The Review Style sets how aggressively the AI flags issues when it compares your document to the Playbook. There are three Styles — Pragmatic, Balanced and Strict — and switching between them changes how many findings you get and how much room the AI gives the document before flagging something. This article explains how each Style works and when to use it.
Tip: Review Style only affects how the AI uses your Playbook. It has no effect on the Drafting Quality, Jurisdiction or Market Practice lenses, which apply the same checks regardless of Style.
How the three Styles compare
Each Style benchmarks the document against a different position in your Playbook clauses. Your Playbook can hold up to three positions per clause: Must Have (the non-negotiable minimum), Want (your preferred outcome) and Will Accept (the lowest position you'd accept). The Review Style decides which of those positions the AI uses as the bar for the document.
Strict
Benchmarks the document against your Want position — the ideal outcome. Anything that falls short is flagged for negotiation. The AI also checks the language used in the document, not just the intent, when example clause text is provided in the Want position. Strict additionally calls out Playbook clauses that appear to be missing from the document entirely.
In-product description: Flags any departure from preferred clause language, including minor phrasing variations. Use for high-risk engagements or compliance audits.
Use Strict when you're reviewing a high-risk agreement, performing a compliance audit, or want maximum visibility into every gap between the document and your preferred wording.
Expect the highest finding volume of the three Styles.
Balanced
Benchmarks the document against your Will Accept position — the lowest acceptable wording. Anything that meets or exceeds Will Accept is left alone. Anything below Will Accept is flagged. The AI applies commercial reasonableness, distinguishing minor drafting differences from substantive deviations.
In-product description: Flags clauses that materially deviate from your playbook. Minor wording differences tolerated where intent is preserved.
Use Balanced when you want a sensible day-to-day review that surfaces material issues without flooding you with stylistic noise.
Expect a moderate finding volume — the default for most reviews.
Pragmatic
Benchmarks the document against your Must Have position — the non-negotiable floor. The AI applies legal judgment across all positions, surfacing only what it assesses as materially significant. It focuses on clauses approaching or breaching Must Have, provisions exposing the organisation to uncapped or disproportionate liability, and clauses that conflict with the governing law.
In-product description: AI accepts language that broadly achieves the playbook intent. Only material gaps or direct contradictions are flagged.
Use Pragmatic when you're triaging a stack of low-risk documents quickly, or doing a final sanity check on a draft that's already been negotiated.
Expect the lowest finding volume of the three Styles.
Choosing the right Style
There's no single correct Style — the right choice depends on the deal, the counterparty and the stage of negotiation. A useful rule of thumb:
Strict for first-pass reviews of new counterparty paper, compliance audits, and high-value agreements.
Balanced for routine in-house reviews where you want to focus on what genuinely matters commercially.
Pragmatic for fast triage, low-risk documents, and final read-throughs before signing.
You can switch Styles and re-run the analysis at any time using Re-analyze.
Note: The quality of the review under any Style depends on the quality of your Playbook. A Playbook with only Must Have positions populated will produce useful Pragmatic and Balanced reviews, but Strict will have nothing additional to compare against. See Creating a Playbook for guidance.
Related articles
Need help?
If you have any questions, please contact us at support@xakiatech.com or reach out to your Customer Success Manager.