.5f521cec.png "Trivas.ai")
See where the year lands, and test what an ad-spend change would do.
AI Forecasting projects revenue and units for the year against your target and last year. Actuals to date are joined to the forecast for the rest of the year, and an accuracy table shows how each model has performed. A simulation wizard lets you plan ad spend by month and channel, then compare the results.
The window is the current calendar year, compared with the previous year.
Targets are set once and checked too late.
Teams set a revenue target in January and find out in the autumn whether they are on course. Forecasts in a spreadsheet go stale, and nobody can say how reliable they were.
No live view of the gap to target
Actuals sit in one report and the target in another, so the gap is worked out by hand, if at all.
Forecasts without a track record
A number is only useful if you know how close past forecasts came. Few teams keep that score.
Budget questions are guesses
What happens if we spend more on Meta and Amazon in Q4? Answering means building a new model each time.
Inside AI Forecasting
Each tab describes one part of the page, with its elements numbered.
Tabs describe one part of the page each, with the parts numbered.
Headline cards and charts show projected revenue and units, split between what has happened and what is forecast.
Forecast: the year against target
Headline cards and charts show projected revenue and units, split between what has happened and what is forecast.
- 01
Projected Revenue
Actuals to date plus the forecast for the rest of the year, shown against your target with a badge for the difference.
- 02
Projected Units
The same actuals plus forecast view for units sold.
- 03
Revenue Forecast charts
Weekly or monthly, and cumulative, with a toggle between Revenue and Units. The legend shows Actual, Forecast, Target and Previous Period.
- 04
Forecast Starts marker
A vertical line on the chart marks the date where actuals end and the forecast begins.
- 05
Filter bar
Choose the platform, the model and the forecast timestamp, and group by Weekly or Monthly. The latest model and timestamp are selected for you until you pick another.
- 06
Model fixed by category
When you filter to specific categories or sub-categories, the model is fixed for that selection. Select all to choose a different one.
What Forecast vs Target does
Actuals plus forecast
For each period up to yesterday Trivas uses what actually happened. After that it uses the forecast, so the projection for the year is always current.
Gap to target, shown plainly
Projected revenue appears next to your target with the difference as a percentage of that target.
Against last year
Charts overlay the previous period so you can see the year's shape against the last one.
A track record for each model
The Model Accuracy Table gives sales and quantity accuracy by month, so you can trust or challenge a forecast before you present it.
Platform split
Switch between Platform Overview, Shopify Detail and Amazon Detail to see where the forecast comes from.
Ad-spend simulations
Enter planned spend by month and channel, run it, and see projected revenue and units without exporting data.
Compare simulations
Put several completed simulations side by side against the baseline to choose a budget path.
Weekly or monthly view
Group the charts and tables by week or by month, and filter by platform.
Who uses it, and for what
Founder or CEO
Checks projected revenue against target and the delta badge to answer one question: are we on track for the year?
Finance lead
Reads the monthly forecast table by platform and the accuracy of past months before presenting numbers to the board.
Growth marketer
Creates a simulation with higher Meta and Amazon spend for October to December and compares it with the baseline.
Ecommerce manager
Compares Shopify Detail with Amazon Detail to see which platform carries the gap to target.
What changes for your team
Planning from a refreshed model
The forecast is regenerated by the model and not rebuilt by hand, so the plan reflects the latest actuals.
Numbers you can defend
An accuracy table beside the forecast lets you show how reliable the numbers have been, not just what they say.
What-if budget testing
Test a spend plan by month and channel inside the same workspace, without exporting data into a separate model.
What it needs, and what it doesn't do
What it needs
- A forecast model set up for your brand. Book a demo and we will confirm what your account includes.
- Connected sales data for the platforms you want to forecast.
- A revenue target, to see the gap to target.
- For simulations, a connected Shopify or Amazon store, and the ad channels you plan to spend on.
What it doesn't do
- Forecasts are estimates, not guarantees. Actual results may vary.
- It reads data and plans scenarios. It doesn't change anything on Amazon or in your ad accounts, and a simulation doesn't spend any budget.
- Simulations in the wizard cover Amazon and Shopify, driven by planned ad spend. They are not promotion or launch simulations.
Frequently asked questions
See where your year lands.
Book a short walkthrough and we will show Forecast vs Target on your data.