Legal

AI costs and ROI for law firms

AI costs money. The question is not whether it is expensive. It is whether the numbers add up, and most law firm boards are not being given enough information to decide.

By Zakir Hoosen, Director, Fuzzelogic Solutions. Board-level guidance in plain English.

A managing partner is told an AI tool will save two thousand hours a year across the conveyancing team. The number is on a slide. It looks impressive. But no one has said what the tool costs, what the implementation costs, or how many of those two thousand hours are actually billable. The board approves the spend. Twelve months later, the hours saved have not translated into revenue saved, and no one can explain why.

This is the AI cost problem in law firms. The numbers that get presented are the numbers that look good. The numbers that matter, the total cost, the real savings, and the payback period, are often missing.

What AI actually costs

The licence fee is the visible cost. It is usually the smallest one.

The real cost of AI in a law firm includes the time to get the data ready, the time to configure the tool, the time to train the people who will use it, the ongoing support, and the management time to keep it running. Add those together and the total is usually three to five times the licence fee, sometimes more.

Boards that rush tend to approve on the visible cost and discover the hidden ones later. That is not a technology failure. It is a planning failure.

What it actually saves

The savings side is where the numbers get vague. An AI tool that reviews contracts might save thirty minutes per contract. But how many contracts does the firm review? At what fee rate? And would those hours have been spent on contract review, or on something else that also generates revenue?

The honest answer is that AI saves time on specific tasks. Whether that time translates into money depends on what happens with it. If the lawyer uses the freed time to take on more client work, there is a real revenue benefit. If the lawyer uses it to do the same work in fewer hours, the benefit is to the client, not the firm, unless the firm can reprice.

For law firms, the savings calculation also has to account for the cost of errors. If an AI tool produces a contract review that misses a key clause, the cost is not the thirty minutes saved. It is the potential negligence claim, the insurance premium increase, and the reputational damage.

The 10-20-70 rule

Fuzzelogic uses a framework that comes from BCG's research on where AI value actually comes from.

Ten percent is the technology. Twenty percent is the data and the model. Seventy percent is people changing the way they work. For a law firm, that means the real cost is not the software. It is the management time, the process redesign, and the training. The firms that count only the licence fee are counting one tenth of the bill.

The questions a board should ask

When the AI spend is presented, ask for three numbers.

First, the total cost over three years, including implementation, training, support, and management time. If the number is only the licence fee, ask for the rest.

Second, the real savings. Not the theoretical hours saved. The actual pounds saved or earned, with a clear line from the time saving to the financial outcome. If the savings are presented as percentages, ask for the starting point.

Third, the payback period. How long before the firm has spent more than it has saved? If the answer is more than eighteen months, the board should be cautious. AI moves fast. An investment that pays back in three years may be obsolete before it pays back.

Where law firms waste money

The biggest waste is the pilot that never scales. A team tries an AI tool, it shows some promise, the pilot runs for six months, and then nothing happens. The tool sits there, the licence fee continues, and next year someone proposes a new pilot with a different tool.

The second biggest waste is buying for the demo. The vendor shows a polished presentation. The output looks impressive. But the demo is not the firm's data, not the firm's workflows, and not the firm's quality standards. What works in a demo often does not work in practice without significant configuration.

The third waste is not counting the cost of inaction. Every year a firm delays on AI is a year where competitors are getting faster, cheaper, and more responsive. That cost is real, even if it does not appear on a balance sheet.

The honest version

Fuzzelogic does not sell AI tools. We help law firms work out what AI would actually cost, what it would actually save, and whether the numbers make sense for the specific firm. Sometimes the answer is yes. Sometimes the answer is not yet. Sometimes the answer is no. The honest answer, given early, is always cheaper than the expensive lesson.

Your systems were built for a world before AI. Most can get there. We tell you which ones cannot.

Start with the assessment. Two to four weeks, fixed price, and you own the verdict and the roadmap whether or not we build any of it. When you are ready to talk AI, call Fuzzelogic Solutions and ask for Zak. www.FuzzelogicSolutions.com | info@FuzzelogicSolutions.com | +44 (0)1624 618950