The key idea

A new book without sales history needs a launch assumption, not a fabricated track record. Start with the publisher's confirmed commitments, campaign plans, relevant comparisons and a limited inventory budget. Treat the first production decision as provisional, then update it as actual orders and replenishment performance become observable.

Forecasting guidance discusses the limitations of very short histories. This article proposes a practical launch control rather than a guaranteed model. [1]

Use comparisons carefully

A previous title may help estimate an initial range, but document the similarities and differences. Author audience, price, format, language, market and promotional support all affect comparability. Avoid selecting only the most successful previous launch. Include an ordinary or weaker comparison to expose downside risk. External interest can justify investigation, but attention to an author is not automatically demand for this particular edition through the publisher's fulfillment route.

Make the first feedback loop short and explicit

Choose a review point based on the speed at which meaningful evidence will arrive. Record actual orders, unavailable periods, cancellations and the time required to replenish. A title that sells out quickly produces incomplete evidence about unconstrained demand; a title with no orders while its listing was inactive does not prove lack of interest. The review should distinguish these situations before reducing or expanding the target.

Worked example · illustrative assumptions

A hypothetical publisher has 12 confirmed event orders and an uncertain expectation of another eight retail orders. It already has four usable copies. A decision to print 16 covers the combined 20 under that scenario, but only the 12 are commitments. Show those components separately. A more cautious policy might cover the confirmed demand first and approve a smaller speculative allocation until the retail route is tested.

Put it into practice

  • Separate confirmed commitments from launch assumptions.
  • Choose comparison titles using documented product and market similarities.
  • Set a cash limit and an early review point before the first run.

Watch out for

Do not display a confident percentage simply because a model produced a number. With limited evidence, an explicit provisional assumption is more useful than a precision claim that has not been tested.

How many weeks of history are enough?

There is no universal number. The required evidence depends on how often orders occur, the forecast horizon, changing conditions and the decision being made.

Sources and further reading

  1. Hyndman and Athanasopoulos: Practical forecasting issues. Sparse counts, missing data and practical forecasting limitations.

AI assisted editorial content with original planning examples. Source links support the stated source notes; operational suggestions are guidance rather than guarantees. Provider rules and product coverage can change. Editorial approach · BooksAhead.ai connection status.