Assessment Before AI: Why Law Firms Are Stepping Back

Collin McKee

Welcome back to the Endeavors AI Podcast. I'm your host, Collin McKee, founder of Endeavors AI, where we talk about the intersection of artificial intelligence, law, business, and the new economy being built around all of it. Law firms are usually the last to chase a trend. With AI, a lot of them grabbed every shiny tool they could find, and now they're quietly stepping back. My guest today is the person they call when that happens. Viveca Hess is a former lawyer herself who spent more than a decade helping attorneys grow their practices on LinkedIn, then went all in on AI. Today she runs a deep assessment for law firms before anyone touches an agent, and she'll turn work down if the firm wants to skip that step. If you're tempted to just implement AI, listening to this one first would be a benefit. Viveca, welcome to the show.

Viveca Hess

Thank you. Thanks for having me. I'm excited to be here. I've been listening to your podcast for a while, and you have some really good ones, so I hope I can live up to it.

Collin McKee

I'm sure you will. Great to have you on. I wanted to start with your story, because it's kind of an unusual path. You grew up in Sweden, you were on a junior ski team, and you came to California planning to stay for one year. What happened?

Viveca Hess

Well, actually, I was planning on staying for the long haul. I'd had enough of the cold and darkness in Sweden, so I was ready for California. I came over and was surprised that, because I hadn't been living in California, I had to pay out-of-state tuition to continue college. My idea was to continue my university studies here. So I thought, I'll take a year off, establish my in-state residency, and avoid the much higher tuition. It seemed a natural fit to teach skiing for that year. That was so much fun it turned into three years. After that I applied to UCSD, political science, and then law school right after that. And here I am, still in beautiful, sunny Southern California. I do go back to Sweden to visit every summer. Toward the end of law school I dove into international securities regulation and got published, and I got a job offer with the San Diego County Water Authority, working with the general counsel. From there I went into estate planning, and because of my background and language skills I ended up working with international clients on the estate planning and corporate side. Then I realized I very much enjoyed the personal interactions, the marketing, and the psychology of marketing, especially with people you haven't had a chance to meet face to face. After quite a few years I backed out and said, I'm going to give this marketing thing a go. I started a marketing agency and had absolutely no clue what I was doing, because law school didn't teach us that. It was a steep learning curve, but a fun one that I don't regret in any way. It's been evolving, and it hasn't been a straight trajectory. In the last two years I've gone from being known as the LinkedIn lawyer lady to really diving into AI as a support tool for growing business and strategizing with lawyers and law firms. There's a lot of fear and trepidation and FOMO around AI right now, but also immense opportunities. It's a matter of what order we do things in to get the best return for each law firm.

Collin McKee

Was there a particular inflection point that drove you away from the law firm and into marketing for firms?

Viveca Hess

There actually was. One of the partners was almost like a father figure, a mentor, and he was pulling back and getting ready to retire. I'd had my little girl at the time, and I thought, what's a way I can have the flexibility, without the hard billable hours and expectations and working through the weekends? So a few things came together: my mentor pulling away, and conversations with him where I said I enjoyed the marketing, the psychology of marketing, and business growth much more than the legal research I was doing. There was a confluence of those events, and I pulled back. Like I said, it took a while to figure out the whole marketing thing, and I'm still learning. We're always learning, especially with AI. Anybody who comes to me now and says they're an AI expert, I personally think that's a red flag, because nobody's an expert. It's changing every day. None of us are experts, and we might as well fess up. We're learning the best we can and assimilating as much information as we can to serve our audience in the best way.

Collin McKee

There might be a small handful out there that applies to, but there are way too many claiming it. To your point, it's progressing way too fast for anyone to say they know it all. So you started to pivot to AI, and last year you went all the way in. You said it felt scary to let go of the label, the LinkedIn lawyer lady. Why was that?

Viveca Hess

It was something I was comfortable with, and that people associated me with. If I went to a conference or was speaking somewhere, there was a certain recognition and level of authority I got used to. So it was a little scary deviating from that. Our identity is closely tied to our work, and it's hard to pull that apart, especially if we're passionate about it. There was a sense of loss of identity for a while, and then I had to get out of my own way and see that I'm still implementing everything I've learned over the 15 years since I left the law track. It just has a different layer on top of it. And I think you and I both see that any professional, in any industry, who hasn't figured out that AI or automations or workflows or agents, at whatever level they need, are going to play a role is going to struggle. For law firms specifically, if they haven't figured out that they need to adapt and adopt in some way, I have to say, small to large firms are going to be out of business. That sounds a little drastic, but I really think they need to figure out how they're going to use it in their firm, and there are all sorts of ways to do so. There's no one-size-fits-all template. That's my angle, and I find a lot of vendors implementing one-size-fits-all too much. We have to be very sensitive to the different needs of solo practitioners through big law. Everyone has a different set of needs, so we need to listen very carefully.

Collin McKee

I'd agree a hundred percent. So how did you gain credibility in the world of AI? How did you get credible fast?

Viveca Hess

Well, I didn't. You're doing it for me. No, I think it's just being really honest about it. I conducted a number of beta tests with kind attorneys who have been in the private membership I've had for four years. I said, can you hop on a call, can we do a full assessment and diagnostic? I learned as much as I could from about twenty use cases, where I transcribed the responses and listened carefully, not to what I thought they needed, but to how they expressed what they needed. That was a learning curve, and I'm so happy I took the time. It gave me the ability to speak directly to my audience in their words, to their pain points and their aspirations, and I think that quickly made it sound like I might know what I'm talking about.

Collin McKee

I'm sure the whole of your background adds to your credibility, but that's a very good point. You're not trying to sell something to someone. You're having a conversation about what they need.

Viveca Hess

I do try to emphasize that, and I think it removes a certain level of suspicion, because everybody has some trepidation about this whole world of AI. I say that I'm tool agnostic and technology neutral, so I'm coming in to listen and talk and find out what's the best fit. My mantra is always strategy first, the foundation first, tools second, and humans in the loop every step of the way. This isn't a tool or technology or era that's going to take over entirely. You continue doing what you're best at, and this is just a tool. It's a formidable tool if you use it the right way, but it is just that. That's where, as we spoke about last time, I'm finally seeing a turnaround. Early on, law firms, fearful of missing out on the AI trend, grabbed the shiny, sexy tools as quickly as they could, or whichever vendor came in and spoke well. When I asked these attorneys and firms why that tool, or why those technologies that could cost quite a bit and take quite a bit of time and training, many of them didn't know how to answer. So I felt I had to take on the less-than-exciting task of saying, let's step back and do an assessment of your particular firm: your culture, practice area, your history. Then we'll have a value map and a process map to know exactly which tools fit, because we have thousands of tools now. We have choices, but let's dig in and find what suits your firm, style and culture. Culture especially is a big thing, and if anything sets me apart from many other AI consultants, it's that emphasis on culture. If I interview anywhere from five to eight key stakeholders, maybe partners, and two of them are gung-ho and say let's roll this out, but the others don't buy in, and you've got staff and younger associates who aren't buying in either, then you'll have continued politics or tensions within the firm. And suddenly you have the problem of shadow AI, where everybody ends up doing their own thing.

Collin McKee

Unpack that a little bit for those who don't know the term, because it's an important one to know.

Viveca Hess

Shadow AI is very common, and it's not necessarily a bad thing. It's gotten a bad rap. It's what happens when there's no overall policy or guidelines or guardrails to get folks on the same page. Part of what I try to do is help firms reach a common understanding and buy-in on what's going to work for that firm. When you don't have that, each individual, especially young associates who are trying to impress the partners if they're on a partner track, ends up doing their own thing with a different LLM or a different harness. That leads to a lot of risk, because everybody has different guardrails for how much they'll use a harness like ChatGPT or Claude or Perplexity. Nobody's on the same page, and that can lead to a huge amount of risk. One of the things that's been surfacing is that law firm insurance carriers are now asking law firms, how are you using AI? Guess what? They have no solid answer. And guess what the insurance carrier might do? Okay, we're not renewing. That's huge. Your firm has to have an insurance carrier for malpractice and for many other reasons. And guess who else is asking? Clients. Clients are asking, how are you billing us, why are you billing us this way, how are you using AI, are you doing what you normally would do in half the time? You have to have answers to those things. Unless you've done a pretty thorough assessment and have everyone who needs to be on the same page, so you can answer those questions and have the ability to not only survive but thrive within two years or so, you'll struggle. So, to answer your question in a long-winded way, shadow AI is basically when everybody is doing their own thing with their own harness, meaning their own AI tool like ChatGPT or Claude, and you get different responses and different sets of risks colliding with each other.

Collin McKee

I've got a previous episode, and the title escapes me, but it's something to the point of avoidance is not a policy. To your point, AI is going to get used. You need to set up the guardrails on how it can be used and which tools, rather than everyone using their own, for all the reasons you just outlined. Without going into the specifics or giving away your special sauce, what are some of the basic steps of your assessment?

Viveca Hess

The assessment is twofold. The first part assesses where the gaps, bottlenecks and opportunities are. The second part takes those results into a second tier where we evaluate them, and I develop a roadmap for the firm, like a ninety-day roadmap, for how they can implement AI. Sometimes it's not even a matter of AI. Agents and going agentic are so trendy right now, but there are two questions I can ask somebody who claims they have fifty agents doing everything in their firm, and I'll know right away they have no clue what they're talking about. Sometimes it's not a matter of building out agents or a whole command center at all. That can come later. Sometimes it's just a matter of an automation. Especially in a firm that's been in business for a while, people have certain habits ingrained and don't really think about them. My assessment interviews a number of stakeholders in the firm, then I synthesize that in a few proprietary ways, using some of what I learned in my certifications with MIT and others, and put together a key findings report. I don't think there's been a single time where they weren't completely surprised. They thought there was an issue or a bottleneck or a gap without understanding what was really going on behind it. Once that surfaces, you can tackle it and minimize the risk of continuing blindly. What that has practically resulted in is, conservatively, saving four to five hours a week for many partners I've worked with. If you think of their hourly fees, at least five hundred to a thousand an hour, we're looking at ten to twenty thousand a month in billable time. That makes my fees look like rounding errors. Not every hour saved would otherwise be a billable hour, but there's a whole other side to the value of saving time. Maybe you finally get to take half of Friday off and see your kid or grandchild play baseball after all these years. That has huge non-monetary value too. It's very satisfying work for me. Once the bottlenecks and gaps surface properly, I take those results and put them into a map, another configuration I synthesize afterward depending on the outcome, and then I walk through it with them. Every time, they've been completely blown away by what surfaced, and that's very satisfying. I'm very passionate about it.

Collin McKee

And you say you'll even walk away from business that could be very lucrative if they want to skip that step. What do you tell them?

Viveca Hess

Absolutely. Usually I sense quickly whether we're going to hit it off and be a good fit. They have to be receptive and open-minded. They're not going to hire me just to impose their older ways on AI. And if we have some fun along the way, even better, then we're probably a good fit. But a few times I've been approached by firms who want to hire me, and I've learned to detect this more quickly now, to do the interviews with certain staff to justify a reason to get rid of people and downscale their staff. I just won't be used that way. If you have good people working for you, then by all means, let's use everything AI or automations can do to enhance their position and let them grow, as staff or associates or partners at whatever level. But if you're going to use me to ask questions so you can replace somebody, that's not my business model at all, and if I detect it, I'll back out. They say, well, don't you think we can scale back? Do we really need somebody in HR, if it's a bigger firm, or what about getting rid of the paralegals? And I say, are these people you value, that you chose for a certain reason? Let's enhance their ability. Let's make them superstars. But don't use me to try to eliminate positions and save money that way. What we're seeing now is a backlash. Without naming names, there have been big firms that laid off large numbers of staff and associates, and they're now suffering because they're realizing AI or automations can't do what those people did. AI will never have their judgment, their experience, their taste. That's why AI is just a tool, but a formidable one if we use it to enhance human beings doing their jobs.

Collin McKee

I like the phrasing you used on our first call: pay now or pay dearly later. I think that's pretty appropriate if you just throw AI at it.

Viveca Hess

You can hop on the shiny tool syndrome, which some still do, and you'll pay dearly for it later, and quite a bit more. And in the worst case, you'll find yourself out of business, outpaced by smaller firms that are able to use AI intelligently, effectively and strategically.

Collin McKee

Let's keep moving. We're getting close on time, but I did want to touch on a very important point: the AI policy, or AI SOP. You said there are three different kinds of firms when it comes to AI use policies. Can you walk me through them? The first is no policy, obviously. That's the scary one.

Viveca Hess

Yes, thank you. No policy is no good. You'll run into problems with clients and insurance carriers, and probably many other areas. Then there are firms that say yes, we checked the box, we have a policy, and it's great. That's good. Then I ask a few more questions: are you evolving it as AI, or whatever technology you're using, changes? Is it dynamic enough to reflect those changes? Because this is certainly not a one-and-done deal. If we've ever seen a policy that needs regular updates, it's now, with this whole era moving us so quickly. Sometimes it's a policy only two or three partners are aware of, and they think they're responding to the insurance carriers or whatever powers are demanding it. But everybody has to understand what those policies and guardrails are. In that case I usually recommend being dynamic enough to at least re-evaluate it every quarter. I think anything longer, even six months, is too long. We're in a different time, with the way things are moving, and it needs to be a very dynamic policy, not a one-and-done. And then there are the firms, like the ones you and I have worked with, that are very dynamic. They're doing all the right things and re-evaluating sometimes monthly, or at least quarterly. They have their finger on the gauge, they know how to communicate it, they hold regular meetings on what they're changing and updating, and it gives them opportunities to take on new technology or tools as they evolve. It's a good way to check in.

Collin McKee

I'd agree with that cadence too. You definitely pull it in with any news or new regulations that come down the pike, but with the speed this technology is changing, it doesn't hurt to get more creative thoughts on how you can use it, or learn from mistakes that have happened. I agree with all of it. You and I are very simpatico on this. I like to keep these under half an hour, so let me ask a couple of final questions. What's one thing every firm could or should do this week before buying another AI tool?

Viveca Hess

Well, of course, they should call me. We should chat. But besides that, I think they should look at some of their patterns and workloads. Look at the things you do day to day on a repetitive basis. That's one really easy, quick win for a firm: what do we take for granted that actually adds up, time-wise, throughout the week? Go back and evaluate that, and ask ChatGPT or Claude or Perplexity, or even Grok, how it could be made more efficient. Take the time. I know that's valuable time for them, but have each partner or associate take one small slice of their day-to-day, look through their work patterns for something repetitive, analyze it a little, and see if you can shave off some time or do it more efficiently.

Collin McKee

It does take time, but you get it back.

Viveca Hess

It does, but it pays off well.

Collin McKee

Threefold. This is always a tough one to answer, but where do you see AI in two years? Is every firm going to have three or four agents working the front desk? What are your thoughts?

Viveca Hess

I'm going to give a very lawyerly answer: it depends. That's what every law professor told me when I wanted a pat answer. It really does depend on the size of the firm, the practice area, their goals, whether it's a legacy firm or a startup firm. But if we take a snapshot right now, I think the smaller firms, closer to the startup period, have an amazing opportunity and ability to compete with some of the big dogs, if they're doing it right, because they're nimble enough. They may not have all the compliance hoops to go through to make a change, so they can adapt swiftly. I think we're going to be using agents quite a bit more. But again, if you have a really solid foundation and strategy, and you keep your policy and your goals and aspirations dynamic enough, I think you can adapt. Command centers in various configurations depending on practice area, yes, but we're definitely going to see second brains. What I'm doing right now, and it sounds funny, is building brains for partners. Sometimes we'll give the agent a name, a position and a title, but it's basically mirroring all of their knowledge so they have more time for their sweet spot, and yet they can dip into specialized knowledge. Everybody's using general knowledge from AI now, but I think we're heading toward specialized knowledge: your brain, your experience, your history, and another partner's. Everybody, including my husband, is using ChatGPT instead of Google, and that's where we are for the most part now. But specialized knowledge is going to show up more and more, and we're going to build out brains for our agents more and more this coming year. Being any more precise than that would be a bit misleading or disingenuous.

Collin McKee

Very fair, quality answer. That's another thing that takes time, and just like our own brains, it's not set-it-and-forget-it. You still need to nurture it.

Viveca Hess

You need to feed it.

Collin McKee

I can't say I'm satisfied with where mine is. I still need to feed it.

Viveca Hess

Steve Jobs always said, "Stay hungry, stay curious."

Collin McKee

Exactly. Lastly, where can people find out more about you and follow your work?

Viveca Hess

Not surprisingly, I'm on LinkedIn: Viveca Hess. I'd love for anyone to connect with me. What I usually offer is a fifteen-minute brainstorming session, complimentary. It's an opportunity for me to give you a quick win. You tell me what you're stuck with or what you're considering, and we have a live session for fifteen minutes with absolutely no pitch, no spam, no push. It's just giving you some value, so that if you ever need more help or support, I'm here.

Collin McKee

That's the point of the podcast. I really appreciate it, and I'm sure our audience has gained some value from this. If there's anyone out there, please do reach out to Viveca. She's got a lot of knowledge she's willing to share, and hopefully she can add some additional value. And if you have questions about your own business and feel like you're lagging behind on AI and all of these automations and tools, reach out at endeavorsai.com. We'd love to be of value to you as well. Viveca Hess, thank you again. We really appreciate it. I'm Collin McKee, this is the Endeavors AI Podcast, and we will see you next time.

Viveca Hess

Thanks so much, Collin. Thanks, everybody.