How does a decision engine turn a situation into a score? Step by step
The essentials in 30 seconds
- A decision engine turns a situation into a score in six moves, from question to threshold.
- Each move can be inspected alone, so a surprising score can be traced to its cause.
- A weighted sum of three rated factors shows the arithmetic in a small illustration.
- The EQOS Asia standard workflow has 4 steps, from NDA and scoping to deliverable and debrief (EQOS Asia method page, as of October 2026).
- The EQOS Asia decision engine reports a φ(S) score between 0 and 1, with a STOP zone below 0.618.
A decision engine looks like a black box until its steps are laid out one by one. This guide follows a situation from the first question to the final score, using a small worked illustration, and then maps the stages onto the standard EQOS Asia workflow. The overview in what a decision engine is introduces the same stages in brief.
On this page
Compare the EQOS Asia offers and request a 30-minute pre-diagnostic.
How does a decision engine start from a single question?
Every run begins with one written question: invest or wait, buy or walk away, enter this market or another. The sentence fixes the scope of the model and tells everyone what the score will answer. A vague question produces a vague score, so teams spend their first minutes on wording. Writing the question first also gives the later steps a fixed target to aim at.
The question also names the options and the owner of the final call. With those three elements in place, the rest of the process becomes a matter of filling in the model. A short, precise question is the cheapest quality control a decision engine offers, and it costs a few minutes.
How does a decision engine collect and date its inputs?
Inputs are the facts and ratings that describe the situation. Some are measured, such as revenue, dates, and exposure. Others are rated by experts on a scale, such as the clarity of a contract or the trust inside a team. A few well-defined inputs with a clear source serve the model better than a long list of loose ones.
Each input carries its source and its date. This record lets a reviewer see where a figure came from and lets the team refresh the score when a figure changes. The EQOS Asia method collects inputs through structured interviews and documents that are collected and vectorized (EQOS Asia method page, as of October 2026).
How does a decision engine put different inputs on one scale?
Revenue in US$ and a trust rating from 0 to 100 cannot be added directly. The engine first converts every input to a common range, often 0 to 1, where 0 marks the weakest position and 1 marks the strongest. This step is called normalisation, and it makes unlike things comparable. With every input on the same range, the next step can compare them fairly.
The conversion rules are part of the model and stay visible to reviewers. A change in the range, for example a new definition of strong revenue, changes every later step, so teams document the rule next to the input. Clear scaling lets every factor contribute in proportion to its weight.
How does a decision engine weight each input?
Weights express how much each input matters to the outcome. A factor that drives the result strongly receives a high weight, and a minor factor receives a low one. Weights come from the design of the model, from expert judgment, or from back-testing against past cases, and the choice is stated openly.
Stated weights invite a useful challenge. A reviewer may argue that stability deserves a higher share, and the team can test that view by changing the number and rerunning the score. The disagreement becomes specific, measurable, and quick to settle, which keeps the meeting focused on the decision.
How does a decision engine combine weighted inputs into a score?
Aggregation combines the weighted inputs into one score. In a small illustration, a team rates three factors at 0.8, 0.6, and 0.4 and gives them weights of 50, 30, and 20 percent. The weighted sum is 0.40 plus 0.18 plus 0.08, which equals 0.66. This toy example shows the arithmetic and is not the EQOS Asia formula.
Real models add interactions between factors and many more inputs, and the principle stays the same. A written rule turns many ratings into one number on a known scale. The figure summarises the full sequence, from the first question to the threshold that turns the number into an action. Readers can reproduce the toy computation with a pocket calculator in under a minute.
How does a score become a recommended action?
A threshold converts the number into a decision rule. Taking the EQOS Asia cut-off of 0.618 for illustration, the toy score of 0.66 lands in the OK zone, and a score below the cut-off signals a pause. The threshold is chosen before the run, so the result speaks for itself. Everyone in the room knows in advance what each outcome means for the action list.
The EQOS Asia Audit pillar applies this logic: a φ(S) score between 0 and 1, a STOP zone below 0.618, an OK zone above it, and the 3 priority risks identified for follow-up (EQOS Asia method page, as of October 2026). The action list then focuses on those risks.
Readers who want the full structure behind these steps can follow the decision engine architecture guide, which describes the layers from inputs to outputs, the place of thresholds, and the records that keep a verdict reviewable. It also lists 5 questions to ask of any engine before relying on its score.
What four steps take a situation to a score in EQOS Asia?
The EQOS Asia method page lists a standard workflow of 4 steps. Step one covers the NDA and scoping, with an initial briefing of 60 to 90 minutes by video in SGT (UTC+8). Step two is structured collection, step three is the EQOS analysis, and step four is the deliverable and debrief (EQOS Asia method page, as of October 2026).
The durations listed are 1 day, 2 to 5 days, and 3 to 10 days for the first three steps. To see the workflow applied, request the free 30-minute pre-diagnostic, which applies the equation to a real case, and keep the method page open for the details of each pillar.
See also in this guide
Frequently asked questions
How many inputs does a decision engine need?
Enough to describe the factors that move the answer. A handful of well-defined inputs with sources and dates serves the model better than a large pile of loose data. The EQOS Asia Audit pillar measures 9 structural dimensions (EQOS Asia method page, as of October 2026).
Who sets the weights?
The model designer proposes them, experts review them, and back-testing against known past cases can refine them. Because the weights are written down, any reviewer can question one and see how the score would move. In the EQOS Asia engine the weighting coefficients are proprietary (EQOS Asia method page, as of October 2026).
Can the score be refreshed later?
Yes. Because each input keeps its source and date, the team can update one figure and rerun the computation. The EQOS Asia method page describes a model reusable monthly through Sentinel, with real-time alerts when a threshold is approached (EQOS Asia method page, as of October 2026).
How long does a full EQOS Asia analysis take?
The method page lists 1 day for scoping, 2 to 5 days for collection, and 3 to 10 days for analysis, followed by the deliverable and debrief. A free 30-minute pre-diagnostic comes first and applies the equation to a real case (EQOS Asia method and contact pages, as of October 2026).
Test your knowledge
1. What does normalisation do to inputs?
- Deletes weak inputs
- Converts them to a common range
- Translates them into text
Show the answer
Converts them to a common range.
2. In the worked illustration, what is the weighted sum?
- 0.66
- 0.60
- 0.80
Show the answer
0.66.
3. What does a threshold do?
- Collects inputs
- Rates experts
- Turns the score into an action
Show the answer
Turns the score into an action.
4. How long is the initial EQOS Asia briefing by video?
- 10 to 20 minutes
- 60 to 90 minutes
- 3 days
Show the answer
60 to 90 minutes.
5. Which step of the EQOS Asia workflow lasts 3 to 10 days?
- EQOS analysis
- NDA and scoping
- Structured collection
Show the answer
EQOS analysis.
About the author
Stéphane Jambu
President, EQOS Forecast Asia
Stéphane Jambu leads EQOS Forecast Asia and works as a consultant in e-commerce, search visibility, and AI. LinkedIn profile
Sources: EQOS Asia method page (eqos-asia.com/method.html) and contact page (eqos-asia.com/contact.html), consulted on 21 October 2026. Page updated on 21 October 2026.