Clients are asking a fair question: If AI speeds up drafting and review, how does that change the way you staff and price matters?
I’m the CEO of Gavel and a former practicing attorney at Sidley Austin. This question has been relevant since before the generative AI boom in legal. I still remember clients “forcing” us to use Relativity for document review in 2013. The firms that answer the pricing question clearly don’t just avoid friction; they win work.
This article offers a practical guide for transactional practices based on my experience talking to our law firm customers at Gavel, particularly those using our AI contract analysis and redlining tool. It covers what to disclose, how to structure quality controls, and four pricing models that reward efficiency. It also shows how to measure your baseline so fees reflect value, not just elapsed time.
What’s actually changed since the legal AI boom, and what hasn’t
AI like Gavel Exec can help you throughout the drafting and negotiation lifecycle, with lawyer-approved drafting assistance, playbooks that automate issue spotting and redlining, and benchmarking against market terms. That not only compresses time on repetitive work, but it allows you to do more tasks that you never would have done without AI.
Meanwhile, what hasn’t changed is where clients see value: sound judgment on tradeoffs, tailoring language to context, strategic negotiation, and risk allocation. Your pricing should reflect both realities.
Two principles guide the rest of this piece:
- Separate value from method. Clients pay for outcomes and judgment. AI is one method among many to reach those outcomes.
- Be explicit about risk controls. Efficiency is only persuasive when paired with clear safeguards on quality and confidentiality.
New engagements: the questions sophisticated clients want answered
Put these in plain language at the start of an engagement:
- Scope: Which tasks may be AI‑assisted, and which stay lawyer‑only?
- Quality: How do you verify accuracy before anything leaves the firm?
- Data: How is client information protected inside tools and vendors?
- Pricing: How does AI change the fee on this matter today?
- Change control: What events trigger a re‑scope or revised fee? (Note: hourly billing might still be the answer!)
If you can answer those five questions in a single page, your team will relax, with fewer back‑and‑forths later.
Build the one‑pager you’ll attach to your engagement letter
Use this structure. And edit it for your practice area and client expectations.
1) Statement of use (plain English)
We may use AI tools to accelerate drafting and review. A licensed lawyer supervises all outputs and remains responsible for the work. We do not rely on AI for legal advice or factual assertions without independent verification.
2) Quality controls
- Written playbooks for each document type (e.g., MSA, DPA, lease).
- Issue lists tied to your negotiation posture.
- Second‑lawyer checks on key risk sections (indemnity, limitation of liability, data security).
3) Data handling
Your information is handled under our confidentiality obligations. Our vendor settings prevent training on your data, and we never use public or non-legal-specific AI tools. Access is limited to the team on your matter. We retain logs of AI‑assisted work for audit and defensibility.
4) Pricing approach
Pick the model for this matter (see below). Explain it in one paragraph and tie it to the scope you just defined.
5) Exceptions and change triggers
List the objective signals that move you off the agreed price: non‑standard counterparty positions, unexpected regulatory issues, or deal dynamics that require new workstreams.
6) Optional disclosure language
If a court or regulator requires an AI‑use attestation, we will provide one consistent with these quality controls.
This one‑pager sets expectations and becomes the anchor for pricing conversations.
Four pricing models AI-forward firms are using
After speaking with many of the lawyers on our platform, these are the pricing models that are designed for transactional work. Make sure to choose per matter, not firm‑wide.
1) Fixed fee anchored to a baseline
When to use: Repeatable documents with stable risk, like vendor contracts, NDAs, DPAs, standard SaaS agreements, routine consents.
How to set the price:
- Measure your historical baseline time and variation.
- Decide the portion of the efficiency you’ll share now. Clients don’t expect you to give away the entire dividend.
- Price the matter as a fixed fee that reflects some expected AI‑enabled efficiency but still pays for expertise and review.
Why it works: Clients get predictability; you get a higher effective rate than hourly would produce after time compression.
What to watch out for: Define change triggers. Third‑party paper can explode scope if you’re not clear about what “standard” means.
2) Portfolio subscription (retainer with SLAs)
When to use: A steady flow of similar items, like vendor agreements, reseller addenda, DPAs, SOWs, and routine amendments.
Structure:
- Monthly fee includes a forecasted volume (e.g., 20 agreements per month).
- Include access to the AI playbooks you have created in an AI platform like Gavel.
- Overages are priced per unit at a published rate.
- SLAs tied to business impact (e.g., first‑pass redline in 24 hours; partner review within 48 hours; emergency lane with a surcharge).
- Quarterly tune‑up: adjust volume tiers and overage rates to reflect real demand.
Why it works: Clients get capacity on tap. You get predictable revenue and batching economies across the team.
What to watch out for: Be precise about what “counts” as an item and how rounds of negotiation affect the unit. Define when a redraft becomes a new unit.
3) Hybrid caps and floors
When to use: Third‑party paper, complex commercial deals, or anything with wide variance.
Structure:
- Fixed fee covers the central 60–70% of scenarios.
- Floor: If the work is unusually light, you still charge the fixed fee to protect against under‑recovery.
- Cap: If predefined variance triggers fire (e.g., more than three material negotiation rounds, five or more non‑standard positions, major regulatory add‑ons), you re‑scope with the client.
- Offer an add‑on package at a preset price for the most common overages (for example, cybersecurity exhibit rewrites or unusual data‑transfer language).
Why it works: Clients get predictability with honest guardrails. You avoid being trapped in outlier scenarios.
What to watch out for: Document the triggers and keep them objective to prevent disagreements later.
4) Outcome‑linked bonus (where allowed)
When to use: Clear, objective outcomes that meaningfully affect business value: cycle time, acceptance rate, or securing a specific commercial position.
Structure:
- Fixed fee plus a modest bonus if the team hits the agreed outcome (e.g., final execution within ten business days or acceptance of your indemnity formulation).
- Avoid subjective metrics. Keep the bonus small enough that it doesn’t distort advice.
Why it works: Aligns incentives without turning legal work into contingent compensation.
What to watch out for: Confirm ethics rules in your jurisdiction and client policies, especially for anything that looks like contingency or success‑based compensation.
How to set prices with real measurement, not guesswork
AI can make you faster, but quoting needs evidence. A light‑weight, four‑step approach:
- Choose two workflows that are frequent and bounded (e.g., DPAs on third‑party paper).
- Measure the baseline for ten recent matters per workflow: total hours, rounds of negotiation, time per round, and sections that took the most effort. Track the median and heavy scenarios.
- Document the AI‑assisted process for the next ten matters: first‑pass generation time, lawyer review time, rework due to AI errors, and final acceptance time.
- Set pricing bands by anchoring fixed fees to the median, with an add‑on or re‑scope clause for 90th percentile‑type variance.
How to present this in a pitch
Use simple, confident language:
- “You get predictable pricing. We keep senior‑lawyer judgment on the final call.”
- “Here’s the quality assurance we run before any document leaves our hands.”
- “This fee reflects time saved on routine edits. You’re paying for expertise and results, not idle time.”
- “If the matter becomes novel or contentious, we pause and re‑scope before charges change.”
Bring a one‑slide version of your playbook: which sections are AI‑assisted, which are lawyer‑only, and what the review path looks like.
A 30‑day implementation plan
Week 1
- Pick two workflows (e.g., vendor MSAs and DPAs).
- Pull ten recent matters per workflow to measure time on each matter, how many rounds, and pain points.
- Draft your one‑pager using the structure above.
Week 2
- Pilot the fixed‑fee model on one workflow and the share‑the‑savings model on the other.
- Set SLAs that reflect your current reality, not your hopes.
- Configure tools with correct vendor settings and document your steps.
Week 3
- Run five live matters through the pilots. Capture time, variance triggers, and acceptance rates.
- Tune your fixed fee and define add‑ons for the common overages.
Week 4
- Debrief with your team and one friendly client sponsor.
- Lock the one‑pager, invoice language, and pitch slide.
- Publish a short client explainer on your website.
A short case study
Client: Growth‑stage SaaS company negotiating DPAs on third‑party paper
Before: Median of 6 hours per DPA across two rounds; $3,000 billed; high variance when counterparties insisted on non‑standard security exhibits
After: AI‑assisted first pass with lawyer supervision; 2.5 hours median; fixed fee $2,125 using a 50% share‑the‑savings model; add‑on package for non‑standard security exhibits at $600
Results:
- Client: $875 savings vs. historical baseline, consistent 48‑hour first‑pass SLA, fewer internal escalations.
- Firm: Effective rate $850/h on the 2.5 hours of real work, cleaner invoices, faster collections, and a three‑month retainer expanded to include vendor MSAs.
What’s next for your firm
Firms don’t need to choose between guarding efficiency gains and racing to the bottom on price. Share the dividend in a way that makes the client’s life easier: predictable fees, clear quality controls, faster turnarounds, while keeping your economics healthy. The mechanics are straightforward: a one‑page disclosure, a pricing model matched to the work, and a governance checklist you actually follow.
Pick one workflow, one model, and one client. Run the play this quarter. The trust you build, and the win rates that follow, will do more for your practice than another slide about innovation.
