AGENCYSCRIPT
CoursesEnterpriseBlog
đź‘‘FoundersSign inJoin Waitlist
AGENCYSCRIPT

Governed Certification Framework

The operating system for AI-enabled agency building. Certify judgment under constraint. Standards over scale. Governance over shortcuts.

Stay informed

Governance updates, certification insights, and industry standards.

Products

  • Platform
  • Certification
  • Launch Program
  • Vault
  • The Book

Certification

  • Foundation (AS-F)
  • Operator (AS-O)
  • Architect (AS-A)
  • Principal (AS-P)

Resources

  • Blog
  • Verify Credential
  • Enterprise
  • Partners
  • Pricing

Company

  • About
  • Contact
  • Careers
  • Press
© 2026 Agency Script, Inc.·
Privacy PolicyTerms of ServiceCertification AgreementSecurity

Standards over scale. Judgment over volume. Governance over shortcuts.

On This Page

The Triage Play: Deciding How Much RigorTrigger and ownerThe movesExit conditionThe Quick Play: Low-Stakes, Reversible DecisionsThe movesOwner and exitThe Full Rigor Play: High-Stakes SelectionsThe moves in orderOwner and exitThe Recurring Play: Comparisons You Run RepeatedlyThe movesOwner and exitThe Governance Play: Keeping It Safe at ScaleThe movesOwner and exitSequencing the Plays TogetherHow they chainWhen to escalate mid-playThe Roles the Plays Depend OnThe requesterThe analystThe reviewerThe practice ownerCommon Ways the Playbook Breaks DownSkipping triageLetting the reviewer role lapseTreating templates as permanentFrequently Asked QuestionsHow do I decide how much rigor a comparison needs?What is the minimum for a low-stakes comparison?What makes the full rigor play different?When should a comparison become a template?What does the governance play add?Can I switch plays partway through?Key Takeaways
Home/Blog/Run the Right Comparison Play for the Stakes at Hand
General

Run the Right Comparison Play for the Stakes at Hand

A

Agency Script Editorial

Editorial Team

·October 31, 2021·6 min read
prompting for comparative analysis tasksprompting for comparative analysis tasks playbookprompting for comparative analysis tasks guideprompt engineering

A playbook is not a tutorial. A tutorial teaches you to do a thing once; a playbook tells you, when a specific situation arises, exactly which play to run, who runs it, and what has to be true before you move on. For AI-assisted comparative analysis, the difference matters because comparisons come in recognizable shapes — the quick gut-check, the high-stakes vendor selection, the recurring competitive review — and each shape calls for a different sequence. Treating them all the same is how you over-engineer a trivial choice and under-engineer a consequential one.

This article lays out the plays. For each, you get the trigger that calls it, the sequence of moves, the owner, and the exit condition that says the play is done. The plays build on techniques covered elsewhere; here they are organized into an operating system you can run from the moment a comparison request lands to the moment a defensible recommendation ships.

The Triage Play: Deciding How Much Rigor

Not every comparison deserves the full treatment. The first play is deciding which play to run.

Trigger and owner

Trigger: any comparison request arrives. Owner: whoever receives it. This play takes two minutes and prevents the two classic failures — drowning a small decision in process and rushing a big one.

The moves

Ask three questions: How reversible is this decision? How costly is being wrong? How often will we make this exact comparison again? High stakes or high recurrence routes to the full rigor play. Low stakes and one-off routes to the quick play.

Exit condition

You have named which downstream play applies. Do not start comparing until triage is done.

The Quick Play: Low-Stakes, Reversible Decisions

For choices that are cheap to reverse, speed beats ceremony.

The moves

Supply the options and your criteria, ask for a table plus a short ranked recommendation, and verify only the single most decision-critical fact. This is the structured first-pass approach from Your Path From Zero to a Trustworthy First Comparison, with verification scoped to the essentials.

Owner and exit

Owner: the individual making the decision. Exit: a ranked recommendation with its load-bearing fact checked. Done in minutes — anything more is over-engineering.

The Full Rigor Play: High-Stakes Selections

When the decision is costly or hard to reverse, run the complete sequence.

The moves in order

Define and weight criteria with anchored scales. Score options in a separate step before applying weights. Have the model show the weighted arithmetic. Run an adversarial pass where it argues against its own recommendation, then reconcile. Shuffle option order and re-run to check robustness. Verify every load-bearing fact. These techniques come from Advanced Prompting for Comparative Analysis.

Owner and exit

Owner: the analyst, with a named reviewer. Exit: an auditable comparison — recommendation traceable through weighted scores to evidence labels — that a reviewer has signed off and whose facts are verified.

The Recurring Play: Comparisons You Run Repeatedly

Some comparisons come back every quarter. These deserve standardization, not reinvention.

The moves

Build the comparison once to full rigor, then template it: criteria, weights, scales, and output format become reusable. Each cycle, refresh the facts, re-verify time-sensitive claims, and date-stamp the result. This is where the playbook connects to Building a Repeatable Workflow for Prompting Comparative Analysis.

Owner and exit

Owner: a named keeper of the template. Exit: a refreshed, date-stamped comparison produced from the standing template with current facts verified.

The Governance Play: Keeping It Safe at Scale

Running across a team adds a layer the individual plays do not cover.

The moves

Enforce a written verification standard, require that inputs be recorded, and keep accountability with the human who ships each comparison. Spot-check a sample of outputs for quality. These safeguards come from When a Confident AI Comparison Quietly Steers You Wrong.

Owner and exit

Owner: the practice owner for the team. Exit: standards are enforced by default in the workflow, not left to individual discretion. For the human side of getting there, see Getting a Whole Department to Compare Options the Same Way.

Sequencing the Plays Together

How they chain

Every comparison starts at triage. Triage routes to quick or full rigor. Recurring comparisons graduate from full rigor into a template. Governance wraps all of it once more than one person is involved. The plays are not alternatives — they are a system that scales rigor to stakes.

When to escalate mid-play

If a quick play surfaces unexpectedly high stakes — a fact that makes the decision suddenly consequential — stop and re-triage to full rigor. The plays are meant to flex, not to lock you in once chosen.

The Roles the Plays Depend On

A play without clear ownership stalls. Name the roles before you need them.

The requester

Whoever asks for a comparison owns supplying the real decision context — what the choice is for, what the hard constraints are, what timeline applies. A comparison built on a vague request is shallow no matter how well it is run. The requester's job is to make the decision legible.

The analyst

The analyst runs the play: framing criteria, supplying facts, driving the model, and verifying load-bearing claims. This is where the analytical judgment lives, and it is the role that determines whether the output is defensible. The analyst owns the conclusion until a reviewer signs off.

The reviewer

For full-rigor plays, a named reviewer checks the comparison against the standard — math shown, facts verified, recommendation traceable. The reviewer is not a rubber stamp; their job is to find the buried flaw before a decision-maker acts on it. Separating analyst and reviewer is what catches errors a single person misses.

The practice owner

Across many comparisons, someone owns the templates, the criteria library, and the standards. Without this role the plays drift and decay. The practice owner keeps the system coherent as the work evolves, a responsibility carried into Turning One Good AI Comparison Into a Repeatable Process.

Common Ways the Playbook Breaks Down

Skipping triage

The most frequent failure is running every comparison at the same intensity — drowning trivial choices in ceremony or rushing consequential ones. Triage is two minutes and it is the play that makes every other play efficient. Never skip it.

Letting the reviewer role lapse

Under deadline pressure, the review step is the first thing teams quietly drop. That is exactly when it matters most, because rushed comparisons carry the most errors. Protect the reviewer role as non-negotiable for high-stakes plays.

Treating templates as permanent

A recurring-play template that never gets refreshed becomes a confident source of stale conclusions. Schedule fact refreshes and re-verification, and date-stamp every output so nobody mistakes last quarter's analysis for current truth.

Frequently Asked Questions

How do I decide how much rigor a comparison needs?

Run the triage play: assess reversibility, cost of being wrong, and recurrence. High stakes or high recurrence earns full rigor; low-stakes one-offs get the quick play. Triage first, always.

What is the minimum for a low-stakes comparison?

Supply options and criteria, ask for a ranked table, and verify the single most decision-critical fact. Anything more on a cheap, reversible decision is wasted ceremony.

What makes the full rigor play different?

Weighted criteria with anchored scales, separated scoring and weighting, visible arithmetic, an adversarial self-critique, an order-shuffle robustness check, and verification of every load-bearing fact — signed off by a named reviewer.

When should a comparison become a template?

When you run the same comparison repeatedly. Build it once to full rigor, then template the criteria, weights, scales, and format, refreshing and re-verifying facts each cycle.

What does the governance play add?

It wraps the individual plays once a team is involved: an enforced verification standard, recorded inputs, clear human accountability, and quality spot-checks. It turns personal discipline into a default.

Can I switch plays partway through?

Yes, and you should. If a quick play uncovers high stakes, stop and re-triage to full rigor. The plays flex to the situation rather than locking you in.

Key Takeaways

  • Start every comparison with a two-minute triage that routes it to the right level of rigor.
  • The quick play scopes verification to the single load-bearing fact for cheap, reversible decisions.
  • The full rigor play adds weighting, separated scoring, visible math, adversarial critique, robustness checks, and a named reviewer.
  • Recurring comparisons graduate into date-stamped templates with refreshed, re-verified facts.
  • The governance play wraps everything at team scale, and plays should flex — re-triage upward the moment stakes rise.

Search Articles

Categories

OperationsSalesDeliveryGovernance

Popular Tags

prompt engineeringai fundamentalsai toolsthe difference between AIMLagency operationsagency growthenterprise sales

Share Article

A

Agency Script Editorial

Editorial Team

The Agency Script editorial team delivers operational insights on AI delivery, certification, and governance for modern agency operators.

Related Articles

General

Prompt Quality Decides Whether AI Earns Its Keep

Prompt quality is the single biggest variable in whether AI delivers real work or expensive noise. The model matters, the platform matters — but the prompt you write determines whether you get a first

A
Agency Script Editorial
June 1, 2026·10 min read
General

Counting the Real Cost of Every Token You Send

Tokens and context windows sit at the intersection of AI capability and operational cost—yet most business cases treat them as technical footnotes. That's a mistake that costs real money. Every time y

A
Agency Script Editorial
June 1, 2026·10 min read
General

Rolling Out AI Hallucinations Across a Team

Most teams discover AI hallucinations the hard way — a confident-sounding wrong answer makes it into a client deliverable, a legal brief, or a published report. The damage isn't just to the output; it

A
Agency Script Editorial
June 1, 2026·11 min read

Ready to certify your AI capability?

Join the professionals building governed, repeatable AI delivery systems.

Explore Certification