"Can you put together a business case for the learning investment?"
Every L&D leader has been asked that question, and most of us have handed back a document that opened with a quote from a McKinsey report and a pie chart of training hours delivered.
Three weeks later the request was politely shelved.
There is a better way. The business case that gets funded does three things that training hours never will:
- It names the cost of the status quo in dollars
- It attaches the learning investment to a specific commercial outcome the business already cares about
- It is defensible: the finance team can audit the numbers without calling you
This is the model we use with clients.
Three questions you have to answer
Before you write a number on a page, you need a clear answer to three questions.
1. What problem are we actually solving?
Not "improving capability." That is a means, not an end.
The problem is usually one of these:
- Productivity gap: we are leaving output on the table because people cannot yet do the thing well
- Quality / risk exposure: errors, rework, compliance breaches, customer churn
- Speed to competence: new starters take 9 months when they should take 4
- Retention of key populations: people leave because they do not see a development path
- Strategic delivery risk: a specific strategic bet is at risk because the capability is not there yet
Pick the sharpest version of the problem your sponsor actually feels. That becomes the headline.
2. What does the status quo cost us?
This is the number that makes a CFO lean forward. Not "what will this program cost?" but "what are we already losing every quarter because we have not done this?"
Some examples from real client cases:
- A bank found that underwriting errors from newly-trained staff were costing $3.2M a year. A targeted capability program cost $420k.
- A retailer calculated that an extra 8 weeks to competence on the shop floor cost $11k per new hire in lost productivity. At 400 hires a year, that was $4.4M.
- A health service mapped that 40% of voluntary exits in the first 18 months came from one cohort, nurses who never got the structured skill progression they were promised. Replacement cost: $65k per head.
If you cannot point at the status quo cost, you do not have a case yet. Go back and find it. Finance already tracks most of this somewhere.
3. What will we do differently, and how will we measure it?
Say this out loud: "If we fund this, in 12 months, the following metric will have moved by this amount."
Then pick 2 or 3 leading indicators that tell you early whether you are on track:
- Time-to-competence for cohort X
- First-pass quality rate on task Y
- Voluntary regretted exits in population Z
- Manager confidence scores on capability A
Leading indicators keep you honest for the 8 months before the headline outcome shows up.
The five line items every case needs
Structure the case on a single page. Five sections.
1. The problem in one sentence. Plain English, backed by a number.
"We lose $3.2M a year to underwriting errors from newly-trained staff. The errors cluster around three decision points in the first 6 weeks of the role."
2. The status quo cost. Same period, same units, finance-signed-off.
3. The proposed intervention. What you will do, not how. Three to five bullets, no jargon.
4. The expected return. Named metric, size of the move, timeframe.
"Error rate reduces from 4.1% to under 2.5% within 9 months, avoiding $1.8M of the annual loss."
5. The investment required. Total cost, phased. Include internal time.
Keep the whole thing on one page. If a 30-page version is needed, it is an appendix.
The pitfalls that kill credibility
Counting training hours as a benefit. Nobody cares that you delivered 4,200 hours of training. They care what moved.
Using industry benchmarks as proof. "Top-performing organisations spend 2.5% of payroll on learning" is interesting. It is not a business case for your organisation.
Claiming ROI that is too big. A 10x return on a capability program is possible, but a case built on it reads as fiction. Be conservative. Beat expectations in delivery.
Ignoring the manager layer. Most learning interventions that do not include the manager fail. Bake manager time into the cost and the design.
A note on measurement
You do not have to prove causation to a peer-reviewed standard. You have to be defensible enough that a reasonable finance partner accepts the logic.
Three things help:
- Run a comparable baseline. What are the current numbers for this population today?
- Protect a control. If you can leave one cohort outside the intervention for 3 to 6 months, you get a natural comparison.
- Report leading indicators monthly. You will see drift early, and you will be in the conversation instead of defending a year-end number.
A template that does most of the math for you
We built a L&D Business Case Calculator that implements this model directly, inputs for cohort size, status quo cost, intervention cost, and expected uplift. It spits out a one-page business case you can take to the CFO.
The same sheet also includes the leading-indicator dashboard we use with clients. Grab it, customise the numbers, and put a real case in front of your exec team.
Welcome to better.