infrastructure invisible from the outside but which determines your trajectory for the next five years.

A growth machine is not a tool you buy. It’s a discipline you install.

Most companies never build it — because it’s slower and less heroic than a big deal closed by the founder. Those who build it are the ones private equity funds target, strategic acquirers court, and who pass the thresholds others stand watching from afar.

The others remain good companies. But capped.

If you want to know which of your segments deserves to have the machine installed first, that’s exactly what I do in one diagnostic session.

Thibaut Vanderhofstadt is co-founder of Sortlist — Europe’s leading B2B matchmaking platform, active in over 140 countries. He now works with SaaS founders, marketplace operators, and ambitious SMB and mid-market executives through MetSaaS. Book a diagnostic →