contract negotiation

Contract negotiation software: what it does in each round and how to test one

Meera HaldenContracts and legal ops writer10 min read

Written by Meera Halden Contract negotiation software is the set of tools that sit between "the other side sent their draft" and "both sides agreed the final wording". It helps a lawyer or a commercial team mark up the draft against the company's own positions, decide what to concede, keep track of which version is live, and see later which clauses always cause a fight. Some products sell this as a stand-alone AI contract negotiation tool inside Microsoft Word; others bundle it into a full contract lifecycle platform. This guide explains what each part of the job looks like in practice, how four vendors that market negotiation features describe them, and how to test a contract negotiation platform on your own documents before you pay for it. ## Negotiation is six separate jobs, not one feature Vendors use the word "negotiation" for very different things. Before you look at any demo, split the work into the jobs your team actually does each week. Most teams find that two or three of these matter and the rest are nice to have. 1. Playbook-driven redlines. The tool compares the incoming clause with your preferred wording and proposes a markup. This only works if somebody has written the preferred wording down. If your playbook lives in a senior lawyer's head, the software has nothing to apply. 2. Fallback positions. A good playbook says what you want, what you will accept if pushed, and where you walk away. Software that understands fallbacks can offer the second position when the first is rejected, instead of starting from zero each round. Our page on playbooks vs fallback positions explains how to structure that ladder. 3. Review of counterparty paper. When the customer or supplier drafts, you are reacting to their structure, their definitions and their order of clauses. The tool has to find your issues inside someone else's document, which is harder than checking your own template. 4. The issues list. Many deals are closed through a short table of open points sent alongside the markup: clause, their position, our position, proposed compromise. A tool that can produce this table saves the most tedious hour of every round. 5. Version tracking. Round three of a supply agreement can involve four files with similar names in three inboxes. The tool should show which draft is current, what changed between rounds and who changed it. 6. Negotiation analytics. Across fifty or five hundred deals, which clauses are conceded most often, how long each round takes, and which counterparties always push back on liability. This is useful for large teams and close to useless for a team doing ten deals a quarter. ## What AI contract negotiation can and cannot do today Language models are good at the first, third and fourth jobs. They read a long clause quickly, compare it to a standard, and write a sensible counter-proposal in the right register. They are also good at summarising a counterparty's markup into an issues list. They are weaker where judgment depends on facts outside the document. A model does not know that this customer is your largest account, that the sales director has already promised a lower liability cap on a call, or that your insurer will not cover the indemnity the other side wants. It also does not know which of your fallbacks are genuinely approved and which are a junior lawyer's guess, unless the playbook says so. Treat AI output in negotiation as a first draft by a capable junior. It will catch the missing limitation of liability and the one-sided termination right. It will not decide whether to trade the payment term for the warranty period. That decision stays with a person who knows the deal. ## Four vendors that market negotiation, in their own words The table below summarises how each vendor describes its negotiation features on its own website, read in November. Feature names change often, so use this as a list of questions for the demo, not as a verdict. | Vendor | Where the work happens | How positions and fallbacks are described | Ask in the demo | | --- | --- | --- | --- | | Luminance | Markups inside Microsoft Word; Outlook also connected | Its negotiation page says playbooks become structured standards with preferred positions, fallback language and review criteria, and that precedent agreed with the same counterparty is brought into new deals | How precedent is chosen when two past deals conflict | | Ironclad | Inside the platform and a Word add-in | Its Contract Review Agent, formerly called Jurist, redlines full agreements against playbooks covering preferred positions and fallbacks | Who builds and maintains the playbook, and how long that takes | | SpotDraft | VerifAI, a Word add-in, and inside the SpotDraft platform | Its help article on guides describes typical wording plus up to two fallback positions, and you choose which one to apply as the redline | What happens when the other side rejects both fallbacks | | Juro | A browser-native editor rather than Word | Its negotiate page stresses threaded comments, separate internal and external versions, and an audit trail of views and approvals | Whether counterparties will accept working outside Word | Two patterns are visible. Luminance, Ironclad and SpotDraft meet lawyers in Word and put the playbook at the centre. Juro tries to move the whole negotiation into its own editor, which suits high-volume sales contracts where the counterparty is happy to click a link, and suits heavily lawyered deals less well. Neither approach is wrong; they solve different problems. ## Sell-side and buy-side teams want different things A sales contracts team negotiates the company's own paper again and again. Its pain is volume and speed: hundreds of similar NDAs, order forms and master agreements, each with small variations. For this team the value sits in self-serve playbooks for sales staff, approval routing for non-standard terms, and analytics that show which concessions keep repeating so the template can be fixed at source. If the same liability carve-out is conceded in most deals, the right answer is often to change the template, not to negotiate faster. A procurement or in-house team buying services usually works on the supplier's paper. Its pain is variety: every vendor has a different structure, different definitions and a different idea of what "standard" means. Here the value sits in reading unfamiliar documents accurately, finding the missing clause (no data return on exit, no service credits, no cap on price increases), and producing a clean issues list the supplier can respond to. Playbook redlining still helps, but the hard part is finding the issue in the first place. Law firms sit in between. They negotiate for clients whose positions differ from deal to deal, so a single company playbook fits them less well. A firm usually wants per-client positions and strong version control more than business-user self-service. Know which group you are before you shortlist, because the same contract negotiation platform can be excellent for one and poor for another. ## Running a fair two-week test of a contract negotiation platform Demos use the vendor's sample contracts, which are clean and predictable. Your contracts are not. A short test with real documents tells you more than any feature list. Week one: prepare. - Pick eight to twelve agreements you negotiated in the last year, including at least three on counterparty paper and one that went badly.

  • Write down your current playbook for five clauses: limitation of liability, indemnity, termination, payment terms and governing law. Include one fallback for each. The contract negotiation checklist by clause is a useful starting list.
  • Record how long a first-pass markup takes today for two of the documents. This is your baseline. Week two: run and score. - Load the playbook into the tool, or ask the vendor how long they need to do it. That answer is itself a data point.
  • Run each document and score the output with the sheet below. Have the lawyer who negotiated the original deal do the scoring. | Test | What a pass looks like | Score 0 to 2 | | --- | --- | --- | | Finds the issue | Flags every clause that departs from the playbook, including ones phrased unusually | | | Proposes the right position | Uses your preferred wording, not generic text | | | Offers the fallback | When told the first position was rejected, moves to your approved fallback | | | Handles counterparty structure | Works when definitions and numbering differ from your template | | | Produces an issues list | Clause, their position, ours, suggested compromise, in a form you can send | | | Shows what changed | Clear comparison between rounds, with author and time | | | Explains itself | Each suggestion carries a reason a lawyer can check quickly | | A tool that scores well on the first three but poorly on the rest is a review tool, not a negotiation tool. That may be all you need. ## Questions that separate negotiation software from a clever Word macro - Can a business user run the playbook alone, and what stops them conceding a position that needs legal approval?
  • When the counterparty sends back a clean file with no tracked changes, does the tool still find what they changed?
  • How is precedent selected, and can you exclude deals that were closed on unusual terms?
  • Where do negotiation analytics come from, and how many deals are needed before the numbers mean anything?
  • What happens to your playbook and your documents if you leave the vendor? Ask for the export format in writing.
  • Is customer data used to train the vendor's models? Get the answer from the contract, not the sales deck. ## Where Adira fits in a negotiation, and where it does not Adira is a contract workspace rather than a dedicated negotiation platform, and it covers the review and editing part of the job. When the other side's draft arrives, its Document Intelligence Dashboard marks each position as favourable, neutral or unfavourable to your company and lists red flags, missing clauses and a single risk-posture reading for the document. Clause-Level Editing then lets you ask for a change in plain English, for example "cap liability at twelve months of fees everywhere", and Adira finds the affected clauses, makes the edit and returns a diff you can check before anything goes back. Company Persona, built from your own documents, records your preferred terms and risk posture, and Adira uses it to tune drafts toward your company. Files and the Activity Timeline keep each round of the document with its history, and you can export the result to Word or PDF to send to the counterparty. If your priority is negotiation analytics across hundreds of past deals, or a multi-step fallback ladder that business users run without a lawyer, one of the dedicated tools above is the better fit. Adira's published pricing starts at USD 89 per seat per month billed annually on the Practice plan (USD 109 month to month), with a minimum of three seats, so a three-seat team on annual billing pays USD 3,204 a year. There is a 7-day free trial, which is long enough to run the week-two test above, and Adira does not train its models on customer contracts. Details are on adiralaw.com. ## A short buying summary - Decide which of the six jobs you need before you book demos.
  • Write five clauses of playbook with fallbacks first; no tool works without them.
  • Test on counterparty paper, not on your own template.
  • Score with the lawyer who did the original deal.
  • Buy analytics only if your deal volume will feed them.

Frequently asked questions

Do I need a written playbook before buying contract negotiation software?
Yes, for anything beyond basic review. Playbook-driven redlining applies your preferred wording and fallbacks; if those are not written down, the tool falls back to generic suggestions. Start with five clauses and one fallback each.
Can AI contract negotiation tools negotiate without a lawyer?
Some vendors let business users handle routine contracts within set limits and route non-standard terms to legal. For anything that trades one clause against another, a person who knows the commercial context should decide.
Is a Word add-in or a browser editor better for negotiation?
A Word add-in suits deals where the counterparty insists on Word and lawyers on both sides mark up. A browser editor suits high-volume sales or procurement contracts where the other side will work in a shared link.
How many contracts should a negotiation software trial use?
Eight to twelve real agreements from the last year is enough, with at least three on counterparty paper. Score each output against a written playbook, using the lawyer who negotiated the original deal.
Does Adira replace a dedicated contract negotiation platform?
Not fully. It reviews a draft for favourable and unfavourable positions and makes clause-level edits with a diff, but teams that need negotiation analytics across many deals or business-user fallback ladders should look at a dedicated tool.
Was this useful?

See how Adira drafts in your voice and reads contracts from your side.

Explore the showroom

Working through a contract like this? Weave is Adira’s free tool to read, mark up, and connect any contract in your browser — no account needed.

Try Weave — free

Book a demo

See Adira on your own contracts

A 30-minute walkthrough with our team. Tell us a little about you, then pick a time that suits you.

Prefer email? [email protected]