The other side of growth
The other side of Lesson 18's first lever: EBITDA grows not only from revenue, but from cost. Cost reduction removes spend that isn't earning its keep. Automation replaces manual, repetitive work with software, support ticket triage, invoice processing, freeing people for higher-value work rather than simply cutting headcount (Lesson 34 returns to this distinction directly). Procurement renegotiates what a company buys, cloud hosting, software licenses, vendor contracts, which is often the fastest, lowest-risk EBITDA lever available, because it requires no change to the product or the customer relationship at all. Operational efficiency covers everything else: fewer handoffs, less duplicated work, tighter processes.
Procurement and automation both tend to be higher-confidence, faster-time-to-value initiatives by Lesson 27's own framework, worth remembering when prioritizing against flashier but slower revenue initiatives.
| Lever | Confidence | Speed |
|---|---|---|
| Procurement | High | Fast |
| Automation | Medium-high | Fast |
| Cost reduction | Medium | Medium |
| Operational efficiency | Medium | Slow |
Confidence and speed are useful for sequencing, but a plan still needs to add up to a number. Put a £ figure on each of the same four levers:
| Lever | Illustrative 5-year EBITDA impact |
|---|---|
| Automation | +£1.2m |
| Cost reduction | +£1.0m |
| Procurement | +£0.8m |
| Operational efficiency | +£0.5m |
| Cost-side subtotal | +£3.5m |
| Entry EBITDA + both sides (Lesson 28's £12.0m + this £3.5m) | £20m + £15.5m ≈ £35.5m |
That £35.5m lands inside the £35-40m range Lesson 5's thesis committed to, which is exactly what "testable" was supposed to mean: a plan a reader could add up and check.
Checkpoint
- Four cost-side levers: cost reduction, automation, procurement, operational efficiency.
- Procurement in particular: is often high-confidence and fast, since it touches vendor contracts, not the product or customers.
If anything here still feels unclear, ask before moving to Lesson 30.