At one multinational manufacturing group we have observed up close, every team in every department had an OD function, short for organisational development. Not just at headquarters. The plant floor had OD teams. The maintenance shop had OD teams. So did the warehouse, the office, the laboratory.
The premise was straightforward: company goals are not delivered from the top down. They are delivered by everyone, at every level, contributing what their workplace can. To make that work, every team needed a structured way to surface improvements, evaluate them, and act on them.
The mechanism was an improvement proposal. Anyone in the team could file one. The team's OD lead (often a respected colleague rather than a manager) would help quantify it. If the proposal passed muster, it got funded. If it delivered, the next round started. The cycle ran annually, and at the end of each cycle teams presented their deliveries to the wider organisation at gatherings designed to make the work visible.
That pattern, written down in an easy-to-learn process and lived in by every team, did something that most innovation programmes fail to do: it made ROI calculation a frontline skill, not a leadership skill.
The model is associated with large manufacturers in the public imagination, but the underlying mechanic carries to any size of organisation. A twelve-person SMB needs the same skill more, not less, because every euro of operational waste lands proportionally larger on a smaller revenue base, and there is no central analyst team to bypass.
Why first-line ROI matters more than executive ROI
The traditional org-chart view is that ROI is calculated by finance, presented by leadership, and acted on by everyone else. That model has a hidden cost: every improvement opportunity that requires translation through finance to be evaluated never gets translated. The ones that get through are the loud ones, the ones with executive sponsors, or the ones large enough to justify the consulting cost of a business case.
Everything else dies in the gap.
The operator at the loading dock watches a forklift make the same wasted detour twenty times a day. They know the fix. They know roughly what it costs. They do not know how to phrase it as a business case, and the people who can phrase it as a business case never see what the operator sees.
The manufacturer's answer was not to bring more analysts to the loading dock. It was to give the operator, and their OD lead, the skill and coaching to do the calculation themselves. The numbers do not need to be perfect. They need to be defensible enough to land funding for a proposal that, in absolute terms, would never make a CFO's quarterly review.
What the calculation actually contained
The proposals had a shape. OD leads learned to fill in:
- Current cost of the existing way of working. Hours, materials, defects, downtime, customer complaints, anything quantifiable.
- Cost after change. Same buckets, with the proposed change applied.
- Investment required to make the change. One-off costs, tools, training, downtime to install.
- Soft factors that mattered. EHS (environment, health, safety) impacts had their own multiplier. A change that prevented a near-miss had a defensible additional weighting in the formula, even if the financial saving was modest. That meant safety improvements competed for funding on equal terms with cost-saving improvements, instead of waiting for an audit to force the issue.
- Payback period. Time until the investment was returned.
The format was the same whether the proposal was for a new label printer (€800 investment, paid back in three months on label-stock waste) or a redesigned night-shift handover process (€0 investment, fifteen minutes recovered per shift, EHS factor for fewer mistakes during early-morning fatigue windows).
A finance professional would call this directional. That is exactly the point. It was not meant to replace the audited business case for a multi-million-euro production-line investment. It was meant to make the €800 label printer reachable.
The compounding economics of small wins
A €5,000 annual saving on one workstation looks negligible against a multinational manufacturer's revenue. Multiply that workstation by 800 across the company and run it for three years and you are inside the territory where finance starts paying attention. For a smaller business, the same €5,000 saving on a 30-person operation often clears the bar on day one.
But you only get there if those workstations (whether 800, 30, or 5) all submit, evaluate, and ship their own ROI-quantified improvements in parallel. That is the structural reason the OD model worked. It parallelised improvement work the same way modern engineering organisations parallelise feature development. No central bottleneck. No queue at the CFO's office. Every team a unit of agency.
On incentives: what the research actually says
A common question about programmes like this is whether contributors should be paid extra for the savings they generate. Some employers attach a small bonus to realised improvement value. The honest answer from behavioural research is that incentives are not simple.
Two findings carry weight. First, self-determination theory (Deci and Ryan, with decades of replicated work since 1985) shows that extrinsic rewards can erode intrinsic motivation when the work is already meaningful to the person doing it. The risk is that an improvement bonus reframes contributing as a transaction, which then requires payment to continue. Second, hedonic adaptation is well established in behavioural science: a recurring reward that excited people in year one becomes baseline expectation by year three, and the motivational lift fades.
Neither of those findings says incentives never work. They say the design matters. Public recognition often outperforms cash for ongoing programmes. Annual gatherings where teams present their deliveries, of the kind the manufacturer ran, attach status to the work without setting up the per-proposal bonus dynamic. Peer acknowledgement, visible results, line managers actually responding to proposals, those are incentives too, just not financial ones.
The cleanest framing we have seen work: contribution to operational improvement is part of the job, the calculation skill is part of the toolkit, and recognition is structural rather than transactional. That puts the load-bearing weight on skill and agency, with recognition reinforcing the behaviour rather than driving it.
ROI calculation as innovation thinking
If this sounds like proof-of-concept methodology dressed up in shop-floor clothes, that is because it is the same skill in different packaging.
A proof of concept exists to answer a question: would building this be worth it? The answer requires:
- A model of the current state
- A hypothesis about the future state
- A cost estimate for the change
- A measurement plan for whether the change actually delivered
That is an improvement proposal template. It is also what every PoC charter we have written for clients contains. The structure is identical because the underlying question is identical: is this small investment likely to compound into a large saving, and how would we know?
Innovation programmes that treat ROI as a separate, after-the-fact discipline keep a highly useful part of innovation thinking out of the hands of the people most likely to spot opportunities. Innovation programmes that bake ROI calculation into the everyday work of every team end up with hundreds of small validated wins instead of a handful of celebrated megaprojects.
What this means for organisations that do not have it
Most organisations do not have an OD model at every level and will not get one by reading an article. The transferable patterns are smaller than that.
Make ROI estimation a teachable skill, not an inherited one. A two-hour workshop covering hours-saved, cost-avoided, payback-period, and how to handle EHS multipliers is enough to take a team lead from "I have an idea" to "I have a proposal".
Lower the threshold for what counts as a fundable proposal. If your minimum business-case size is €100,000 of impact, you have already excluded almost every front-line improvement worth doing. Pick a threshold that lets the €800 label printer through.
Tie the formula to outcomes other than cash. The EHS multiplier is the underrated piece. It signals to the workforce that safety, environmental impact, and quality of work life are not parallel programmes competing with operational improvements; they are first-class benefits that pay for themselves in the same currency.
Resist the temptation to centralise. A central improvement office is the wrong shape. The reason 800 workstations can each file, evaluate, and ship a proposal is that no central body needs to read all 800 proposals. The line manager and the team are the audience. Aggregation happens after, for reporting, not before, for approval.
For SMBs, the version is simpler still. The "OD model" might be a single shared document, a once-a-quarter team meeting where someone presents a calculated proposal, and a leader who actually listens and acts. The maths is the same. The institutional ceremony is not needed at small scale, and the ROI skill compounds faster because the line of sight from operator to outcome is shorter.
Where to start
Pick one team. Pick one obvious improvement. Spend an hour with the team's lead walking through the calculation: what is it costing now, what would it cost after the change, what is the investment, when does it pay back. Make the calculation visible to the team. If it lands, fund it. If it delivers, repeat.
The hard part is not the maths. The maths is week-one of any finance training. The hard part is getting the calculation into the hands of the people who actually see the opportunities, trusting their numbers, then giving them the ongoing coaching to get sharper at it. The manufacturer's OD programme bundled leadership training and coaching support around the calculation skill, so OD leads and their teams felt continuously supported rather than parachuted into a process they were expected to figure out alone. The same applies at SMB scale, where the coaching might be a fortnightly thirty-minute call rather than a formal training programme, but the principle is identical: nobody should be left alone with a new skill.
The manufacturer was not unusual in having sharp operational people. Every manufacturing company has those, every SMB has those, every services firm has those. It was unusual in having drilled the ROI skill down to where those people work, and then keeping it sharpened. That is the move worth copying.