Monthly Reporting for an Industrial SME, Without Manual Consolidation
In this industrial SME, monthly reporting required six days of manual consolidation before management could read it. We made the process reliable first, then automated what could be, using the tools already in place.
The company
An industrial SME of about 150 people, whose monthly reporting is consolidated by hand from several files and system extracts.
The problem and its cost
Each month, the finance team gathers production, sales, inventory and accounting data from separate files and systems. Consolidation takes about six working days. Figures arrive late in the month, two versions of the same indicator often circulate, and management decides on a picture that is already out of date.
The cost is direct and can be calculated: 6 days × 12 months × 8 h × CHF 95, or about CHF 55,000 per year of skilled time spent copying, reconciling and checking figures. Two less visible costs come on top: the risk of error at each re-keying, and the time the team cannot spend analysing variances.
The strategy
We started by working with management on scope: which indicators really drive the business, how each one is defined, and who answers for it. Without that agreement, automation would only have produced questionable figures faster. So we narrowed the scope before touching the process.
Next, we mapped the sources and reconciliation rules, then automated collection and consolidation within the existing tools. Finally, we set up a closing calendar and a monthly management review. The order is deliberate: define, simplify, automate, then steer. Reversing these steps would have weakened the result.
The execution
Team: We mobilised a dedicated contact, the WALVIS core team in finance, operations and data, and a sector expert with the profile of a management controller from process manufacturing.
- →A single indicator dictionary, stating for each figure its definition, its source and its owner.
- →An automated collection and consolidation flow, fed by ERP exports and existing files, with no new software.
- →Consistency checks that flag variances before the review, so that time goes into understanding them.
- →A monthly dashboard in the reporting tool already in use, with a closing calendar and a management review format.
We used AI in two specific places: classifying entries whose allocation was not explicit, and proposing a first draft of variance commentary. In both cases the finance team validates, and the consolidation itself relies on deterministic rules.
Most of the gain comes from organisational choices: shared definitions, a calendar that is respected and one owner per figure. These decisions require no technology, yet they determine how reliable everything automated afterwards will be.
The result
Expected annual saving: 75% of CHF 55,000, about CHF 41,000. For an investment of about CHF 35,000, payback comes in 10 months.
The gain is not 100% because some work remains necessary: reviewing variances, validating figures, handling exceptional cases and writing the commentary for management.
- Figures available earlier in the month, so decisions rest on a more recent picture.
- One version of each indicator, which reduces debate about data reliability.
- A finance team that spends more time on analysis and less on data entry.
The assumptions
- 6 consolidation days per month, 12 months per year, 8 h per day.
- Loaded hourly cost of CHF 95, giving about CHF 55,000 per year.
- 75% reduction in consolidation time, or about CHF 41,000 per year.
- Investment of about CHF 35,000, giving payback in around 10 months.
Scenario type, built from real missions and anonymized.
In the same cocoon
