Navigating Legal Tech Marketing: Kurt Dunphy Shares Spellbook’s Journey

Thanks for having me on.

Yeah, sure. Before I was at Spellebook, I was working for a chain of retirement homes. So we were selling beds for seniors, and there for a few years. I joined Spellebook about four years ago when it was Rally, which is a whole other product. We pivoted pretty hard about two years ago to AI and helped us grow pretty aggressively. Yeah, my journey started, I guess I started in finance in college and then went to be a solopreneur doing agency work with very little experience, but learning on the job on building sites, doing ads, doing the whole gamut of marketing, which has been helpful because, for the longest time in Spellebook and Rally, I was a one-man marketing team. We’ve since built it out to three more people. And now, yeah, I run marketing at Spellebook, and we grow leads. The marketing team grows leads, and we’re pretty much solely focused on revenue.

Yeah, we’re up to about 2,600 now clients on spellbook. We’re one of the fastest-growing products in law. Okay, this isn’t a plug. I’m not trying to plug your audience. Yeah, the strategy we’ve done. So we’ve been, I think the biggest The highest level thing is we’ve been super focused on lead gen and driving revenue. Everything we’ve done, we’ve tied it back to the actual performance of things with our attribution, which has been the biggest thing. We haven’t focused on brand building or other things. It’s been purely lead gen into driving sales and then tying that back. If campaigns or strategies don’t work, we just kill them if they’re not driving revenue. We experimented a lot when we started, so we tried pretty much everything, like organic. We tried pretty much everything except for direct mail, which is still not off the table. We might do some postcards and flyers. But the biggest thing for us was we got into some paid media, that worked, we scaled it up, and then we tightened up the funnel so that we had folks coming in as a lead and then getting right past over to sales, which drove up our conversion rate and then moving into a fast sale time.

Yeah, I’m a pretty big-numbers guy. I started in finance, so I didn’t start in finance. I didn’t believe in advertising. I didn’t think it worked. Everything I’ve done, I’ve tied it back to attribution. So we experiment a lot. When we started with Spellbook, we were experimenting with everything from organic social, organic search, Reddit posts, and everything you can do digital marketing, we tried. And then we tied it back to what worked and then basically just killed anything that wasn’t driving leads into meetings, into sales. So it was a big thing was experimenting. We tried I can’t give all the secret sauce away, but we tried weird places that you wouldn’t think lawyers were, and they were. And I guess the secret sauce is super relevant for the audience, but it’s just we tried weird stuff in weird places. And places our investors were Why should you be there? Should you be spending this much advertising dollars in these places? And that was part of the experimentation. We just trying different stuff. I think when we launched Spellebook, we tried 10 to 15 channels. I was looking at our experimentation board yesterday.

That’s a good question. I don’t have a great answer for this. I was talking to a founder of a small company yesterday, and they’re asking me the same thing. I don’t have, Okay, spend two grand, and you’ll know. You get a feeling when you’re doing it a lot. If you try 10 places, and you go in there for two weeks, you have a really… It’s hard to stay because everything’s different. Small companies are going to need a lot more leads if they’re selling $30 a month versus an enterprise that might need one if they go to a conference. But when you do it enough and you experiment enough and you try enough things and you record the process and you feel how bad it feels when things don’t work, you get a good feeling for what does work. And at the 10, usually one works. And you’ll see four to five times the results on that one. And it’ll be easier and leads to come easier. And it won’t feel as gut-wrenching because everything’s not failing. It’s just a feeling. I’ve never really found a hard-set number, but generally, I’ll run an experiment on a platform and give it an honest shot for two weeks.

Sometimes four, and if it doesn’t work, I’ll move something else. I don’t know if that’s short. When you ask this question, do people give it a longer life, or is that short?

Is that a short life cycle? When you ask this question, do other people say two months, or is it two weeks normal?

Yeah, I’m with you on this. It’s multiple things.

I search the MetaSuite and Bing. Bing is so good and cheap, but the volume is so wildly low. It’s so low that we’ve never been able to increase it, but it does work at a wildly good cost.

We were talking before the call, It’s still a new strategy. We haven’t completely cracked it and figured out the playbook. I’d say for us, a lot of what we’ve done is brand searches. We’ve driven that through advertising. And then we show up. Anytime there’s a brand search, our site lights up like a Christmas tree. We’ve got technical SEO done pretty well on the main site. We haven’t done pretty well. If you look up spellbooks, we came up with Magic: The Gathering stuff, which took us a while because initially when we launched, if you looked up spellbooks, I don’t know if Magic: The Gathering is, the card game. You would get magical-related things, and we never came up. It was infuriating. We had to advertise on our main brand. Now we’ve come up within the rankings, so we beat Magic: The Gathering a lot. That was a big win for us, which was just technical SEO and getting backlinks and looking a little more legit. We are doing a lot more. We’re doing a lot more blog posting. We’re just getting involved a little more with that.

But so far, it’s been just driving it from a pushing perspective through advertising.

Oh, yeah. We track everything in HubSpot. We had a really good team called Big Time Data set up our data tracking in Metabase. Shout out to them. They’re really good. So we track everything from every single campaign. We go over it weekly. We go over leads, meetings, and booked, sales are driven every week, and we go through which source is driving each of that. Or SEO is a little tougher because we would have to attribute some cost back to that. That’s not necessarily advertising cost, but a lot of that we do attribute to advertising because we get our biggest thing is brand searching. But yeah, we’re checking every week. We’re analyzing every single thing we do on HubSpot with dashboards there and then a little bit on Metabase. But the HubSpot dashboards are fantastic. If you get your UTMs right, which I’ve gotten them wrong in the past more than once, but I’ve learned my lesson. All right.

Yeah. So it’s funny that you asked this because we just launched it today. Again, I’m not trying to plug your audience here. Probably we have different audiences, but it’s this agent thing. So right now, the biggest thing for the last couple of years was ChatGPT and this generally generative AI, which was the biggest search thing ever. I think OpenAI is the fastest-growing company that’s ever lived in the modern day. But now, I think, I can never know, but I think the next big thing is going to be a genetic AI. Just a quick summary. If you compare it, ChatGPT and just generative AI, like any of them, like perplexity, Anthropicic, any of them are you’re living in a chat window and you’re asking It’s telling you questions and it’s telling you answers. Then you have to, if you wanted to write a blog post, I’m sure you’d use this for blog posts or other things, you take that, you put it where it has to go. It’s okay, it’s great. The output is wildly faster than anything we’ve seen, and it’s pretty accurate. But the problem is you’re jumping between documents, and it’s hard to get tasks done.

With GenTic, it’s going to be able to run programs. You can set up automated things. You get an email, it comes back, you then fix documents, you update it, and it sends a draft, r gets a draft email ready to send to your client. It can just jump between programs. So you break out of that chat window, and you can go between a lot of different things. And if you Google trends this, if you Google trend, generative AI or ChatGPT, you can see when it’s spiked. You can see now, that I’ve been looking every day since we launched Agent, go and look up Agent AI, and you’re going to see a massive spike within the last month or two. It’s starting to spike, so it’s starting to get up there. I think that’s going to be the next big thing. If I was going to put my money on it, it would be that. The one-two punch we’ve seen for SEO with this is when we launched, we were the first generative AI company for law. And when we launched it, there was no SEO because you’re not this way better than I am.

You’re not searching for stuff that you don’t know. If it’s a new product, if it’s a new category, you’re not searching for it. So we did a lot of push advertising for our category or our segment. And when we did that push advertising, what we could see was search followed it like six months, maybe three months behind. So then we pushed that and we captured it on search. I think Because you’re asking about SEO strategy, I think that’ll probably be the SEO strategy again. Do you push it for your niche or your market, and then you capture it on the back end? But you have to push because it takes a while. Man, it takes a while for that to catch on in Google search. The SEO is beautiful, and I’ve had some success in the past in generating hundreds of thousands of impressions. But it’s hard in a new and emerging category before people get educated for it. But if you’re the first one in there, you’re just going to crush. If an agent takes off, we were the first one for Gen AI for law. By the time everything caught up, we had so many pages and posts.

For what type of content?

We don’t have the best process right now. There’s a small team. If we want to set up a new landing page, most of us can work on the Web. We use Webflow, so you can go in, you launch something. We’ll do copy-edits between the team on Google Docs. A lot of times for new ads, we’ll rapid-fire stuff and then just go into Figma. And between two, or three of us, we’ll just build things as fast as humanly possible. We get this for this campaign we just launched. It’s not the perfect best strategy. In probably bigger companies, they would cringe at me. But we just try to build as many things as fast as possible and see what works. And then just… Honestly, the best thing is just putting it live and testing it. It’s hard to know what’s going to work until you get it out there. But we’ll jam on Figma a lot. We’ll jam on Google Docs for headlines, a copy, and that stuff. Building a better process now because eventually, as it scales, it’s not just me. It doesn’t make sense. But we try to get everyone in Figma and not break it apart by silos.

This is your job. This is my job. And we look at it. We all just get in there and try stuff. And whoever’s the best at design will tweak other people’s work to make it look good if that’s what we go with.

About marketing in the legal industry?

That we don’t care about driving revenue, I think. We’re all about brand impressions. I think it’s a pretty big conception overall that marketing is about brand building and stuff. I can’t think of any big ones, to be honest with you. I can’t think of it. Like, lawyers don’t spend a lot of time thinking about marketing, I think. They don’t like that I bombard them with ads, I think.

Yeah, you know what? The biggest, probably, misconception is that marketing has to be boring and business-like, and it has to reach customers through business avenues. I’d say that’s the biggest one. You hear a lot of people We have to reach them where they make business decisions on LinkedIn, and the content should be very business-esque, and it should show all the value props, and it should be professional, ready to print. Some of the best results I’ve had are just going into not business, business places, coming up with really conversational tone advertising, showing one interesting thing, and not making your ad look like a brochure. This might be I’m not trying to call any of our competitors out, but it might be a B2B SaaS thing. It’s definitely in legal tech. B2b SaaS is pretty buttoned up. When you go into legal tech, it’s even more buttoned up. It’s because you’re selling to lawyers and you don’t want to make your product look like a… I don’t know how to say this kindly. You don’t want to make it look like a bad tool that no one wants to invest in. You want to look professional and really, you’re just extra buttoned up.

You have a suit on top of your shirt in the advertising space in legal tech. A lot of it doesn’t work because I’ve tried it. I know for a fact it doesn’t work that well. I’ve talked to people and I know it’s not driving a ton of leads and it’s not driving a ton of meetings because when you see a brochure, when you’re scrolling through, honestly, even when you’re scrolling through LinkedIn, because a lot of people are using LinkedIn in the evening, on the weekend, they’re using it to break up their time, get away from work. They’re not scrolling it. Yeah, I’d love to read a brochure. I don’t know about you but have you ever scrolled LinkedIn, a broch you’re looking at and you’re like, This is great. This is everything I need.

Something close to my heart. One of the big things we did was our legal tech conference conference. I don’t know if there are photos, but there are on my LinkedIn. But it was pretty interesting. We went to two conferences right when generative AI was getting into law, and we just had a whole different vibe. If you look at our booth, it’s wild, and people might say, Okay, so if you go to the conference, it’s piggybacking on what I just said. You go to these legal set conferences, it’s very buttoned up. Everyone is a different shade of blue on their corporate brand. Every vendor is buttoned up. Everything is very rigid. We came in with pink neon lighting blowing up our whole space. It was really strange. We were like, one of the first ones to put AI in all of our banners because the felt like it was just straight-up an AI company. We’re not like, We didn’t take our old product and slap AI on it. It’s an entirely new AI native product. It was pretty wild. We were talking to other vendors, and by the end of it, we were getting more leads and meetings booked a day than most vendors were getting for the It was like a three-and-a-half, four-day conference.

We stood out like a sore thumb or maybe like a good thumb because it worked, but it just stood out like crazy. We ran ads leading into it. We emailed all our customers. We did some press. We sponsored part of the event, and we stood out because everything else was just very typical of B2B SaaS, to be fair. I’m not trying to rip on the industry. It was just very standard stuff. Blue and gray white, and the messaging was perfect, and everything had long words and complicated, like simultaneously updating complex projects, very professional words that you didn’t get. We were just like, Our My tagline was like, draft 10 times faster with AI. People were like, Okay, I get that. Yeah.

It is now.

It is very abused now, and it’s almost impossible to use that word. People are like, Okay, here we go. It’s just not working so well anymore. A year or two ago, it was very good. Yeah.

Yeah. Probably I’m really that I don’t have to do it all myself. I should have expanded the team faster. We could have produced more. I spent years doing everything, from web design to graphic design to videos to screen recordings for product demonstrations to running ads to keyword searches, everything. I could have built up a team and it probably scaled a little faster, to be honest. Probably I would have done that if I went back and hired a few contractors. We have a few now that are just helping us scale everything from design to web to advertising. I probably would have gone back and done a little bit more of that.

Yeah.

Give a man a nail with a hammer. He uses a hammer. Give him anything else. He still uses a hammer. I don’t think it was probably the most efficient thing, but yeah.

Yeah. Agents for Law. It’s live now. Been doing, working on today, working on everything from ads to funnel to attribution. We went pretty big on social, which is cool. I would recommend any company that is launching stuff, just go big on social. We were just plastered there. We brought in a lot of leads today, a few hundred of our ICP, just from the organic stuff because the whole team is starting to get bigger now. So when we do organic stuff, there’s 50 of us that like dog pile onto it. And then we share stuff. We got a bunch of press. We got ads rolling out, probably end of the day. Maybe tomorrow. New landing pages, new videos. It’s a big endeavor. You need it when you launch a new product.

Yeah, do you see it?

Oh, yeah. We should have it up there. I was hoping you saw it. It should be live today. Okay. We’ll do that, too. It’s a good checkbox for us, but it’s not like we’re ICP, our ideal customer profile. They don’t live there, but it is helpful. We’re trying to check off all the proper product launch update checkboxes, and get your email out, and into the press. A lot of it’s going to be tested now. To scale it, it’s going to be figuring out which advertising, which SEO tactics, which content scales it. And then I honestly don’t know yet what it’ll be.

Okay. Yeah, I’m ready.

Find a It was a typo in this blog post. It was a typo.

Right now, I’ve been looking up a lot of if aliens are real or not. Trying to find an answer for that. My YouTube history is a mess.

I’ve seen Postcards work, which I thought… I’ve seen Postcards like Smash, which I thought was weird because I thought postcards were over. And I heard people put them on their fridge and stuff. Just totally smashed. The ROI was like 100 to one. And that was wild. I thought the day of postcards was over.

The weirdest place? The doctor’s office, I think. I don’t I don’t know if that’s normal… That might be a normal place. It’s not a techno doctor’s office. It’s like a walking clinic because we have a low… I’m in a small province. We have a low doctor-to-person ratio, both in the walking clinic.

I like LinkedIn. I know we’re disagreeing on this one. I don’t know if you’re a Twitter guy. I knew LinkedIn. I like LinkedIn. It’s positive. There’s a lot of interesting stuff. It needs to be getting better, too. Less of the flaunting stuff and more really useful content. I saw someone posting on Apple’s history of ads today, and it was awesome. The work people are putting in on LinkedIn is just good. I’m not a Twitter guy myself. I don’t know if you’re on the Twitter Club. All right.

We have to keep some things secret. We’re going to keep some secret thoughts.

Hopefully, it was helpful in the problem.