Data & Analytics

Retail Demand Forecasting with Inventory Recommendations

Forecast product level demand and turn the forecast into reorder recommendations that account for lead time, holding cost and the cost of running out. The forecast is the easy half; converting uncertainty into a decision is the interesting one.

Difficulty

Intermediate

A small team, or one strong student willing to learn something new.
Effort

1 semester, 2 students

Deliverables

5 to ship

3 optional extras

Suggested stack

PythonpandasProphetLightGBMStreamlit
A suggestion, not a requirement. Swap anything for what you already know.

What you should ship

  • Cleaning pipeline handling promotions, stockouts and returns, since a stockout is censored demand rather than zero demand
  • Baseline forecasts including seasonal naive, compared against a machine learning model
  • Forecast intervals rather than point estimates, with calibration checked against actuals
  • Reorder recommendations combining the forecast with lead time, holding cost and stockout cost
  • Backtest reporting forecast error and simulated inventory cost against the actual historical policy

If you have time left

  • Hierarchical forecasting reconciled across product, category and store levels
  • New product forecasting from similar items with no sales history
  • Explicit treatment of promotional uplift and cannibalisation between products

The problem

Retailers over order and hold dead stock, or under order and lose sales. Both are expensive and both come from forecasting by intuition or by last year plus ten percent.

What you build

A cleaning pipeline, forecasts with uncertainty intervals, and a reorder policy that uses those intervals to make a decision.

The data trap that catches everyone

Stockouts. When an item was unavailable, sales were zero and demand was not, and training on that teaches the model to predict low demand for exactly the products that sell out. Detecting and handling censored demand is the single most important step and the one most student projects miss.

Why intervals matter more than the point forecast

The reorder decision depends on the spread, not the middle. If stocking out is expensive and holding is cheap, you order against the upper end of the interval. A point forecast throws away the information the decision actually needs, so evaluate calibration, not just error.

The evaluation that counts

Backtest the whole policy, not just the forecast. Simulate what your recommendations would have ordered and compare total cost against what actually happened. A statement such as fourteen percent lower inventory cost with the same service level is a result; a lower error metric alone is not.

Scope warning

Use a public retail dataset. Forecast and reorder policy only, no supplier integration or warehouse management.

Ideas and guidance, not finished projects

These are project ideas and scoping guidance, published free for students to use as a starting point. I do not build, write, or sell final-year projects, and I do not complete coursework for anyone. Take an idea, make it yours, and build it.

Not sure this one fits you?

I help students pick an idea that matches their skills and their deadline, then scope it down until it is finishable. Guidance only, never done for you.

Get guidance
Work with me

Stuck on something specific?

Writing only gets you so far. If you want an answer to your situation rather than the general case, book a session and we will work through it together. Every session is free; a few slots open each week.

Follow along

New writing, resources and project ideas land here first.