Finding and working opportunities in industrial SaaS

The problem isn't the 'who'

In classic B2B SaaS, the business developer's question is 'who do I talk to?'. The market is vast, the list is almost infinite, and the work is about slicing it: ICP, segments, personas.

In industrial SaaS, that question is already settled. If you sell CMMS to manufacturers of 50 to 500 employees, your market fits in one file. A few thousand accounts. A few hundred if you specialize in one sector. You know them by heart, and your two competitors have the exact same list you do.

The consequence is brutal: you can't play the volume game. A finite list you spray twice a year is a burned list. Every off-target contact is expensive, because the account doesn't renew, there is no other one behind it.

So only one variable remains: when.

A production manager doesn't buy a CMMS because a BD sent a good email. He buys it because he just lost three days of production to a breakdown, because an audit flagged a gap, because he hired a maintenance manager arriving from a site that was already equipped, or because an investment budget was just voted through. Your email lands at that moment, or it doesn't land at all.

A signal is a way of knowing that this moment is happening.

A signal is a public, dated, observable fact that raises the odds an account meets the three conditions of a purchase: a mandate, a budget, a deadline.

Three criteria, all three required:

• Observable - someone published it. It's not a hunch, nor an extrapolation from headcount.
• Dated - you know when it happened. No date, no window. No window, no decision.
• Actionable - it implies a concrete consequence inside your own scope.

What looks like a signal without being one:

• 'Company is growing' - that's a state, not an event. True for three years, so useless today.
• '200 employees, food industry' - that's a targeting criterion. It says who, not when.
• 'Visited our website' - that's intent. Useful, but it tells you they found you, not that they have a project. In industry, the person with a project doesn't start with a Google search: he calls his counterpart in the region.
• 'Liked a post about Industry 4.0' - noise.

1. Hiring - the most reliable of all

A company doesn't create a role by accident. A role created is a mandate assigned, a budget voted, and an implicit deadline. And the posting is public, dated, and often talkative: it describes the project better than any discovery call would.

Which roles to watch depends on what you sell, but the logic is constant: look for the role whose very existence presupposes your product.

• CMMS / EAM → 'Maintenance Manager', 'Maintenance Methods', and the most explicit of all: 'CMMS Technician'
• MES / industrial performance → 'Continuous Improvement Manager', 'Lean Engineer', 'Digital Factory Manager'
• Quality / traceability → 'Quality Manager', 'QHSE Engineer' at a certified site
• Supply chain → 'Buyer', 'Scheduler', 'S&OP Manager'

The decisive nuance: role created or replacement. A replacement is continuity, nothing is going to change. A created role is a project. The posting almost always says so: 'newly created position', 'as part of our restructuring', 'new department'.

2. Industrial investment

New line, extension, new site. A building coming out of the ground is a plant to equip. Strong signal, but early and narrow window: the useful moment sits between the announcement and commissioning. A site that's been inaugurated is a site already equipped, by someone else.

Where it shows: press releases, regional press, trade press, building permits, chamber-of-commerce announcements, site-location files.

3. Leadership change

New site director, new COO, new CIO. An incoming leader has roughly a hundred days to show they're changing something: it's the only period where inertia works for you rather than against you.

The detail worth gold: look at where they come from. If they arrive from a site equipped with a tool in your category, they won't wonder whether they need one. They'll wonder which one.

4. Regulatory or normative constraint

It's the best kind of deadline, because it isn't negotiable. European directives, sector standards, certifications to renew, audits: it doesn't matter which one, what matters is that it imposes a date and makes 'we'll see next year' impossible.

But careful: the regulation itself is not a signal. Everyone knows it, and everyone sends the same email the same month. The signal is the account that just ran into it: an audit passed, a gap flagged, a certification to redo, a QHSE hire.

5. Money coming in

Fundraise, LBO, buyout, grant. The first three are over-exploited, everyone sends 'congrats on your round' the day after the announcement. The signal is still good, but the useful window isn't the next day: it's a few months later, when the money turns into a plan.

Productive-investment subsidies, on the other hand, are public, dated and under-exploited. A grant obtained is an earmarked budget with a disbursement schedule. Few people read them.

6. The public incident

Product recall, line stoppage, non-conformity, accident. The most powerful and most delicate signal. Powerful, because it unlocks in three weeks a budget that had been stuck for two years. Delicate, because the way you talk about it decides everything.

You never write 'I saw your product recall'. You write about the subject, never about the incident.

7. The end of a contract

Rare, hard to see, decisive when you do. It shows up in the negative, almost always in a job posting: the words 'migration', 'overhaul', 'replacement of the existing tool', 'data history takeover'.

This is the point most BDs miss, and it's the one that costs the most. A signal isn't true or false. It's on time, or not.

These durations aren't laws: they're orders of magnitude to recalibrate on your own market. What matters is the principle - a signal has an expiry date, and it's shorter than you think.

Two symmetrical mistakes:

Too early. You arrive before the mandate exists. Nobody has the power to say yes. You get a polite reply and get filed under 'revisit later' - that is, nowhere.

Too late. The project is scoped, the shortlist is set, or the tool is already installed. You're the third courtesy quote, the one used to validate the price of the other two.

You have the signal, you have the window. What's left is the most thankless part: finding who to talk to, and how to reach them. This is where most sequences die, not on the message, on the recipient.

Three roles around every signal

A signal sets three people in motion, and they're rarely the same:

• The mandated one - the person handed the problem. He has the pain, and very often, not the budget.
• The sponsor - the one who voted the envelope. He ignores the detail, but he knows the date.
• The future hire - in the case of a recruitment, he isn't there yet. Writing to the role is writing to no one.

The rule that works: talk one level above the problem. The maintenance manager suffers the breakdown, but he doesn't sign. The CEO signs, but he'll never reply to you. The industrial director has both: the pain and a hand on the envelope.

On a job posting, it's written in black and white: the person to contact isn't the role being hired, it's who they'll report to. The posting almost always spells it out.

LinkedIn doesn't start conversations in industry

In tech SaaS, your buyer lives on LinkedIn. In industry, no. A maintenance manager at a 120-person SME has a profile created ten years ago, a photo of the same vintage, and logs in three times a year. A site director spends his day on the shop floor, not on a feed.

If your sequence relies on LinkedIn messages, you're talking to a wall. That's the bad news, and it's the one everyone already knows.

But acceptance is a dated signal

Flip the problem. The fact that your contact doesn't live on LinkedIn is exactly what makes the invitation interesting.

An invitation asks for nothing. It costs nothing to accept, nothing to ignore, and it contains no sales attempt. It's the only zero-cost move in your entire sequence, and the only one that can sit pending for six weeks without expiring.

And when it's accepted, it tells you something nothing else ever will: this person just logged in. Of the three times a year they open LinkedIn, you've got one. It's your only window of attention on that account, and it lasts about forty-eight hours.

An acceptance isn't a lead. It's a timestamp. And it's the only signal in this whole guide that comes from the person themselves rather than a public source.

The real problem isn't sending, it's not missing

You send twenty invitations on a Monday. Three accept on Tuesday. Two on Friday. One three weeks later, another in October.

And then: nothing. No useful notification, no queue, no view of 'who just accepted me and I've never approached'. LinkedIn announces a new connection in the middle of forty other notifications, and that's it.

The result: the vast majority of acceptances die on the spot. You rediscover them six months later scrolling your connections, with that faint unease of recognizing a name without remembering why you added it. Meanwhile, the signal that triggered the invitation has expired.

The job isn't sending the invitations. It's not missing the acceptances. It's follow-up work, not volume, and it's exactly the work nobody does, because it's neither gratifying nor visible.

Cadence

LinkedIn caps invitations. The cap isn't public, it varies by account, and it moves. Never mind the exact number: the principle is that an account sending everything at once gets restricted, and a restricted account is no longer a channel at all.

Spread them out. A small daily dose held for months beats a burst that costs you the account, especially since in a finite market, you don't get a second list.

The upshot of all this: multichannel isn't sophistication, it's the bare minimum. Each channel reaches a different population, not the same population three times.

Email: the pattern, then the verification

Finding the address is the easy part. Most companies fit five formats: *first.last@*, *f.last@*, *flast@*, *last@*, *first@*. One domain, one rule, and everyone follows it.

Verification is the real subject, and that's where it's decided.

A bounce isn't a lost email. It's your sender reputation dropping, so the *next* emails, to other accounts, land in spam. When your market is 800 accounts and not one more, torching your domain is a mistake with no way back.

And a single verifier is never enough:

• catch-all servers reply 'valid' to any address, including *anything@company.com*
• greylisting makes a perfectly valid address look invalid on the first try
• some servers simply don't answer verification requests

A single verifier will hand you either false positives or false negatives, and you won't know which. Cross-checking several providers is the only way to decide: when five sources say yes and one says no, you know what to do. When you have only one, you know nothing.

The phone: the channel everyone neglects, and it works

The paradox of the trade: it's in industry that the phone works best, and it's in industry that it's used least.

The reason is dumb: your contact is physically on site. Not remote, not in an open space with headphones on. There's a switchboard, and the switchboard picks up.

Two routes:

• The direct mobile, via databases. Partial and very uneven coverage from one provider to the next: each has its sources, its regions, its favored sectors. Here too, cross-checking several providers changes everything, where one finds nothing, another has the number.
• The switchboard, widely underrated. And it doesn't just place the call: it's a free source of information. 'Hello, I'm looking for the maintenance manager, that's Mr. Dubois, right?', you've just confirmed a name, its spelling and a function in thirty seconds.

A practical rule: don't call before you've verified the email. If the domain bounces, it often means you don't have the right spelling of the name, and you were about to mispronounce it.

The sequence

The order isn't cosmetic:

• LinkedIn invitation, no message. It asks for nothing, your name exists. It runs in the background.
• Email at D+2, with the signal and the inference. It does the work.
• Phone at D+5 if there's no reply, and only if the email didn't bounce.
• LinkedIn message the day the invitation is accepted, wherever you are in the rest of the sequence.

That last point breaks the sequence logic, and it's deliberate. An acceptance doesn't wait for your calendar: it arrives when the person logged in, and that's the day to write, not at D+7 as planned on paper.

For the rest, the phone isn't a fallback. In industry it's almost always what closes: the email made the subject known, the call gets the date.

The universal reflex: *'I saw you were hiring a Maintenance Manager, I'm reaching out...'*

That's not a hook. It's proof you set up an alert.

What makes the difference is the inference, not the observation. The signal is what you saw. The hook is what you deduce from it, and what your contact may not have formulated yet themselves.

The structure that works, in this order:

• The fact - one line, no flattery, no 'congratulations'.
• The likely consequence in his day-to-day - this is where all the value of the message sits.
• What you've seen elsewhere - a comparable reference, not a product pitch.
• A small, dated ask - twenty minutes next week, not 'a chat'.

And one simple rule: if you can't write the second line, you don't have a signal.

Four bricks. None requires a tool.

Define your contexts

Not your ICP - your contexts. The difference: the ICP describes an account, the context describes a situation. 'Food SME of 100 to 500 employees' is an ICP. 'Food SME that just created a quality role' is a context.

Write five to ten, always in the form: *'when X happens, then they need Y within Z months'.* If you can't write five, the problem isn't your monitoring, it's your positioning.

Wire up the sources

By decreasing yield:

• Job postings - by far the best signal-to-noise ratio, and the sources are public
• Trade and regional press - that's where industrial investments are announced
• The accounts' own news pages - under-read, because tedious
• Legal notices - disposals, capital operations
• Public grant and subsidy databases
• LinkedIn, for role moves

Most are free. None reads in five minutes.

Filter

This is where it all breaks. A watch that surfaces everything is a watch you drop after three weeks. For each signal, three questions, three binary answers:

• Is it in one of my contexts?
• Is it in the window?
• Is there someone to talk to, with a mandate - and do I have a way to reach them?

Whatever doesn't pass all three doesn't surface. A 'maybe' is a no.

Close the loop

Note what you handled and what it produced. Without a feedback loop, you'll never learn that your 'new site' signals systematically arrive six months too late. With it, you know in a month, and you stop working them.

• Real-time alerts - 90% noise, switched off in three weeks.
• Mass scraping - you rebuild the volume logic you were precisely trying to escape.
• The cosmetic signal - 'congrats on your round', spotted by everyone the same day, including your competitors.
• Intent data alone - in industry, your buyer doesn't search online. He calls his network.
• A watch with no filter - abandonment guaranteed. Filtering isn't a comfort option, it's the brick that decides everything.

None of the above requires software. A methodical BD, with a spreadsheet, five well-written contexts and thirty minutes on Monday morning, will beat the vast majority of his market. Method is the real lever, not the tool.

The problem isn't the method. It's holding it. The sources move, the windows close while you're in a meeting, and filtering - the brick that decides everything - is reading work, not a rule you program once.