Case study

Marketplace Pricing & Demand Forecasting

Turning noisy demand and capacity signals into actionable pricing decisions — before operational pressure becomes obvious.

The situation

A logistics marketplace needed to make pricing decisions before capacity problems became visible operationally. Late, manual intervention meant pressure could become obvious only after much of the opportunity to respond had passed.

A demand forecast alone would not solve that problem. The useful questions were when the business should change price, by how much, how confident it should be, when it should leave price alone, and how the decision could be validated safely.

A decision system, not a model hand-off

The work combined demand forecasting with live supply and capacity signals to identify periods where the marketplace was likely to come under pressure. Price sensitivity and commercial scenario modelling then helped distinguish between a forecast worth monitoring and one that justified action.

Operational guardrails mattered as much as the model. A pricing response that improved short-term revenue but damaged conversion, customer experience or supply behaviour would be a poor commercial decision. The workflow therefore connected four stages:

Scenario modelling compared holding price with progressively stronger interventions under the same demand and capacity assumptions. That made the trade-off explicit: a larger response could create more commercial upside, but only by accepting more conversion and customer risk. Recommendations could therefore be calibrated to the strength of the pressure signal instead of applying one rule everywhere.

  1. Diagnose
  2. Forecast
  3. Simulate
  4. Validate

Make the intervention window visible

The executive view showed whether the cumulative booking trajectory was likely to exceed usable capacity by the service date. That forecast created time to intervene, regulate subsequent bookings, and still allow enough gross demand to fill capacity after expected attrition.

Forecast demand early, regulate bookings toward capacity

Illustrative data
Cumulative bookings forecast and pricing intervention Illustrative cumulative bookings from 90 days before service to the service date. At 35 days, the current trajectory indicates demand will exceed 100 jobs of usable capacity. A recommended illustrative price increase of 12% slows subsequent booking growth. Projected gross bookings finish at approximately 105 jobs so expected cancellations and rescheduling bring fulfilled jobs back toward usable capacity. 100 90 60 30 0 Cumulative jobs / bookings Usable capacity — 100 jobs Pricing intervention Recommended price: +12% Expected cancellations & rescheduling Gross-booking buffer Days before service
Illustrative data. Pricing is used to slow booking growth when forecast demand exceeds capacity. Gross bookings deliberately finish slightly above usable capacity to allow for expected cancellations and rescheduling.

The recommendation

The answer was not a blanket surcharge. Pricing intervention should be targeted to periods where forecast demand and available capacity justify it, start conservatively, and be validated through controlled experimentation. When the signal was weak or guardrails were at risk, the right action was to hold price.

That framing gave commercial and operational teams a shared decision: not “is the forecast accurate?” but “is there enough evidence and enough upside to intervene now?”

Outcome

The work contributed to reducing blocked days from 21 to 1 and to an approximately 10% revenue premium from the resulting surge-pricing approach.

It demonstrates the full path from ambiguous commercial problem, through forecasting and scenario reasoning, to a recommendation the business could act on and test safely.