Fratz CC Tools

Fratz WFM or All Access

Volume forecast

Preview

Forecast the next weeks of interval volumes from your history, see how accurate the method would have been on your own recent weeks, and send any day to the Day Planner. Your data stays in this browser.

History

The preview uses 16 weeks of sample history. Unlock to use your own data.

Needs a date, an interval start (or one date-time column) and offered volume. For messy exports, use Import first and tick “Remember in this browser”.

Forecast

One date per line (YYYY-MM-DD): holidays, outages, one-off campaigns.

Date and a factor per line: 0 = closed, 50% or 0.5 = half volume, 120% = 20% busier.

Fratz WFM or All Access

You're looking at a forecast of sample data

With the Fratz WFM or All Access tier, forecast your own imported or uploaded history, export every interval to CSV and send any forecast day straight to the Day Planner.

Join on Patreon

Already on the Fratz WFM or All Access tier? Log in with Patreon and this tool unlocks in this browser.

Forecast

–

Next 7 days

–

forecast contacts

Backtest accuracy

–

interval WAPE

Daily MAPE

–

bias –

Trend

–

per week

–

    One day, by interval

    –

    Show the intervals

    Backtest: how it did on your recent days

    –

    How the forecast works, in plain English
    1. Daily totals. We add up each day in your history. Days with no data count as closed.
    2. Weekday pattern. We compare each weekday's average with the overall average. If Mondays run 15% busier than an average day, every forecast Monday gets +15%.
    3. Level and trend. With the weekday effect taken out, we draw the best straight line through the daily totals. That line gives the level and the weekly growth or decline. Untick “Follow the trend” to use the average of the last 4 weeks instead.
    4. Intraday shape. For each weekday we average what share of the day arrived in each interval over the last few weeks.
    5. Forecast. Level × weekday factor × interval share, times any adjustment you entered.

    Accuracy is measured honestly: we pretend the last days of your history haven't happened, forecast them, and compare with what really came in. WAPE is total interval error ÷ total volume (10% or less is very good for interval-level volume, 20% is good). MAPE is the average daily error. Bias shows whether we ran high (+) or low (−). We also show a naive “same day last week” forecast so you can see whether the method beats the simplest guess.