The firms getting real value from AI started with clean processes. Everyone else is just pouring gasoline on a dumpster fire.
I had a conversation with a managing partner last month that stuck with me.
He was frustrated. His firm had spent six figures on AI tools for their marketing department. Content generation, social media scheduling, automated email campaigns, the works. Six months in, his marketing director was more overwhelmed than before. Output had tripled, but so had the complaints from attorneys about unapproved content going live, off-brand messaging, and recycled ideas that weren't connecting with clients.
"The AI is broken," he told me.
The AI wasn't broken, but his content process was. It had been broken for years. The AI just made the brokenness visible. And fast. I worked in professional services marketing and ran a digital marketing agency that served law firms before it was acquired. I've been inside these firms and sat in the meetings where partners wave off process conversations because they want to "get to the AI stuff." The firms winning with AI right now are the ones that did the boring work first.
The insight here is strategic, not technological.
When a firm adopts AI without fixing its underlying processes, something predictable happens. The firm produces more of whatever it was already producing, only faster. If the firm was producing good work through a clean process, AI accelerates that good work. If the firm was producing mediocre work through a chaotic process, AI accelerates the mediocrity.
Speed is only valuable when you're headed in the right direction.
This is like driving. If you're on the right road with a clear map and good brakes, a faster car gets you there sooner. If you're lost, driving faster just means you end up further from where you need to be. And you burn more fuel doing it.
The professional services firms I talk to (law firms, accounting firms, consulting firms, agencies) are often driving fast in the wrong direction. They've bought the faster car (the AI tools). They just haven't bothered to check the map (their processes).
Content creation and approval is probably the most common broken process in professional services firms. And it's the first place most firms try to deploy AI.
The broken version looks like this, and I'd bet money this sounds familiar to anyone who's worked in a law firm marketing department:
An attorney has a vague idea for an article. They mention it in passing to a marketing coordinator. The coordinator drafts something based on a five-minute hallway conversation. The draft sits in someone's inbox for three weeks. It eventually gets routed to a partner who rewrites half of it. Another attorney weighs in with conflicting edits. Nobody is sure which version is current. Six weeks later, a watered-down version gets published that nobody is proud of and that doesn't speak to any specific client pain point.
Now add AI to this process.
The AI generates a first draft in three minutes compared to three days. Wonderful. But that draft still sits in the same inbox for three weeks. It still gets routed through the same tangled approval chain. Partners still rewrite it based on gut instinct without using data about what content actually connects with clients. The only difference is that the firm now has five AI-generated drafts stuck in the same broken pipeline where there used to be one human-generated draft.
You've accelerated the production of content that nobody reads and nobody is proud of. Congratulations.
Now look at what happens when a firm fixes the process first.
A firm that takes process seriously maps out every step of the content lifecycle: ideation, creation, review, editing, approval, publishing, and performance tracking. Each step has a clear owner, a defined timeline, and specific criteria for moving forward. The marketing team knows who approves what, what "approved" actually means, and how long each stage should take.
With that clean process in place, AI becomes extraordinarily useful. During ideation, AI can analyze which topics have performed well historically and identify gaps in the firm's thought leadership coverage. When creating content, AI can produce drafts that adhere to the firm's established brand voice and style guidelines because those guidelines actually exist and are documented. In the review phase, AI can check drafts against the firm's tone standards and flag inconsistencies.
Most importantly, AI can add institutional memory to the process. It can track which articles drove engagement, which topics generated inquiries, and which approaches fell flat. Over time, the system gets smarter. It recommends topics based on actual performance data instead of relying on a partner's hunch about what clients want to hear.
That institutional memory component is something I want to emphasize. Many professional services firms don't know what content works and what doesn't. They publish and move on. AI layered on top of a clean process can tell you, with real data, that your articles about cybersecurity compliance for healthcare companies generate three times the engagement of your general data privacy pieces. That kind of insight changes your entire content strategy.
But only if the underlying process is sound enough to capture and use that data. Garbage process in, garbage data out.
Content creation gets all the attention because it's visible. But there's another broken process hiding in plain sight at most professional services firms, and it might be costing them even more revenue: cross-selling.
Every managing partner I've ever met says the same thing. "We need to do a better job of cross-selling." And every year, cross-selling stays stuck at the same underwhelming levels. The reason: the process is broken. Or more accurately, there is no process.
At most firms, cross-selling works like this. An attorney finishes a matter. They might mention to a colleague that the client could use help with a different issue. That colleague might follow up, or might not. Nobody tracks the conversation. Nobody has a structured way of understanding the full scope of a client's business challenges. Cross-selling is treated as an organic byproduct of good relationships when it should be a disciplined business development function.
Now layer AI on top of that. The result is predictable.
Some firms buy AI tools that analyze client data and flag cross-selling opportunities. The tool surfaces a recommendation: "Client X has exposure in area Y and should be introduced to your regulatory practice." An email goes to the attorney who manages that client. The attorney ignores it because they don't have the context, the relationship mapping, or the conversational framework to act on it. The recommendation dies in an inbox, just like the content drafts.
The AI identified the opportunity correctly. The process failure killed it.
The fixed version looks like this. A firm that takes cross-selling seriously builds a structured client discovery process. Attorneys are trained (and expected) to conduct regular business reviews with key clients. These conversations go beyond the current matter. They ask about what concerns the client most. They explore the client's strategic priorities, competitive pressures, regulatory concerns, and growth plans. The firm documents these insights in a systematic way.
With that process in place, AI becomes incredibly powerful. AI can synthesize information across hundreds of client conversations and identify patterns that no individual attorney could see. It can match a client's stated challenges against the full range of the firm's capabilities. It can even prepare briefing documents for attorneys before client meetings, pulling together recent news about the client's industry, relevant regulatory changes, and the firm's track record in related areas.
AI can also bring something to cross-selling that many firms lack: continuity. People leave and memories fade. Relationship intelligence walks out the door when a partner retires. AI, built on top of a solid process, preserves that knowledge and makes it accessible to the entire firm.
But again, and I cannot stress this enough, the AI is only as good as the process feeding it. If attorneys aren't having those deeper conversations with clients, there's nothing for the AI to analyze. If the firm isn't documenting client needs in a structured way, the AI has nothing to synthesize.
So what does it actually look like to fix your processes before adding AI? I've found that firms that get this right tend to follow a similar path.
First, they map what actually happens. They don't map what the firm handbook says happens or what the managing partner thinks happens. They map what actually happens, day to day, when a piece of content gets created or a client conversation occurs. This is often humbling. I've worked with firms that discovered their "two-week approval cycle" was actually averaging seven weeks. You can't fix what you can't see.
Second, they identify the friction points. They look at where things stall, where handoffs break down, and where people make decisions based on instinct instead of information. In content workflows, the friction point is almost always the approval stage. In cross-selling, it's usually the discovery stage. Attorneys simply aren't gathering enough information about their clients' broader needs.
Third, they simplify before they automate. This is the step many firms skip, and it's the most important one. Before adding any technology, strip the process down to its essential steps. Eliminate redundant approvals, clarify roles and responsibilities, and set realistic timelines. A clean, simple process that humans can execute consistently is the foundation that AI needs.
Fourth, and only fourth, they add AI to specific points in the process. They use AI for ideation, first drafts, consistency checking, performance analysis, and client intelligence synthesis. Each application is targeted, measurable, and built on a process that already works without the AI.
This sequencing is what separates the firms getting value from AI and the firms throwing money at vendors and wondering why nothing improves.
I want to be direct about something. Broken processes don't stem from incompetent marketing directors or business development teams. They stem from firm leadership that has never treated operational processes as worthy of serious attention.
In a typical professional services firm, the practice of law (or accounting, or consulting) gets all the investment and intellectual energy. The business of running the firm (marketing, business development, HR, knowledge management) is treated as overhead. Something to be tolerated and managed down. That attitude was always expensive. In the age of AI, it's becoming untenable.
The firms that will pull ahead over the next few years are the ones where leadership recognizes that operational excellence is a strategic advantage. The managing partner cares as much about the content approval workflow as about the client intake process. Business development is treated as a discipline with real methodology behind it, not as a vague expectation that attorneys will "do some networking."
AI is an amplifier. If you've built something worth amplifying (clean processes, clear accountability, disciplined execution) AI will take you to places you couldn't reach on your own. If you haven't done that work, AI will just make your existing problems louder and more expensive.
If you're reading this and recognizing your own firm, many professional services firms are in the same position. Process improvement doesn't require a massive technology investment or a multi-year transformation initiative.
Pick one process, just one, and map it honestly. Fix the obvious broken parts. Get it running cleanly with your current team and current tools. Then, and only then, start exploring how AI can make it better.
That's the unsexy advice nobody in the AI industry wants to give you, because there's no software license attached to it.
But it's the advice that will actually move your firm forward.
Guy Alvarez is Co-Founder and Managing Partner of InnovAItion Partners, an AI consulting firm that helps professional services firms and marketing agencies integrate AI into their operations. He previously founded Good2bSocial, a digital marketing agency for law firms that was acquired in 2023. Have a process you're trying to fix before adding AI? I'd love to hear what you're working on.