Point the model at your history and set the horizon.
Choose the source, a monthly grain and the date the forecast starts. Units, revenue and cost are summed from your own sales data, so the plan starts from what the business actually did.
Build the revenue and margin plan from your own actuals, forecast demand, compare pricing scenarios against a baseline, then track the plan against actuals each time your data refreshes.
The model reads six months of fittings sales, keeps them as actual periods, and plans the next six. The baseline is today’s price and cost, with nothing changed.
Choose the source, a monthly grain and the date the forecast starts. Units, revenue and cost are summed from your own sales data, so the plan starts from what the business actually did.
Price change, unit cost and price elasticity are inputs with a fallback value. The baseline uses those defaults, so every other case is a change from it.
A demand forecast for each fittings product gives the units the plan works from, with a likely range and a record of how each run scored.
History rows show actual demand. Forecast rows show the expected value and the low and high ends of the range 8 in 10 outcomes should fall inside.
Each run scores several methods on past periods and keeps the best. Version 4 kept gradient boosting, at 7.9% weighted error.
Version v4: what the run used
| Method | WAPE | Used | Weight / note |
|---|---|---|---|
| Seasonal smoothing | 10.8% | No | |
| Seasonal ARIMA | 9.1% | No | |
| Prophet | 11.6% | No | |
| Gradient boosting, per series | 7.9% | Yes | |
| Pooled Gradient Boosting (all series) | 8.6% | No |
The decision model at the top of the page puts three cases on one chart: the baseline, a 12% resin cost rise, and the same rise with a 4% price increase. Here is how each case is defined, and how wide its outcome could be.
A scenario starts from the baseline inputs. One raises unit cost by formula; the other raises it and sets a 4% price change.
Scenarios are named what-if cases that override input defaults.
| Scenario | Assignments | Trials |
|---|---|---|
| Base Baseline | Uses input defaults | 100 |
| Resin cost +12% Scenario | Unit cost: formula | 100 |
| Resin +12%, price +4% Scenario | Price change: 0.04, Unit cost: formula | 100 |
Give an uncertain input a range and each scenario runs many trials. With the increase, March margin lands between about $285K and $349K in 90% of them.
Resin +12%, price +4% · 100 trials. The shaded band spans the 5th–95th percentiles; the line shows the median.
Publish the chosen scenario, keep the forecast version it was built on, and compare each month’s actuals with the plan in a view. Three months in, fittings margin is $14,170 under the $937,510 plan to date (1.5%), and December is the month to look at.
Plan from the chosen scenario, actuals from the sales history, variance as formula columns. December is 3.5% under plan, still inside the range the trials showed. Illustrative data.
Publishing gives the model’s results a read-only page, with the chosen scenario against the baseline, ready to print.
A January refresh trained version 5, which scored worse, so version 4 stays in use. You switch versions only when you choose to.
The capabilities that do the work, what every subscription adds, and where to start.
Related solutions: Price Increase Management, Cost, Tariff & Freight Recovery, Customer & Product Profitability.
No. Revomo holds the commercial plan: revenue, price and margin built from your own sales data, with scenarios. Consolidation, currency conversion and corporate budgeting stay in your FP&A tool or ERP. Revomo reads from and publishes back to the systems that run the business; it does not replace them.
No. A demand forecast tries several methods on each series, scores each on past periods and keeps the best, and each version shows what it tried and its error. In the plan itself, every lever is a named input, and each scenario shows which inputs it changes from the baseline.
The baseline, the forecast and the actuals read the same data in one place. A scenario changes only the inputs that differ from the baseline, forecasts keep their versions, and plan against actual is a view that updates when the data refreshes, rather than a file rebuilt every month.
Not with a separate lock. A decision model’s results stay as calculated until someone calculates it again, and role permissions decide who can. Publishing gives everyone a read-only page of the results. For a fixed copy, download the plan view and keep it as a dataset.
Not on its own. The plan against actual view updates when its data refreshes, and a formatting rule can mark any month under your threshold. Reports and dashboards in Analytics & Insights can show the same view to the people who need it.
Sales history by month, with product, customer and quantity, plus price and cost for revenue and margin. Load it as files or connect a supported source in Data & Context Studio. Which plan, forecasts and scenarios to build first is agreed in the Activation and Success Plan.
Bring this year’s plan and last quarter’s miss. We’ll walk through a baseline, scenarios and plan against actual on sample data.