Skip to content

Jonbar

Launches · Roadmap

A product with no history still has a probable range.

On launch day your data says nothing about a new product, so Jonbar starts from a wide range and every week of real sales narrows it.

Roadmap

Not in the pilot yet. Today the pilot rehearses price and discount versions of products you already sell.

Example run · illustrative numbers

Week 1 to week 8, as the range closes

No sales yet80 − 685
After 4 weeks325 − 475

Same product, same engine. Only the real data behind the line changed.

The situation

A new flavour nobody has ever bought.

A price and a first production run, decided with zero history.

The product
A new salted caramel protein powder.
The decisions
Launch price, quantity, opening discount, first channel.
The data
Two years on vanilla and chocolate only.
The deadline
Production is ordered before the first sale.

Where the first answer comes from

Borrowed behaviour, named out loud.

Three sources, each weaker than your own measured sales.

  1. 01

    Your nearest products

    Their sales history lends the new one a shape.

  2. 02

    Published elasticity ranges

    Why the first answer is a range.

  3. 03

    Your own costs

    Unit cost, commission and shipping make it yours.

Across 1,851 published estimates, the average price elasticity is −2.62. Bijmolt, van Heerde & Pieters, Journal of Marketing Research

The range, week by week

Wide on day one, narrower every week after.

Weekly units for the new flavour. Step through the weeks and watch the range close.

Example run · illustrative numbers
Week 8 range80 − 685units per week
Range width605units per week
A fan chart of weekly units for a new flavour over eight weeks. With no sales data the probable range for week 8 is very wide; after each week of real sales the measured line extends and the forward range narrows, until after six weeks it is a fraction of its opening width.
Weeks of real sales

Nothing about the product changed. The engine stopped borrowing and started measuring.

How we keep ourselves honest

The forecast is locked before the launch, then scored against it.

Jonbar writes the range down before launch and compares it with what happens, week by week.

Details: inputs, outputs and limits

What the engine needs

  • Product card: category, price band, pack size, unit cost.
  • Two years of sales on your nearest products.
  • Lead time, minimum order, stock and floor price.
  • Commission, shipping and launch ad budget per channel.

What comes back

  • Weekly units and gross profit as a probable range.
  • A launch price with its trade-off written next to it.
  • A first production quantity that holds at both ends.
  • The band the first four weeks should land in.

What it cannot do

  • A genuinely new category has nothing to borrow from, so the range stays wide.
  • Early weeks are shaped by stock and ranking as much as by demand.
  • Returns and ad cost are not modelled.
  • Coming later: an automatic weekly re-fit and an early warning. Today the range is re-run when you send new weeks.
  • No accuracy figure until the backtest is published.
  • You set the launch price and the production order.

Launch with a range you wrote down first.