Your AI Assessment: How to Rescue a $25K Tool and Break the Adoption Cycle at Hartley & Dunn

Why This Matters

Every legal business tells clients they take security seriously -- but most cannot answer basic questions about their own cybersecurity posture when pressed.

Firm purchased an AI legal research tool. Six months later, only 2 of 15 attorneys were using it. $25K/year subscription with single-digit adoption.

This is not an edge case. Businesses in legal face these challenges every day. The question is whether you act before the incident -- or after.

Quick Answer: An AI-first assessment of a 11-50 employees legal business identified 42 hours/week of recoverable capacity, worth $742,560 annually, with 1-2 month payback period. The assessment delivered a prioritized remediation roadmap with specific costs, timelines, and regulatory compliance mapping. Every finding is actionable within 90 days.

Key Takeaways

At a Glance Names anonymized

The Challenge

Here is what we heard: Hartley & Dunn bought a legitimate AI research tool, ran a one-hour webinar, handed out logins, and then watched 13 of 15 attorneys go right back to Westlaw. Six months later, the firm is 75 days from a renewal decision, two senior partners want to cancel, and you are walking into a partners meeting with no plan.

That is the surface problem. The deeper one is this: this is the third time the firm has been here. The practice management platform.

The client intake tool. Now the research tool. Roughly $65,000 spent across three failed rollouts -- same pattern each time, no champion, no structured process, no accountability.

The $25K tool is not the crisis. The pattern is the crisis.

The adoption failure is the symptom. The absence of a change management process is the cause. And until that root problem is fixed, every technology purchase Hartley & Dunn makes -- regardless of how good the tool is -- will end the same way.

What We Found

Here is the connection that matters most right now: you have two attorneys inside your own firm who cracked the code on this tool, developed validated workflows, and are delivering real results -- and that knowledge has never moved an inch.

Sarah Chen figured out that a spot-check on the first three results builds the confidence to trust the other seven. She knows the tool is 92% accurate on major jurisdiction rulings. She volunteered to run a lunch-and-learn three months ago.

Nobody followed up. That is not a technology problem -- that is a knowledge distribution problem sitting right inside your adoption problem, and it costs you nothing to fix.

Think about what that means for the partners meeting in two weeks. You do not need to defend the vendor. You need to walk in with Sarah's workflow d ocumented, a 30-day adoption plan with specific attorneys assigned, and a concrete metric -- say, six attorneys actively using the tool by day 30.

That is a winnable argument. 'Trust us, people will start using it eventually' is not.

The second connection is the one that will sting a little: 80% of those 50-60 weekly research hours across the 13 non-adopters are being billed to clients. Two corporate clients have already grumbled about research line items. When AI cuts that research time by 60-70%, you face a choice -- bill fewer hours and keep clients happy, or redeploy that attorney time to higher-value work that commands the same rate.

Either path is better than the status quo. But you cannot get there while 87% of your firm is running manual workflows on a paid AI subscription.

The attorneys who won't use the AI tool are not being stubborn -- they're being professional. They haven't been shown a validated workflow. When you give a litigator a tool with no validation protocol, you're asking them to bet their bar license on vendor promises. Of course they said no.

Recommendations

1. Execute an Emergency 30-Day Adoption Sprint Before the Partners Meeting

In firms with similar adoption patterns -- an expensive AI tool, single-digit adoption, leadership asking hard questions -- the fix is almost never more vendor training. It is a structured internal knowledge transfer from the one person who actually understood the tool.

You already have Sarah Chen. She has a documented workflow, a validation method, a 92% accuracy benchmark, and she already volunteered to teach it. The move here is simple: give her 90 minutes and a structured format.

Document her exact workflow in a one-page protocol -- specific to Hartley & Dunn's practice areas -- and make it the firm standard. Then run a mandatory two-session lunch-and-learn in weeks one and two. Not optional.

Not a webinar. Attorneys in the room, working a real research question from a current case using the tool while Sarah walks them through it live.

Assign two attorneys from litigation and one from corporate as adoption leads for weeks three and four. Have them report back at a brief Friday check-in -- not to management, to each other. Peer accountability moves faster than top-down mandates in law firms.

Set a target: six attorneys with at least three research sessions logged in the tool before the partners meeting. That is a number you can defend in the room.

AI helps here in a specific way: we can build a lightweight AI-powered onboarding protocol that monitors tool usage across the firm, sends automated weekly prompts to non-users with a specific use case relevant to their practice area, and surfaces adoption metrics to you each Monday morning. No manual tracking, no nagging emails from the office manager. The system does it.

Setup is typically two to three weeks and runs on existing infrastructure.

2. Install a Change Management Framework So You Never Pay This Tuition Again

Three tools. Three failed rollouts. $65,000 in wasted spend.

The common thread is not the vendors -- it is the absence of any internal process for introducing technology to attorneys who are already billing at capacity and have zero incentive to change their workflow unless someone makes the case directly to them.

The framework we recommend for firms your size has three components. First, every technology purchase above $5,000 gets a designated internal champion before the contract is signed -- not after. That person participates in vendor evaluation, runs the internal rollout, and is accountable for adoption metrics at 30, 60, and 90 days.

Second, no tool goes live without a documented use case specific to each practice group. 'Here is your login' is not a rollout. Third, adoption metrics are reviewed at the monthly partners meeting alongside billable hours -- not buried in an IT report nobody reads.

This is not complicated. It is the kind of process that takes one afternoon to design and prevents the next $25,000 mistake. We can help you build it, but you could also build it yourself starting this week.

The point is: the firm needs a stan dard, and right now there is none.

Where AI accelerates this: we build automated onboarding sequences for new tools -- practice-area-specific prompts, use-case tutorials, milestone check-ins -- that run without anyone managing them. When the next tool gets purchased, the system handles the rollout infrastructure. First-time setup runs $5,000-$10,000 and pays for itself the first time it prevents a repeat of the current situation.

3. Quantify What Manual Research Is Actually Costing the Firm -- and Make That the Business Case

Here is a number worth putting in front of the partners: 65-70 hours per week of attorney time is going to manual research across your 13 non-adopters. At a blended $340/hour rate for that group, that is $22,100-$23,800 per week in attorney time spent on work the tool was purchased to handle.

The AI research tool -- used properly, the way Sarah uses it -- cuts research time by 60-70% based on her own experience (6 hours to 90 minutes on the citation task). Apply that to the full 65 hours per week: you recover 39-45 hours of attorney time weekly. At $340/hour, that is $13,300-$15,300 per week in capacity that can be redirected to billable work, case strategy, or client development.

Annually, that is $690,000-$795,000 in recovered attorney capacity -- from a $25,000 tool. Even if you apply a conservative 40% realization rate (not all recovered time becomes new billable work), you are looking at $276,000-$318,000 in incremental revenue potential per year. The tool is not the problem.

The ROI on full adoption is extraordinary. The partners meeting needs that number on the table.

An AI-powered vCAIO engagement -- starting at $2,500/month -- can build the metrics infrastructure to track this in real time: research hours logged, tool usage by attorney, time-to-completion comparisons, and monthly ROI reports for the partners. That is the difference between defending a spend and proving it.

4. Address the Trust Problem Directly -- It Is a Quality Control Issue, Not a Training Issue

The 13 non-adopters are not lazy. They are cautious. They have heard stories about AI hallucinating case citations, and in a profession where citing a nonexistent case can result in sanctions, sanctions get taken seriously.

The fear is rational. The solution is not more training -- it is a validation protocol that makes the risk explicitly manageable.

Sarah Chen already built that protocol: feed in the specific legal question, request the top 10 cases from the last five years in the relevant jurisdiction, spot-check the top three on Westlaw. Two weeks of that and she had 92% confidence in the output. That is not a secret -- it just has never been written down or shared.

Write it down. Make it the firm's official AI research protocol. Have your managing partner -- you -- put your name on it.

That changes the psychology. It is no longer 'the vendor says trust it.' It is 'our firm has validated this workflow and here is the process we use.'

Beyond that, the ABA has been increasingly clear on attorney supervision obligations when AI is used in legal research -- Model Rule 5.1 and 5.3 apply directly here. A documented validation protocol is not just good practice, it is your ethical compliance posture. Firms that have written AI use policies in place are significantly better positioned when bar ethics opinions land on this topic, and those opinions are coming.

A brief AI governance policy document -- covering acceptable use, validation requirements, and supervision standards -- should be part of the rollout package. We build these as part of our vCAIO engagements and as standalone AI governance deliverables.

5. Build the Firm's AI Foundation -- Before the Next Tool Purchase Creates the Same Problem

The pattern in firms that break the failed-adoption cycle is consistent. Once they have mapped their process gaps and built their AI governance framework, they did not just fix the current tool -- they built the internal infrastructure to evaluate, onboard, and scale any AI tool going forward. Industry benchmarks show firms with structured AI governance achieve 80%+ adoption rates and frequently eliminate redundant vendor relationships within 12 months.

Hartley & Dunn is at the decision point where most firms either course-correct or repeat the cycle. The pattern is clear: without a designated AI leader, a change management process, and a validation framework, the next $25K tool will go the same way.

A vCAIO retainer -- starting at $2,500/month -- gives you fractional AI leadership: someone who owns the AI roadmap, manages vendor relationships, drives adoption, and reports to the partners on ROI. For a firm your size, that is significantly less than the cost of a full-time IT director, and far more specialized than anything a generalist IT hire would bring. The engagement typically pays for itself within 60-90 days in recovered tool ROI alone -- before counting any new capabilities the AI strategy unlocks.

This is also where the assessment you just received becomes relevant: you described your challenges, AI analyzed the data, and you received a personalized strategic plan in minutes -- with specific numbers from your own firm. We build this exact capability for law firms. A client-facing intake system that qualifies leads, captures matter details, and delivers a structured brief to the responsible attorney before the first consultation call.

That is AI working for your firm around the clock, not sitting unused on a subscription invoice.

ROI Analysis

Your numbers, run straight:

You told us 65-70 hours per week of research time across the 13 non-adopters, at a blended $340/hour rate. Sarah Chen's experience shows the tool reduces research time by approximately 65% (6 hours to 90 minutes on a citation-heavy task). Apply that to 67 hours per week:

• Hours recovered per week: ~42 hours
• Weekly attorney capacity recovered: 42 hrs -- $340/hr = $14,280/week
• Annual attorney capacity recovered: $742,560/year
• You are already paying for the tool: $25,000/year subscription already in budget
• Net annual value at full adoption: ~$717,560 in recovered capacity -- before any revenue realization

Apply a conservative 40% realization rate (not all recovered hours become new billable work in year one): $287,000 in incremental revenue potential, Year 1.

Implementation investment: $7,500-$25,000 for the adoption sprint, AI governance framework, and vCAIO retainer engagement. Payback period: 3-6 weeks at 40% realization. Under 2 weeks at full realization.

Multi-year projection (conservative, 40% realization):

• Year 1: $287,000 recovered revenue minus $25,000 implementation = $262,000 net
• Year 2: System is running, adoption is 12+ attorneys, AI governance handles new tools automatically. Realization rate climbs to 60%: ~$430,000
• Year 3: Full adoption, AI handling citation checking, brief summarization, and research threading. 70% realization: ~$500,000+

Cost of doing nothing for 12 months: $742,560 in attorney capacity consumed by manual research. Plus the $25,000 subscription either canceled (sunk cost confirmed) or renewed (same failure pattern). Plus the next technology purchase that follows the same adoption arc.

The three-tool pattern has already cost the firm $65,000 in wasted subscriptions. Without a change management process, the next 12 months add to that total.

The partners meeting framing: This is not a question of whether to cancel a $25K tool. It is a question of whether to recover $287,000 in year-one revenue potential from a tool already paid for -- or write off the spend and repeat the cycle on the next purchase.

Hartley & Dunn didn't waste $25,000 on a bad AI tool. They wasted $25,000 on a good AI tool with no adoption plan. The difference matters -- because the tool is still fixable. The pattern that created the problem is what needs to change.

Implementation Roadmap

Phase 1: Quick Win (Weeks 1-2)

Goal: Walk into the partners meeting with proof, not promises.

Week one: Sit down with Sarah Chen for 90 minutes. Document her exact workflow as a one-page firm protocol -- specific to litigation research, corporate research, and real estate research. Get Mike Torres's citation-checking workflow on paper too.

These two documents cost nothing and are the most valuable thing the firm can produce right now.

Schedule the first lunch-and-learn for end of week one. Mandatory for all 15 attorneys. Sarah runs it live on a real research question from a current case.

Not a vendor demo -- a colleague showing how she does it on work the room recognizes.

Week two: Assign three adoption leads (two litigation, one corporate). Each commits to logging five research sessions in the tool before the partners meeting. Pull the usage dash board and bring the numbers -- sessions logged, time comparisons, attorney names -- to the partners meeting.

That is a plan with a scoreboard, not a promise.

Phase 2: Foundation (Weeks 3-8)

Goal: Build the infrastructure that makes adoption permanent and prevents the next failure.

Weeks three and four: Formalize the change management process. Every technology purchase above $5,000 gets a champion assigned before the contract is signed. Build the one-page evaluation template: designated champion, practice-area use cases, 30/60/90-day adoption metrics, partners meeting review cadence.

This takes one working session to design.

Weeks four through six: Draft the firm's AI Acceptable Use Policy. Cover attorney supervision obligations under ABA Model Rules 5.1 and 5.3, validation requirements for AI-generated research, data handling for client matter confidentiality, and citation verification standards. This is your ethical compliance posture -- and it makes every future AI tool deployment cleaner.

Weeks six through eight: Begin the vCAIO engagement. Map the firm's current research workflows end-to-end. Identify the next two highest-ROI AI opportunities (likely: brief drafting assistance and client intake automation).

Build the metrics dashboard -- research hours logged by attorney, tool usage rates, time-to-completion comparisons -- so the partners see ROI numbers monthly, not anecdotally.

Phase 3: Strategic (Months 3-6)

Goal: Move from fixing the current problem to building compounding competitive advantage.

Months three and four: With the research tool at 80%+ adoption, turn attention to the practice management platform gap. Apply the new change management process to evaluate whether the canceled platform should be replaced and what a proper rollout looks like. The firm now has t he infrastructure to do this right.

Months four through six: Expand the AI footprint. Brief drafting assistance for litigation (AI drafts the structure, attorney refines the argument -- typical time savings of 40-50% on first drafts). Client intake automation -- AI captures matter details, conflict checks, and initial case assessment before the first consultation, freeing attorney time for the work that requires judgment.

Document automation for corporate and real estate (NDAs, standard agreements, due diligence checklists).

By month six, the firm has a documented AI strategy, a track record of successful adoption, a validated governance framework, and an AI system that learns Hartley & Dunn's specific practice patterns over time. That is the difference between a firm that bought a tool and a firm that built a capability.

How AI Helps

AI transforms legal operations by automating the work that consumes the most hours and creates the most risk.

Here is what AI specifically changes for a 11-50 employees legal business:

• Assessment speed: AI-first cyber audits deliver findings in 5-10 business days instead of 4-8 weeks. The assessment you just read was powered by AI analysis.
• Compliance documentation: AI generates policies, gap analyses, and remediation roadmaps that would take a consultant weeks to produce manually.
• Continuous monitoring: After remediation, AI continuously monitors for new gaps, policy violations, and compliance drift -- eliminating the "audit and forget" cycle.
• Cost reduction: AI-first methodology delivers the same depth as traditional assessments at a fraction of the cost. That is why the audit is $2,500-$5,000 instead of $25,000-$75,000.

The assessment you just read is itself a demonstration. You described your situation, AI analyzed it, and you received a specific, personalized plan with your actual numbers, your actual deadlines, and your actual regulatory exposure. That is what AI-first looks like.

Terms and Definitions

Frequently Asked Questions

How much does an AI cybersecurity assessment cost?

Just In Time AI charges $2,500 for businesses with 50 or fewer employees and $5,000 for businesses with up to 500 employees. Traditional assessments cost $10,000-$75,000 and take 4-8 weeks.

How long does an AI-powered assessment take?

AI-first assessments deliver findings in 5-10 business days. Traditional consulting engagements take 4-8 weeks for similar depth.

What frameworks does the assessment cover?

The assessment maps findings against NIST Cybersecurity Framework, CIS Controls v8, and industry-specific regulations. Gap analysis includes specific control numbers and remediation priorities.

Is remediation included in the assessment cost?

No. The assessment identifies and prioritizes gaps. Remediation is a separate engagement scoped from the findings. The $2,500 audit fee is credited toward implementation if you engage within 30 days.

What happens after the assessment?

You receive a 50-page assessment report, gap analysis, and prioritized remediation roadmap. We walk through findings together and scope next steps based on your risk tolerance and budget.

Do I need to prepare anything before the assessment?

Existing policies, insurance agreements, org chart, and any current plans (BCP, DR, IR). If anything is unavailable, those gaps become findings in the assessment.

Ready to Get Started?

You have seen what an AI-first assessment looks like. Now imagine having that same analysis applied to your actual environment -- your real systems, your real compliance gaps, your real dollar exposure.

The AI-First Cyber Audit from Just In Time AI costs $2,500 for businesses with 50 or fewer employees and $5,000 for businesses with up to 500 employees.

Dan Stolts | Just In Time AI
Based on a real assessment scenario. Details anonymized.