What regulated-industry buyers actually need before they'll adopt AI

Selling to law firms is different from selling to most other businesses, and it took me a while to understand why. I want to write down what I have learned so far, partly to organize my own thinking and partly because I do not see this talked about enough.

Law firms are not slow. They are careful. This is the first thing you have to internalize. When a firm takes three months to evaluate your proposal, it is not because they forgot about you. It is because they have a process, and that process exists for good reasons. They deal with client confidentiality, malpractice exposure, and regulatory obligations every day. They are trained to evaluate risk. Expecting them to move fast because your SaaS dashboard looks cool is a fundamental misunderstanding of your audience.

I learned this early at MHSB. Our first few engagements were with firms that came to us through referrals, which meant they already had some baseline trust. When I started doing outbound work, the cycle times were dramatically longer. Not because the firms were uninterested, but because I had not yet earned the right to be trusted with their systems and data.

The real decision-maker is often invisible. In a firm, the managing partner might be the economic buyer, but the actual blocker is usually the IT person, the office manager, or whoever handles vendor compliance. At larger firms, InfoSec has absolute veto power. I have had engagements die because I could not satisfactorily answer a security question about where data would be stored during a migration. That was a fair question. I just was not prepared for it.

Now I lead with the technical details. Before I pitch what the system can do, I explain how data flows through it, where it lives at rest, who has access, and what happens when something goes wrong. This is not natural for most salespeople, but it is essential for this market.

Proof of concept is everything. I have never closed a meaningful engagement without a POC phase. Firms want to see the system working with their data (or a realistic proxy of it) before they commit. The trick is defining clear success criteria upfront. “You will like it” is not a success criterion. “Average intake processing time will drop by X minutes per lead, measured over two weeks, using your real intake volume” is.

The POC phase is also where you build the relationship that sustains the long-term engagement. If you are responsive, transparent about what is not working, and willing to adjust your approach based on what the team is actually telling you, you earn trust that no slide deck can manufacture.

Integration is the product. This is the thing I keep coming back to. Firms do not buy tools in isolation. They buy the outcome of those tools working together. Lawmatics is great, but if it is not connected to Clio or MyCase or whatever practice management system the firm runs, it is just another data island. The integration layer is where MHSB does most of our work, and it is usually where the most value gets created.

The problem is that integration work is hard to sell because it is invisible. Nobody opens their browser and admires a well-built API integration. They just notice that things work. New leads show up in the right pipeline. Follow-up emails go out on time. Conflict checks run automatically. The value is in the absence of friction, which is a tough thing to demo on a sales call.

I have started showing firms their own data flow, drawn out on a whiteboard (or a Miro board, more recently). Here is where a lead enters your system. Here is where it gets stuck. Here is where someone has to manually retype something. Here is where data gets out of sync. This resonates much more than a feature list, because it reflects their daily experience.

Things I am still figuring out. Pricing is hard. Fixed-scope projects are risky because scope always creeps when you are dealing with legacy systems and undocumented customizations. Hourly billing feels misaligned because it punishes efficiency. I have been experimenting with a retainer model (we call it MatterCare) that provides ongoing support, but I am not sure I have the structure right yet.

I also struggle with how much to specialize. Right now MHSB is focused on Lawmatics, which gives us deep expertise but limits our addressable market. I see firms running Clio Grow, or Smokeball, or HubSpot-plus-PracticePanther stacks, and I wonder if we should broaden. But specialization is what makes us good at what we do. I do not have a clear answer.

This is an ongoing thing. I will keep writing about what I learn. If you work with law firms and have figured out things I have not, I would genuinely love to hear about it.