The gap is commitment: knowing what to use, in what order, and for what workflow.

Here's the stack and the rollout plan.

The 72%/8% Tension

Before the tools: why does this gap exist?

Clio's data on the split is useful. Of the 72% who use AI "in some form," the most common tools are generic AI (ChatGPT, 57%) and legal research platforms (54%). These are task-level uses — a lawyer opens ChatGPT to draft a specific email, then closes it. The AI is a one-off tool, not a workflow.

Wide adoption — the 8% — means AI is integrated into how the firm actually operates. Not "I used this last Tuesday," but "every contract of this type goes through this review process before I see it."

The 56% annual growth in solo firm tech spending (Clio, 2025) shows the appetite is there. The 8% wide adoption shows most firms haven't crossed the line from experiment to system. That line is the one worth crossing.

Solo firm tech spending is growing at 56% annually — twice the industry average — but only 8% have committed to wide AI adoption. The cost barrier is not the problem.

The Seven Tools

This stack is drawn from an independent review of small law firm AI tools in 2026 (Bot4orge) and Clio's own practice guidance. The goal: cover every major workflow category without overlap or redundancy. Total cost for a solo firm: $350–$600/month depending on tier.

1. Practice Management with AI — Clio Manage ($49–$149/user/month)

Clio is the operating system. It handles case management, client records, document storage, billing, and time tracking. The AI layer (Clio Duo) adds document summarization, task automation, and AI-assisted client communication. For a small firm, Clio is also the integration hub — most other tools in this stack connect to Clio natively.

If you're not on Clio, evaluate it before adding anything else. Practice management with a built-in AI layer is more durable than a standalone AI tool bolted onto a generic system.

2. Client Intake CRM — Lawmatics ($69/month)

Client intake is where small firms lose the most administrative time and miss the most revenue. Lawmatics automates intake forms, lead scoring, follow-up sequences, and appointment scheduling. Its AI features classify intake submissions and flag high-priority leads.

The firm's first AI-assisted client interaction is intake. If a potential client submits a form and hears nothing for 48 hours, they've already called someone else. Lawmatics closes that gap automatically.

3. Legal Research — Lexis+ AI ($80–$135/user/month)

Lexis+ AI is the established platform for AI-assisted legal research at small firms. Natural language queries, citation-grounded answers, and case analysis that holds up to cite-checking. The AI layer doesn't hallucinate case law in the way general-purpose tools can — it's grounded in Lexis's database and returns citations you can verify.

For a solo attorney, this is the tool that compresses research time most directly. What takes an hour with traditional Westlaw or Lexis search can often be done in 15-20 minutes with a well-constructed AI query.

4. Contract Drafting and Review — Spellbook ($180/user/month)

Spellbook works inside Microsoft Word and Google Docs, reviewing contracts in real time. It flags unusual clauses, suggests alternatives, checks for missing provisions, and identifies risk areas by practice area. For a firm doing transactional work — NDAs, employment agreements, service contracts, commercial leases — Spellbook is the tool that cuts first-pass review time most dramatically.

The ABA's recommended approach for small firms: identify one document type, run it through an AI review tool for 30 days, document what the AI catches versus what you catch manually. Spellbook is the practical implementation of that recommendation.

5. AI Reception and Scheduling — Smith.ai ($140+/month)

Smith.ai handles inbound calls, web chat, and appointment scheduling using AI-powered virtual receptionists. For a solo or two-attorney firm, this replaces the "I'm going to voicemail" problem with a professional intake experience that captures leads and books consults without staff involvement.

The ROI math is simple: one missed client call that would have become a $5,000 engagement covers three months of Smith.ai costs. For firms where the attorney is the only one answering phones, this is the highest-leverage tool in the stack.

6. General-Purpose AI — ChatGPT Business or Claude for Business ($25–$30/user/month)

This is the catch-all layer for everything not covered by the specialist tools: drafting client emails, summarizing documents, brainstorming arguments, researching procedural rules, and preparing for depositions. Both ChatGPT Business and Claude for Business include enterprise data privacy — client data doesn't train the model.

This is not a replacement for Lexis+ AI on research or Spellbook on contract review. It's the Swiss Army knife for everything else.

7. Document Automation — HotDocs or Contract Express (varies by tier)

For firms that produce the same document types repeatedly — engagement letters, standard pleadings, routine agreements — document automation translates your templates into fill-in-the-blank forms. The AI layer populates common fields from intake data. HotDocs is the traditional choice; Contract Express integrates natively with Clio.

This tool is optional in the first 90 days. Add it after you've established consistent use of the first five.

The 90-Day Rollout

Do not deploy all seven tools at once. The agentic law firm that gets covered in the press — agents running autonomously while the owner sleeps — was built one workflow at a time. Firms that try to implement everything simultaneously abandon everything by month three.

Month 1 — Clio and ChatGPT Business

Pick one document type you review repeatedly: NDAs, engagement letters, or a standard agreement type in your practice. Run every new instance through ChatGPT Business with a structured review prompt before you open the document yourself. Note what the AI flags that you would have caught, what it catches that you might have missed, and how long your review step actually takes.

At the end of 30 days, you have a measurable baseline: average review time with AI vs. without, and a refined prompt that reflects your practice standards.

Month 2 — Lawmatics and Lexis+ AI

Add Lawmatics to automate intake and follow-up. Connect it to Clio. The 15-minute setup cost is recovered in the first missed lead you stop missing.

Activate Lexis+ AI for research. Run your next three research tasks through it before you run them through Westlaw or traditional Lexis. Compare the time and output. Most attorneys find AI research is faster on well-defined questions and requires more iteration on complex ones — knowing which is which is the judgment call you develop in month two.

Month 3 — Smith.ai and Spellbook

Smith.ai covers inbound intake. Spellbook adds real-time contract review in your drafting environment. By month three, your practice management, intake, research, and drafting workflows all have an AI layer. What's left is refining: better prompts, faster review, more confident delegation.

Document automation (HotDocs or Contract Express) is the month four or five addition, after you know which documents are worth automating.

The Adoption Gap Is a Decision Problem

Clio's data is unambiguous: the AI is available, the cost is accessible, and 72% of attorneys have already touched it. The 8% who've committed to wide adoption aren't doing anything technically complex. They've made a decision: this workflow, this tool, this month.

The gap is a decision problem disguised as a technology problem. The cost isn't the barrier — $400/month is less than most attorneys spend on continuing legal education annually. The learning curve isn't the barrier — Clio, Lawmatics, and ChatGPT Business are tools you can use on day one. The barrier is the friction of changing a workflow that works well enough, in a practice where you're billing by the matter and every interruption has a cost.

The 90-day rollout is designed specifically for that constraint. One workflow per month. Measure before adding the next. By month three, you have a practice that's structurally faster and structurally lighter than it was in January.

This week's action: Open Clio's feature list and identify one AI feature you haven't used. Run it on the next document of that type that comes through your queue. If you're not on Clio, identify the document type you review most frequently — NDAs, engagement letters, standard agreements — and build one structured ChatGPT Business review prompt for it today. Run your next three documents through it before reviewing manually. That's Month 1, started.

The $400/month stack isn't a commitment to AI as a concept. It's a commitment to getting 90 days faster by May.