Judgmental forecasting
Definition
Judgmental forecasting
Judgmental forecasting predicts outcomes through expert opinion and structured intuition instead of statistical models. Analysts turn to it when data is thin — a new market, a fresh product, or conditions that shifted too sharply for old numbers to still apply.
Key takeaways
- Judgmental forecasting relies on expert opinion when hard data is missing, dated, or unusable.
- Common methods include Delphi rounds, analogy, scenario building, Cooke’s method, and structured surveys.
- Bias controls such as anonymity, calibration, and blended models protect the estimate from drift.
- It suits new-product launches, long-horizon strategy, and shocks that break historical patterns.
- Pair it with statistical forecasting whenever any usable numeric history exists.
Managers and forecasters turn to the approach when the model refuses to run — a new offering has no prior sales, a regulator shifts the rules overnight, or a shock rewrites customer behaviour. Structured methods turn expert intuition into a defensible estimate.
The approach draws on experience, market signals, and disciplined reasoning. It fits pharmaceutical launches, geopolitical planning, and complex operations where a spreadsheet alone would mislead the decision.
How it works
Judgmental forecasting collects expert input in a controlled way, corrects for known biases, and translates opinion into a documented estimate. The steps borrow from social science and decision analysis rather than pure statistics.
Analysts start by framing the question, then choose a method that matches the horizon and data gap. A short-cycle demand call may need only an analogy comparison against a similar past product.
A ten-year strategy call usually needs scenario planning or a full Delphi method round with a mixed panel. Panels stay anonymous so no single voice dominates the answer.
Facilitators then aggregate the responses through medians, weighted averages, or Cooke’s calibration scoring — the method choice depending on how well the panellists have been calibrated on past questions.
Every assumption gets documented against the guidance in Forecasting: Principles and Practice. Post-hoc scoring keeps the panel honest for next time.
The main methods sit in one table:
| Method | Best for | Inputs | Output |
|---|---|---|---|
| Analogy forecast | New product or campaign similar to a past one | Historical performance of the comparable case | Range prediction |
| Delphi method | Long-horizon, high-uncertainty calls | Anonymous expert rounds with feedback | Converged consensus |
| Composite | Complex problems needing more than one lens | Two or more methods combined | Blended estimate |
| Cooke’s method | Risk and probability calls needing calibration | Expert answers to seeded questions | Weighted forecast |
| Scenario building | Strategy under deep uncertainty | Plausible future “worlds” | 3–5 narrative scenarios |
| Statistical surveys | Customer demand, sentiment, intent | Structured polling of buyers or staff | Probability estimate |
Examples
Judgmental forecasting shows its value when history is silent or misleading. Three well-documented cases show how experts turned qualitative reasoning into decisions with real money on the line and clear post-hoc validation.
Apple iPhone forecasting (2014–2020). Apple could not regress the iPhone 6 Plus at launch because no earlier phablet lived in its catalogue. The team compared the design brief to the iPad mini and Samsung’s Galaxy Note line before setting inventory.
The iPhone 12 mini in 2020 repeated the exercise against the iPhone SE. Judgmental analogy filled the estimate gap that no time series could have produced on its own.
Pfizer and BioNTech COVID-19 vaccine (2020). Both firms scaled production capacity before Phase 3 trial data existed.
Executives, epidemiologists, and government forecasters weighed severity signals, viral-spread models, and public-health need against a base case of zero demand.
The resulting scenario forecast let manufacturing lines start months before the statistical signal arrived. Judgment carried the call the numbers could not yet defend.
Royal Dutch Shell scenario planning (1973). Pierre Wack pioneered the technique at Royal Dutch Shell in the late 1960s, then used it to war-game an oil-price shock that mainstream analysts dismissed.
The Delphi method was developed at RAND Corporation around the same era for exactly this class of long-horizon question. Both traditions still shape strategic planning today.
When OPEC embargoed exports in October 1973, Shell had already rehearsed the response. The firm climbed from the seventh-largest oil major to the second by the late 1970s, and the scenario discipline became a case study across strategic forecasting practice.
Related terms
Judgmental forecasting sits inside a family of estimation and planning terms. Each connects to a specific piece of the workflow, from raw data pipelines to workforce coverage against the forecast.
- Forecasting: the parent discipline covering both quantitative and qualitative methods.
- Delphi method: an anonymous, multi-round expert consensus technique developed at RAND Corporation.
- Scenario planning: a structured way to explore several plausible futures side by side.
- Predictive analytics: statistical and machine-learning approaches used when a strong data trail exists.
- Business intelligence: the reporting and dashboarding layer that surfaces the signals feeding a forecast.
- Workforce planning: headcount forecasting where judgmental input covers attrition and growth surprises.
- Capacity planning: matches production or service supply against a forecast of expected demand.
FAQ
Common questions on judgmental forecasting cover method choice, bias control, reliability for board decisions, and how the approach sits alongside statistical forecasting and modern AI tools.
How is judgmental forecasting different from statistical forecasting?
Statistical forecasting projects the numeric past forward. Judgmental forecasting brings in expert opinion when the past is thin or broken. Most mature teams run both and blend the outputs.
When should you use judgmental forecasting?
Reach for it during product launches, geopolitical shocks, regulatory shifts, and any decision with under three years of usable history. It also helps as a sanity check against a purely statistical model.
What are the main biases to control?
Anchoring, groupthink, overconfidence, and recency bias distort every panel. Structured methods such as anonymous Delphi rounds and Cooke’s calibration scoring keep the estimate honest and auditable.
Is judgmental forecasting reliable enough for board decisions?
Yes, when it uses documented method selection, transparent assumptions, and blended checks against any available data. Boards typically want a scenario band rather than a single point estimate.
How does it fit alongside AI and predictive analytics?
AI handles pattern-heavy problems where the data trail is long and clean; judgmental forecasting handles the shocks, new categories, and strategy calls it still struggles with. Pair them for coverage across the whole horizon.
Explore more OA terms and guidance at Outsource Accelerator.
Independent
