The Best of Office Hours

Episode Notes

Transcript

Conrad (00:00) The trade-off should never be cost effectiveness. It should never be experience, right? The trade-off should never be a bad experience on the one-yard line. On the one-yard line, we are not looking for a ten percent cost savings. That should never be the rationalization. You should never compromise for a ten percent cost savings.

Gyi (00:54) Welcome to Lunch Hour Legal Marketing. I’m Gyi from AttorneySync, and I won my first World Cup pool.

Conrad (01:04) Wow — you won your… does that mean I have to buy you tickets to the Mockingbird AttorneySync Lunch Hour Legal Marketing Extravaganza this coming week?

Gyi (01:17) Well, it’s very interesting, because per the terms of the contest, employees of Lunch Hour Legal Marketing are not eligible. So I started to reflect — am I an employee?

Conrad (01:26) Are you an employee? Do you draw a salary from Legal Talk Network?

Gyi (01:30) I don’t get a salary, so I don’t know.

Conrad (01:32) You holding out on me?

Gyi (01:33) I don’t know what my relationship is there, but I’m not going to ask you to pay for my ticket. Half my ticket.

Conrad (01:43) Maybe I’ll buy you a beer. And I’m Conrad Saam from Mockingbird, and I believe that everyone should splurge and buy expensive pasta.

Gyi (01:58) Ooh — did you just buy some expensive pasta?

Conrad (02:00) I always buy expensive pasta, and here’s why: you can buy the highest-end pasta and still only spend like two dollars more than buying off-the-rack Barilla garbage pasta. And it makes your meal that much better. The bang for your pasta dollar goes so far when you’re splurging on expensive pasta.

Gyi (02:23) This episode not brought to you by Barilla. In addition to pasta and my victory lap, what else are we talking about on this episode of Lunch Hour Legal Marketing?

Conrad (02:35) So Gyi, you and I get together every Thursday at one o’clock Eastern, ten o’clock Pacific — we host Office Hours and answer any question sent our way. Sometimes we comb message boards for questions, sometimes we make stuff up, sometimes I ask you a question — but it’s really Q&A based. And we’ve found that our best shows are the ones where we’re answering questions. So, without any further ado — three excerpts from our most recent three episodes of Office Hours.

Conrad (03:11) We are gonna pander to Lawmatics. Is Lawmatics a sponsor, Gyi?

Gyi (03:17) Lawmatics is a sponsor.

Conrad (03:20) And so, are you an investor, Gyi?

Gyi (03:23) I’m an investor, and I would like to alert the FTC and everybody paying attention that we have disclosed that relationship. This is from July 9th: Lawmatics unveils Merlin, a suite of AI features designed to help law firms grow. As we mentioned, Lawmatics is a sponsor, and I am an investor and advisor. I actually had the opportunity — Matt Spiegel showed some demos before this launch, and I was like, this is pretty freaking cool. Let’s go into a couple of the specific things they talk about here. Copilot: a conversational user interface that creates automations, generates reports, and surfaces insights from firm data through plain-language prompts. That sounds good, right? You’re querying your firm data right in the platform.

Merlin Qualify: an AI agent that evaluates and prioritizes incoming leads based on fit, urgency, and engagement signals defined by the firm.

Conrad (05:03) That’s actually — that’s not new from Lawmatics, right?

Gyi (05:08) That’s right — they released that earlier this year. Merlin Qualify has been recognized with an Excellence in AI award from Business Intelligence Group. So that’s not new, but I think it’s more part of their Merlin ecosystem now.

Conrad (05:30) Okay, but let me stick on this for a moment. This is an absolute no-brainer use of AI — qualification of leads and sentiment analysis in general. Sentiment analysis from intake all the way through client representation is a very powerful use case for AI. Very powerful. If you think about the value of identifying the best opportunities and having a different workflow to respond to them — no-brainer for improving how your firm does. Secondly, if you have an automated way of understanding that you don’t have enough information to evaluate whether a lead is good or bad, that can spur you to actually go get that information — of which Lawmatics does both. Again, we’re pandering grossly here, but both of those things — regardless of platform — are obvious, valuable, automatable uses of your data through AI.

Gyi (06:50) All right — and Merlin Engage: agentic AI for dynamically engaging with prospects through SMS, email, website chat, and phone calls. Currently in beta. So essentially agentic AI engaging via communication platforms like text, email, website, chat. What are your thoughts?

Conrad (07:20) Well, let’s get deeper into this. This is basically an automated follow-up sequence through agentic AI, correct? That’s what I’m understanding.

Gyi (07:34) Well, you’re contextualizing it as a follow-up sequence — presumably people will use this as a frontline intake tool. Let’s start there: someone lands on your website, and agentic AI is talking on behalf of the firm.

Conrad (07:52) So, I lean on another one of our sponsors, Ted DeBettencourt (spelling unverified — please confirm), and I’d tend to — this, to me, Gyi, is one of those things where I want to watch others bleed before I bleed myself. And here’s why: you’ve done all this work, and this is where the biggest mistakes happen. You ask people generically in the marketing world what they’d fix in their digital marketing, and to a T, anyone who knows what they’re talking about says, “I would fix intake.” So this is putting AI on the one-yard line to put the ball in the end zone at the end of the game — and you’d better have nailed it to make this work. I would not want to — and by the way, pandering to Matt Spiegel and Lawmatics all along — this is one of those places where I want to see other people bleed before I jump all the way in. I’ll go back to Ted, who uses human-based chat, not AI chat. Not “we’re going to save some money on our chatbot” — we are going to have the most expensive chatbot, but it’s going to be really, really effective, which in fact makes it the least expensive chatbot out there. I think that’s the way I’d look at this at the very end of the system. Is this foolproof? I don’t know. Are there ways to improve it? I don’t know. But we talk internally about the difference between cutting edge and bleeding edge — great to be on the cutting edge. This is not a place I want to be on the bleeding edge.

Gyi (09:56) Yeah, I think for me, I largely agree. You questioned — is it foolproof? Absolutely not foolproof. Neither are human beings, though. So I always approach this through a trade-offs lens. You and I were actually — even though they’re not a sponsor — on the Answering Legal podcast talking to our friends over there, and they asked us: what’s the ideal intake setup? My answer was a lawyer trained in intake with a high level of empathy who can scale — if you’re doing scale, they can answer the phone 24/7. And of course everyone rolls their eyes — that’s not realistic, but it’s the best experience from a consumer standpoint. The other end of the spectrum — we’ve laughed about this too — historically it was voicemail, but it could be a human intake person who answers the phone having a bad day. That’s not even a neutral experience, it’s a liability. And then you’ve got all the mess in between. So I’m always telling folks — I know “it depends” is annoying, but it does depend on which trade-offs you’re willing to accept. If you’re in personal injury doing front-line intake, you have a very short window. If you’re going to put any kind of technology on the front line, you need to be able to filter certain intakes and escalate them to human beings, to lawyers, as a priority — or you’re going to lose out. Now, some people will say, “I’m okay accepting that I’ll lose out on some of those clients because of the efficiencies gained from the technology-first solution.” But there are consequences.

Conrad (12:11) This is where I’d take exception. You talked about trade-offs — the trade-off should never be cost effectiveness. It should never be a bad experience on the one-yard line. On the one-yard line, we are not looking for a ten percent cost savings. That should never be the rationalization. You should never compromise for a ten percent cost savings.

Gyi (12:41) Yeah. So then — what are the contexts where it makes more sense? Certainly on the chase, the follow-up, the nurture — that’s a no-brainer, because you’ve already established rapport, already started building the relationship. You can set expectations, because you already have that rapport: “Hey, part of our system is going to be this chatbot — if you’re ever having a bad experience, we want to hear from you, you can still reach out to a human being.” But set those expectations, be transparent about it. And I think you’ll find it’s continually getting better in more contexts every day.

Conrad (13:29) I had an interesting conversation the other day with someone — I think it’s on an earlier pod, go back three or four episodes from today — about this lead-scoring concept, really big cases, and the importance of not having an aggressive automated follow-up sequence, because that’s the best way to lose them. They were talking about the importance of dialing back follow-up for the larger-value matters — specifically in the PI world, but the analogy probably holds more broadly. Being much less aggressive for the larger-value matters — I thought that was counterintuitive. I think this is Kerri Coby White who said this.

Gyi (14:15) Well — you’re using the term “aggressiveness.” You’re really talking about frequency of follow-up — number of touches, multi-channel — we’re going to call you, text you, email you, advertise the heck out of you…

Conrad (14:37) Yeah, it was an interesting conversation. This goes back to the root of the issue — and it was mentioned in the comments too — on the big cases, you’ve got to get a human on the phone quickly. I agree. The issue is: when do you figure out it’s this big case, or this type of case that needs a different workflow? That’s the trick.

Gyi (15:06) Well, watch me pander to Lawmatics again — this is the value of lead scoring. I think it’s one of the most obvious, underutilized use cases for AI. But let’s talk about your example — are you presuming a human being was the first touchpoint, or Lawmatics was the first touchpoint?

[Note: the next several exchanges involve overlapping/interrupted speech in the source audio — attribution and timing below are best-guess.]

Conrad (15:36) I’m not presuming anything.

Gyi (15:38) I’m presuming that matters.

Conrad (15:40) It matters because — well, in a perfect world, whether it’s a human, a form fill, or a chatbot, Lawmatics can look at that input data and say, “this is a great opportunity,” or “this is a garbage opportunity,” or “we don’t know” — regardless of what that first contact was.

Gyi (16:07) That’s the challenge, though — if you’re interfacing with technology on the front end, there’s a certain number of clients you might lose just because they were engaged with by a bot.

Conrad (16:26) Regardless of whether Lawmatics figures it out or not — that’s part of it. I’m just recognizing there are trade-offs. I think the technology is amazing — I’ve seen it. We talked to the Haskins brothers too — you’re familiar with their Speed AI product.

Gyi (16:50) Yep. The reason Speed AI exists is because of this exact issue, right?

Conrad (16:57) Yep — it’s because of misunderstood intake data, dropped intake balls.

Gyi (17:02) Yep. Can you quickly explain what Speed AI does, and why we’re using this example?

Conrad (17:09) Speed AI does it after the fact. So regardless of whether the intake was completed by an in-person receptionist, a virtual receptionist, a human being, or some other technology — Speed AI analyzes the intake data, essentially the call transcript. I believe — you can check me if I’m wrong — but in the system, if the intake person marks something as “no case,” it’ll review that for whether it believes there actually might be a case worth following back up on.

Gyi (17:44) Yeah, and it’ll alert you: “Hey, go call Mary, because you guys dropped the ball — let’s go try and save this thing.”

Conrad (17:54) And here’s the thing people don’t believe me when I tell them — regardless of whether it’s an in-house receptionist, a virtual receptionist, or technology — you’ll see someone call in with a qualified case, and the intake person says, “There’s no case,” or “we don’t service that kind of case, that’s not our type of client.” And if the law firm owner saw that, they’d say, “This is terrible.”

Gyi (18:30) And by the way, people think this doesn’t happen a lot — they don’t realize it does.

Conrad (18:36) I’m glad you’re sharing this, because this happens all the time. And I don’t mean to besmirch third-party vendors, but it is worse — I can’t prove this from a data perspective, only anecdotally — it is worse when you have a third-party answering service. It varies by service, and even varies within a service based on the reps you get. And obviously you’re comparing that against what you’re paying for an internal person — on average, maybe you’re right, but I’ll tell you, there’s a lot of bad in-house intake going on too.

Gyi (19:16) A hundred percent. But this happens a lot, and I’ll tell you — so yes, automate the process of figuring this out. With Speed AI specifically, there are a bunch of other services doing this now too — a lot of sentiment analysis happening, a lot of analysis of intake calls. Speed AI was on it quickly — they made something that was vaporware to start, then actually backed it up with something interesting, and now a lot of other software has caught up. But the concept of understanding that we’re missing the boat on some of our intakes — I still believe the job of the COO or the director of intake, ideally using technology to make this better, but someone whose job it is to manage the intake people to minimize these drops — that has to be part of your business.

Conrad (20:16) Yeah. Don’t take our word for it — take it from a real law firm owner.

Gyi (20:22) Anyway, as many of you super fans know, Lunch Hour Legal Marketing does have a Slack channel — if you’re not in it, reach out and we’ll get you added. But the reason I bring this up is because our good friend Tim from Iowa, aka Tim Semlroth, posted a question. Okay, here we go: “We filmed each of our lawyers — 17 different lawyers — answering the same five common questions personal injury prospects have. Each answer is approximately a minute long. We did this so we could feature them on each attorney bio page of the new website. They turned out really well, and I was pleasantly surprised to see how each attorney answered the same question in their own voice, in a slightly different way. In reviewing them, I realized these FAQ videos have the potential to be extremely useful if AI starts digesting them. Any suggestions for how we should deploy them on our YouTube channel so we get the most bang for the buck?”

Conrad (22:15) So, framing — we’ve got lawyer videos, FAQ style. When I first read this, I was thinking about a post from long ago on the Moz blog — I think from some of the Wistia folks — about considerations in video hosting: whether you host on YouTube or natively on your site, SEO, all the historical stuff that goes into that. This is similar, but — because the question is specific about how we should deploy them on the YouTube channel — he’s not talking about deploying anywhere else. That’s my first question. So maybe the second thing we talk about is things beyond just YouTube he should be thinking about. But he’s talking about his YouTube channel and bang for his buck, and it sounds like it’s in the context of AI.

Now, the first thing I’ll say: YouTube, owned by Google — it’s very clear from all the Google I/O stuff, and even what we’re seeing anecdotally and what all the platforms are publishing as research, that Google’s AI loves YouTube. That’s probably one of the reasons Tim’s thinking he wants maximum bang for his buck from posting on YouTube in the context of AI. The 101 answer: make sure you’re titling, filling out a description, and the captions are correct — whether auto-generated or uploaded, make sure they’re accurate. Other than the basic blocking and tackling from an AI perspective, the AI kind of handles the rest, in my opinion. You just have to make sure — if you put the wrong transcript in there, you might have problems. If you mention a name wrong, or say something factually untrue, that could be an issue. Or if the closed-caption transcript doesn’t match what’s actually being said, that could be an issue too. But getting the title, some tagging, the description filled out — those are the basics I’d start with. Google’s ingesting YouTube content — there’s not much more to do there with a published video, since we’re talking about videos that have already been created, not video content strategy. What do you got?

Gyi (24:37) Well, a couple things — I was scribbling talking points while you were talking, looking things up. Number one: biggest bang for your buck doesn’t mean limiting yourself to a single channel, especially when you’re creating content. So definitionally, in his question, he’s limiting his own biggest bang for the buck by talking specifically about YouTube. I think that’s a really important thing to note.

The other thing — there was an interesting post or email I read in the last 48 hours from Darren Shaw that made me think about pricing information, and then made me think: what is the content that AI wants to know about you? What does Google AI specifically want to know? And how can I get that content everywhere I possibly can? They actually tell you — this is direct from Google, from when they announced they were listening in on LSA calls, telling you what they’re listening for and how they’ll use it. Let me read this, because it’s important: “New provisions allowing Google to select, modify, display, and use rich content, including but not limited to photos, provider bios, service descriptions, special offerings, pricing information, and discounts.” I highlighted provider bios and pricing information — they’re telling you what they’re listening for. This isn’t going to help Tim, but if your name isn’t Tim and you’re not in Iowa, this might help you.

They want to know more about provider bios. And one of the hardest things for some lawyers — I’ll even say most lawyers — is talking about themselves in a superlative fashion. There are some painfully annoying people who can’t help that, JD or not, but it’s a really awkward thing to talk through your provider bio, your LinkedIn, your resume. “I don’t want to be on film… I’m not going to sit here and say, well, my 3.9 GPA at the University of Michigan Business School, and I ran the Silver Star for being an amazing digital marketer.” You don’t want to say that stuff — and having that be one-way, it feels to me there’s room in feeding the AI. We talked about “feed the machine” at the Lunch Hour Legal Marketing Summit coming up August 11th through 13th — get your tickets now, we’ve almost sold out the room. We did a “feed the machine” session at the last one. And feeding the machine your provider bio — how do you do that with video? I think having a two-way conversation with an interviewer that talks about why you — going back to Tim’s office, Tim and each of these 17 attorneys talking about why they became a lawyer — but doing that with a deliberate context of feeding the machine the provider bio information.

Conrad (27:43) Yeah, absolutely — 100% agree with all that. I’m really trying to think about meeting Tim where he is — he’s got these videos. The other thing I’d be thinking about — and I haven’t seen the videos, I pulled up the YouTube page so we have some context — assuming they’re more raw, longer-form bio FAQs, I would chop them into Shorts.

Gyi (28:10) Well, he said they’re a minute each, so they’re already short.

Conrad (28:17) Yeah, they’re already short — I might even cut some 30-second ones for YouTube.

Gyi (28:25) And would you post that as a regular Short?

Conrad (28:30) Yeah, as a Short. Let’s see if we’ve got any Shorts here. I’m telling you, Shorts get more display.

Gyi (28:39) They do — because no one wants to…

Conrad (28:43) Yeah — again, I’m going to do what I was trying not to do, which is give advice that won’t help Tim but will help the next person: YouTube has its own — any YouTube expert will tell you there are all sorts of things you can do to improve your videos’ performance on YouTube. You can go look this up — thumbnails are extremely important, there’s all sorts of stuff that goes into this, which I’m not going to go deep on because it’s accessible to anyone who wants to learn about YouTube creation. But I think that’s a piece that’s missed — and frankly, Lunch Hour Legal Marketing, our own process — I feel like we talk about this in our show prep — is being more intentional about designing content specific to YouTube. And this is the trade-off. Just for context, because a lot of people are dealing with the same thing: Lunch Hour Legal Marketing was a podcast, we started filming it, so we have video, and we take clips and make social clips out of the video content — but we don’t actually mindfully do the YouTube-specific stuff. For example, instead of just taking a clip, you might do “three things about RSH Legal that are unique” — designed for the social media consumer, platform-specific. That’s what I think is the missing piece for a lot of the stuff I see on YouTube.

Gyi (30:36) Because we always come prepared with some questions we’ve seen in other contexts — today’s question we’re going to ask you, Conrad, is actually a timeless one: how do I get more online reviews? We talked about how Top Dog generates all these reviews. Let’s do this three different ways. Let’s do — I’m brand new, I just set up my Google Business Profile, talk to me about how to get online reviews. Then talk to me about — if I’ve got scale, machine behind me, infrastructure, marketing budget, intake, staff, I’m a heavy marketer — what kind of ideas do you have to get more online reviews for that segment as well? You said three, that was two.

Conrad (31:46) And then you can come up with your own third. Okay — a couple things. The brand-new firm: it’s very important to get an understanding of what you need and what you’re going to do with your reviews. I think you need to very simplistically break this into two components.

Number one: are you going to be able to compete in local search, period, end of sentence? By that I mean — if you look at the firms within the addressable market who are proximal to your office, can you catch up to those guys? And if the answer is no, your objective with reviews is not to generate business from Google Local, because you’re never going to catch up. Okay? So you first need to answer that question. And if that’s the case, you might think, “Okay, fine, Conrad — reviews don’t matter then.” That is not true. The second objective, which may actually be more important — it depends on what game you can play — is to answer the vetting question: “I was just referred to Gyi’s law firm, and I want to learn a little about them.” If I go on Google and there are six reviews, I’m not really going to learn enough about them. If I go to ChatGPT and Gyi has no reviews on Yelp — which is now being heavily featured in ChatGPT — I don’t know anything more about them. So you need to break this into a vetting-versus-new-business conversation in your head. And most people fail to do this, because most of you fallaciously believe you can compete, and most of you can’t, given where your office currently is. Sorry — don’t shoot the messenger.

And then your next question was scale. So this is actually my answer to both profiles: the answer is volume. That’s why it’s so hard for the new solo to play the acquisition game. But you really need to recognize that the primary success — whether I like it or not — in the review game is having a huge amount of volume and a process to turn that volume into reviews. Which is massively unfair to small solos, massively unfair to anyone opening up an office — but that’s a fact. So when I think about volume — the cousin of volume — it’s not just necessarily volume of business you do, it’s the volume of people your firm interacts with outside of “I hired this lawyer.” Because by Google’s guidelines, people who interact with your firm can leave a review about your firm — that doesn’t necessarily mean they retained you for a brain injury matter, it means they interacted with your firm. So expand your definition of where those reviews can come from, to expand the denominator of volume.

Gyi (35:21) That’s right — that’s the tactical takeaway here. Most firms think reviews come from happy clients singing your praises, so they get into things like, “I’m going to create a system and process to ask for reviews after I hand the client a check or something positive happens” — and that’s definitely one part of the review game you should be thinking about. But to Conrad’s point, you’ve got to get creative. Let’s face it — especially in PI, but really in most major direct-to-consumer practice areas, in any reasonably sized market, it’s going to be competitive — you’re going to be in hundreds, if not thousands, of reviews. So how do you do that if you’re a smaller firm just getting started, without hundreds or thousands of clients to leave reviews? You’ve got to start thinking: what can I do to motivate somebody to leave a review? It still comes down to experiences, right — that’s the guideline, anybody who’s had an experience with your firm. Let’s talk experience, intake experience — even if it’s a turndown, if you find a way — maybe you have more time on your hands because you have fewer clients — going the extra mile to really help someone find the help they need, even if it isn’t you, is a way to motivate a review.

Speaking-wise — if you’re doing educational content, seminars, webinars, or in-person events, don’t forget to include, “if you found this session helpful, go leave a review.” I’ve heard of lawyers getting even more creative than that — some partner with a local business, maybe host something at the local pizza shop, whether it’s a focus group or an informational event about accident prevention and safety tips. Then have folks at the in-person event — make it very frictionless for them to leave a review. Now, people will say this crosses the line against Google’s guidelines — we’ve talked about kiosks, and there are specific guidelines to pay attention to — but at the end of the day, in most competitive landscapes, you have to get beyond thinking happy clients will just leave you a review at the end of representation, because you’re just not going to compete.

Conrad (38:06) Not going to compete. Yep. What else?

Gyi (38:10) I think the other thing that’s interesting — you mentioned the Yelp deal with ChatGPT — how are you talking to clients about diversifying review profiles on other sites? Historically we used to say, “if you don’t have a competitive number of reviews on Google, just focus there.” But to your point, now with AI, are you more diversified, or still mostly Google?

Conrad (38:37) 100%. So this all lives in the context of competitiveness. One of the things you could choose to do is lose the Google game and win the Yelp game with your 80 reviews. That’s a great strategy — we just talked about how hard it is for the solo firm. Fine — get all your reviews on Yelp. Now, Yelp sucks, because it’s really difficult to get reviews published — you can’t even ask.

Gyi (39:15) You can’t even ask — what’s that guy’s name? Jeremy — I can’t remember his last name — but he has some asinine perspective about how it’s unethical to ask for reviews. Yelp has this whole utterly stupid policy about asking for and getting reviews. But I’m using this as an example — go win a thing you can win. Don’t try to win the Google game if you’re never going to win it — take your 80 reviews and put them somewhere you can actually win. I think that’s really good strategy.

Conrad (39:57) There’s also — and I think this is increasingly real — back in the day we used to have this multi-channel approach. I used to encourage lawyers to get their reviews written on Avvo, because it was delightfully self-serving for Avvo. People were showing up on Avvo and using Avvo reviews — there were more lawyer reviews on Avvo than on Google. Google Business Profile wasn’t really a thing back in the day — that’s a relatively new addition to Google, this whole profile-about-a-business living within Google search results. I don’t know the exact answer, but it’s less than eight years old, I think.

Gyi (41:07) Well, they were like, “Avvo’s eating our lunch over here, we should just flip a switch and put reviews on Google” — and that’s what happened. And right after that became a thing, guess who dropped out of the game? Avvo sold to Internet Brands. Don’t tell me those two things aren’t closely related. I have no knowledge that they were, but — there’s a great Australianism, “Blind Freddy could see that.” Anyway, where was I going with that — I don’t know. Win where you can win, be strategic about where you put your reviews. Increasingly, we’ve gone from a place where there’s reviews everywhere, to “how do we get the reviews everywhere to show up on Google,” and now we’re going back to the reviews-everywhere perspective because of the importance of this be-everywhere model.

Conrad (42:22) And we talked about this the other day — we should do a whole segment on this — the abandonment of the hub-and-spoke model. The hub-and-spoke model had your website at the center, and the goal was to have everything else support getting traffic to your website. That is gone. And that also held true in the review game — the objective was to get people to your site.