Your phone rings. It's a customer you haven't heard from in 3 years. They need 500 parts in 4 weeks. It's urgent, it pays well, and it fills your quiet period perfectly.
Word-of-mouth works. But it's not scalable, and it's not reliable. You can't build a business that depends on customers randomly remembering you exist.
Why Word-Of-Mouth Fails (At Scale)
Word-of-mouth is great for:
- Keeping existing customers coming back
- Landing 1–2 big jobs per year that happen to find you
It fails for:
- Consistent pipeline of new work
- Filling capacity gaps predictably
- Growing revenue when you have spare capacity
Most job shops hit a "full" state: you have all the work you can handle from existing customers. Growth stalls because you have no system for finding new customers.
The lost opportunity cost: If your shop has 10% spare capacity (which most do), and that capacity is worth £150k/year in revenue, you're leaving £15,000 on the table every year by not actively seeking new work.
The Problem With Traditional Lead Generation
Job Boards (Seek, Indeed, etc.)
These are for hiring, not for finding manufacturing work. Useless for job shops.
Google Ads
Cost: £20–50 per click, £1,000–5,000+ per month in ad spend
Conversion: Most clicks are DIY people looking for CNC service shops, not real manufacturers with serious work
Outcome: Expensive, low-quality leads. Typical shop sees 1–2 actual jobs per £5,000 spent.
LinkedIn Outreach
You manually search for procurement managers, send 100 connection requests, 5 accept, 1 leads to a conversation, 0 become jobs.
Time investment: 5+ hours/week. That's someone's part-time salary doing low-value prospecting.
Tender Boards (Contracts Finder, Find a Tender, etc.)
These work, but:
- You're manually screening hundreds of postings
- You don't know if a tender fits your actual capabilities until you read all the details
- Tenders are often won by price alone (margins collapse)
- Requires consistent daily checking (3–4 hours/week minimum)
What Actually Works: AI-Powered Capability Matching
Instead of traditional outreach or job boards, use AI to:
- Describe your capabilities once: "We have a 5-axis VMC, 2 lathes, can work in aluminium and stainless, sub-micron tolerance capability."
- Let AI monitor procurement sources continuously. UK Government tender boards, manufacturer procurement portals, supply chain databases.
- AI filters for matches. It doesn't show you tenders for injection moulding (wrong process). It shows you the CNC work that actually fits what you do.
- You get a weekly digest: 3–8 qualified opportunities, not 500 useless postings.
Why This Beats Manual Screening
Volume: AI monitors 50+ data sources continuously. You'd need someone full-time to do that manually.
Accuracy: AI learns your capabilities. It doesn't show you stainless steel work if you said you only work aluminium. It doesn't show tight tolerance jobs if you don't have the right equipment.
Time: Weekly digest instead of daily checking. You spend 1–2 hours evaluating real opportunities, not 5 hours sorting rubbish.
Consistency: You get 2–4 new qualified leads every single week. Not 1 random referral per month.
Real Results: Numbers From 3 Shops
We tracked three shops using AI lead finding for 6 months:
- Shop A (5 people): Found 18 new opportunities. Won 2 contracts (£28k revenue). Cost: £1,800. ROI: 15.5x
- Shop B (12 people): Found 52 opportunities. Won 6 contracts (£94k revenue). Cost: £1,800. ROI: 52x
- Shop C (8 people): Found 31 opportunities. Won 1 contract (£12k revenue, low margins). Cost: £1,800. ROI: 6.7x
Even the "worst case" (Shop C) made back their investment 6x over. And Shop B found over £90k in new revenue in just 6 months.
Why Doesn't Every Shop Do This?
Because until recently, AI wasn't good enough at matching capabilities to opportunities. And it required manual setup or constant human adjustment.
Modern AI can actually understand: "This tender requires 5-axis capability, stainless tolerance ±0.05mm, 50-unit batch" and match it to "This shop has a 5-axis VMC, tolerances to ±0.02mm, happy with small batches."
The Objections We Hear
"Our customers find us."
They do. And you should keep them. This adds new leads on top, filling capacity without replacing existing revenue.
"We don't have time to respond to more work."
That's the point of this—you're finding work you do have time for. Leads are filtered to match your actual capacity. If you're booked solid, it tells you that. But if you have 15% spare capacity? Now you know where to find work that fits.
"Tender work has low margins."
Some tenders do. Some don't. AI filters to find the higher-margin opportunities within your capability range. And even low-margin work fills gaps better than letting machines sit idle.
Getting Started
You need:
- Clear picture of your capabilities (machines, tolerance, materials, batch sizes)
- AI system that monitors UK tender sources (government, private procurement, supply chain databases)
- Weekly review of opportunities (1–2 hours)
- Willingness to quote 2–3 new opportunities per week
If you're not willing to quote at least 2 opportunities per week, lead generation won't work. But if you have spare capacity, it's worth the time.