How to pilot a decision engine on one decision in 30 days

Why pilot a decision engine on a single decision?

A single decision keeps the pilot small enough to finish and large enough to matter. Everyone can judge the result against a real choice with a real deadline, and the team learns the full cycle: briefing, data, score, debrief. A pilot spread over many decisions would take longer and make the comparison with your own judgment harder. The same plan can be repeated for the next decision with fewer surprises.

Define success before the work starts. A good test reads like this: one score, one recommendation, one next action, and a record of the effort. If the pilot delivers those four things inside 30 days, the engine has earned a second decision.

Write the success test as a short list of criteria, each with a measure, a target and an owner. 3 groups cover most pilots: measurable, quality and adoption criteria. The page on pilot project success criteria shows thresholds and an illustrative scorecard that turns the debrief into a clear go or no-go.

Which decision is the right one for a 30-day pilot?

Choose a decision with a deadline, an owner, and facts that can be gathered within a week. Typical candidates in the EQOS Asia pricing examples include whether to sign a contract, whether to partner up, whether to hire a key executive, and whether an investment is viable (EQOS Asia pricing page, as of October 2026). Pick the one your team already debates in meetings, because the pilot then answers a live question.

Favour decisions that are still open, and especially those where a wrong call would be costly enough to justify an hour of preparation. Write the question in one sentence and name the executive who owns the final call. That executive attends the first call and the debrief.

A short proof of concept can come first when the team wants to test one narrow claim before a live decision. It uses a past case with a known outcome, a written success test and exit rules. The page on a proof of concept for a decision engine sets out each of the 3 parts and the tiers that fit.

What happens in the first week of a decision engine pilot?

Week one frames the pilot. You send a short description of the situation, with the countries and currencies involved, and EQOS Asia replies within 24 hours on business days with a 30-minute video slot in Singapore time (SGT, UTC+8) (EQOS Asia contact page, as of October 2026). The call identifies the right tier for the question.

The standard workflow then opens with the NDA and scoping, in 1 day, including an initial briefing of 60 to 90 minutes by video (EQOS Asia method page, as of October 2026). By the end of the week, the scope, the data list, and the date of the debrief are written down and agreed by both sides. Share the written scope with everyone who will attend the debrief.

How does the pilot collect the decision inputs in the second week?

Week two gathers the facts. The structured collection takes 2 to 5 days and covers internal documents, targeted interviews, and existing data, in an encrypted, isolated environment (EQOS Asia method page, as of October 2026). Any format is accepted, digital or physical, so the team can send what it holds today.

Assign one person to answer questions from the provider and keep a list of open items. Label each fact with its source and date. A quick reply to each question protects the schedule, because the analysis can start as soon as the collection closes. Short daily check-ins of ten minutes keep the collection on schedule.

What does the decision engine deliver in the third week of the pilot?

Week three is the computation. The EQOS Asia analysis runs for 3 to 10 days: the Codex Universalis core computes the φ(S) score, models the tensions, and projects scenarios (EQOS Asia method page, as of October 2026). A Study at US$1,369 delivers audit, modelling, and prediction in 5-7 days, with an interactive HTML deliverable and a 30/60/90 day action plan.

A Decision at US$910 models one structural choice in 48-72 hours and fits a pilot with a tighter calendar (EQOS Asia pricing page, as of October 2026). Whichever tier you select, ask for the deliverable format in advance, and plan the debrief while the analysis runs.

How do you read the score and decide in the last week of the pilot?

The last week turns the score into a choice. The deliverable lists immediate decisions, deferrable decisions, and a recommended sequence. A φ(S) score below 0.618 falls in the STOP zone and a score above it falls in the OK zone, with the 3 priority risks named (EQOS Asia method page, as of October 2026).

Compare the result with your own judgment from the start of the pilot. Where they agree, the engine gives that call a documented basis. Where they differ, the gap shows which fact or assumption to review. The executive keeps the final call, and the debrief records the reasons for it. Write the decision and its reasons in two lines, and share them with the team.

What can follow a 30-day decision engine pilot?

Three paths open after the debrief. You can apply the engine to the next decision, move to a deeper tier such as the Diagnostic at US$2,288 over 2 weeks, or keep the model current with Sentinel, which re-measures the 9 dimensions monthly and carries a 6-month minimum commitment (EQOS Asia pricing page, as of October 2026).

Review the pilot against the success test you wrote in week one: score, recommendation, next action, effort. The pricing page lists the next tiers, and the free 30-minute pre-diagnostic starts the first week of a new pilot. Each next step starts from the record of the pilot.

Frequently asked questions

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