GE 251: Guillaume Cabane Reveals How 20% of Drift’s Traffic Will Turn into 80% of Their Revenue Down the Road (podcast) With Guillaume Cabane

Guillaume Cabane, the Vice President of Growth at Drift

Hi, everyone! Today I share the mic with Guillaume Cabane, Vice President of Growth at Drift, a conversational marketing platform that helps you turn your traffic into qualified meetings using their BOTS, 24 hours, seven days a week.

Tune in to hear why a one-size-fits-all approach doesn’t work, how AI is affecting this type of marketing, how pitching specific ideas can double or triple your conversion rate, and why he tells potential customers their lead score as a way to draw them in.

Download podcast transcript [PDF] here: Guillaume Cabane Reveals How 20% of Drift’s Traffic Will Turn into 80% of Their Revenue Down the Road PDF

Time-Stamped Show Notes:

  • [00:32] Before we jump into today’s interview, please leave a review and rating and subscribe to the Growth Everywhere Podcast!
  • [01:36] Guillaume started with a website at 15 years old; he published content.
  • [01:55] In college, he moved into online B2B sales.
  • [02:51] He spent the past 5 years in B2B companies, eventually leading him to drift.
  • [03:25] Drift’s mission is to help companies have more relevant conversations with their customers.
  • [05:50] Drift’s support tools help them get ahead of their competitors.
  • [07:00] Drift started with live chat and has since added email to their services.
  • [08:00] Drift uses Clearbit to help them with their framework.
  • [10:15] Drift tells potential customers their lead score as a way to draw them in.
  • [10:38] 20% of Drift’s traffic will turn into 80% of their revenue down the road.
  • [11:08] “One size fits all can only fit so many.” – Guillaume
  • [12:05] Drift’s competitors are often SAS products that require javascript to work properly.
  • [13:15] Drift uses Intellivise for website optimization.
  • [14:12] Clearbit tells you who is visiting your site and MadKudu tells you information about your visitors.
  • [15:12] Guillaume likes to see who is visiting his competitors’ sites and then he makes a point to contact those people.
  • [16:18] Guillaume loves to test new tools. If people pitch him, he will try it out.
  • [16:40] Guillaume is now involved in an Eastern European job board site, because he spent time with the product and thought it was viable.
  • [17:55] It costs a lot of money to use other tools to run Drift (Clearbit, MadKudu, etc.).
  • [18:00] It probably runs $200,000/year, but their revenue is around $4 Million.
  • [18:52] It’s expensive, but he believes the costs to be worth the outcome.
  • [19:12] It’s preferable to paying for traffic via AdWords.
  • [19:18] Guillaume believes it is up to him to maximize revenue.
  • [19:58] Average customer value is in the hundreds of monthly revenue.
  • [21:00] Segment is about 200 employees, 50 sales reps, and they generate 50% of the growth opportunities every month.
  • [22:30] If a company comes to your site, you should be able to see what their field is and be able to pitch them.
  • [23:44] You double or triple your conversion rate by pitching specific ideas.
  • [24:22] Single Grain uses the account-based marketing feature on Drift.
  • [25:01] If you reach out to Drift, they have consultants that will help you set up a marketing strategy.
  • [26:06] The market is moving into more AI.
  • [28:45] We are not far from removing the human element from this type of marketing.
  • [29:10] Guillaume recommends Influence, which is a great book about psychology.
  • [29:45] He has a podcast called “Seeking Wisdom”, which is about marketers telling their stories. He also likes the “SAASter” podcast.

Resources From This Interview:

Leave Some Feedback:

Connect with Eric Siu:

Disclaimer: As with any digital marketing campaign, your individual results may vary.

Full Transcript of The Episode

Show transcript
Guillaume C.: Yeah, so, in terms of the markets, people know Drift, where we started, which was live chats. And yes we have some competition there, but most of the competition we have in live chat are mostly south of the sign up. Right? And its more like support tools. If you look at all, that will help you generate more demand from your site. And that's critical cause your site is what is intent. Its what people, they come to your site [inaudible 00:06:16] most the market is B2B and I think the listeners here are on sites which are B2B kind of, right?

Eric SiU: Yup.

Guillaume C.: And the thing is, your traffic comes to your site, because they're paying but it's got a fund. They think, maybe your product can answer it. You're not Reddit. You're not FaceBook. People don't come to your site just because they two hours to waste. That's not how it works. They are gonna stay on your site for 30 seconds, maybe less. And so if you think in terms of how do we capture that intent? How do we make sure that within that traffic, the right people will capture the maximum. There's not that many companies that help you do that and we're one of them. And we're probably the best one. To help you do that, so yes, we started on the live chat as a way to help you convert more of that high quality traffic that's been on your website, capturing it, with playbooks of information with being able to chose the right message for each company, and now we're venturing into email to do the same thing. Helping the sales rep be more efficient through higher quality, not through more volume.
That's really the game.

Eric SiU: Got it. You know how ... one of the big things I remember reading about you is, you had this crazy case study where you, let's jut talk about this work flow you built that unlocked "Extraordinary growth for segment." Can you speak to that?

Guillaume C.: There's a few, I'd say frameworks that are built like that. I think there's two which I can speak to and both are relevant for discussion. Both actually hinder the same thing, which is the ability to reveal the company that a visitor works for, based on the IP address, and so I use a third party provider called Corbit and their API called Reveal. Which when you hit that API with an IP address it returns the domain of the company and the company attributes. And so basically that visitor comes from and that tells you a lot, because then you can use our scoin models, our predictive scoin models based on all those attributes on the company size review they have, the amount of employees, the market, the technology to see whether they are a good fit for your business. And based on that, we did two things. The first thing is the visitors for high score companies who are leaving our site without interacting with us without signing up or doing a demo request, we would re-target them by email.
And that's unlocked a ton of growth, because they are the right people because the lead score we use that's built by Madkudu. And they have the right intent because they came to your site, so that's something right? It's not a cold email. And on those emails, we have 15 to 20 percent response rates. Which is insane.

Eric SiU: Wow.

Guillaume C.: And then, we thought, hey that's good to re-target them by email, but it's not the best experience ever because they still have to come to your site and then leave the site and then get an email, why not interact while the intent is at it's peak, right on the site. And so when I was at segment, I was a customer of Drift and the purpose was to hack shift so we could compute the IP to domain, compute the score and then ping Drift within a few seconds to force the stable live chat and say, hey we know that you have the right fit, we should talk. You know? Very direct, very honest and that was insane. Cause people say what do you mean I have the right score and how do you know of abc. And that's just the ice beaker. That's the ice breaker that enables us to create the conversation, right there.

Eric SiU: So you tell them their lead score, so it's your lead score is 150, is that what you do?

Guillaume C.: We tell them it's very good. We just use low, medium, good, and very good. And basically what we do is target the people who [inaudible 00:10:31] you gotta see the score from Madkudu, the company that provides the lead score for us. They are able to predict within 20 percent of our traffic. The visitors, the companies will generate about 80 percent of our revenue. One month from now, alright? So it's very predictive, so engaging those people is critical. And let say what's critical is that your website is your one size fits all. You can have large companies, small companies, high pit, low pit, high intent, low intent, coming to the same website, which has the same message and what one size fits all can only fit so many.
If you have, if you have Delta is a good example. Delta more than 100,000 employees, 10,000 people in the IT department, they come to they're going to see the public pricing, which goes from 99 dollars, to 400, 500 dollars. And they're going to think, okay, this is like a mid market solution, it's not for us.

Eric SiU: Right.

Guillaume C.: And they're going to leave. Now I've just lost a hug enterprise deal. What if I could interact with them before they got there and say, hey, yes we also have an enterprise optioning. We can talk about it. We will white log your un boarding. We'll take care of your come down. Don't go on the product. What I discovered is that we also started putting changes on the website to prevent those large scales from going in the product, because their experience is not going to be good. It won't. I mean, in my case, passing on to a drift. These are sass products that require a java script to be installed on your site. That's how it works. Do you think that the person that comes from Delta is going to install the java script?

Eric SiU: Nope.

Guillaume C.: No, of course not. They've got a lot of security review to go through. Which, if they learned the product, they won't be able to do anything. Nothing. They can't even connect their email if they don't use Gmail. So they can't do anything, which means if you let them run the product, you're failing on the user experience. Which means you have to stop them before and you have to hand them over to a very good sales rep. Who's going to build a plantation and explain to them the features without needing to learn the product. And that I have seen improve conversion rates and of course, the average customer by a lot. Prevent the high score, high market companies to learn your product.

Eric SiU: Got it. So you talked about the conversion rates going up and the customer lifetime value going up. Can you talk about what percent you see them increasing by? Just so the audience gets an idea.

Guillaume C.: Sure. So preventing the people from seeing the pricing, which is one of the things we did, we actually created a variation, which means that we will call and televise, which does website optimization based on the date I talked about before. And one of the variations was hiding the pricing page from large prospective customers, large enterprise. That doubled the contacts capture rates, our rate of capturing the email from about two percent to four percent. Which is huge in that case. And it has enormous impacts in terms of revenue.

Eric SiU: Got it. Okay. This is the stack we are talking about right now, so we have Drift, we have Madkudu and then you talked about intelligent, is that the one?

Guillaume C.: Yeah, I talked about [inaudible 00:14:11] usually [inaudible 00:14:11] sits at the top and tell you who's going to your sites, then you have Madkudu who tell you whether that company is a company you should sell to or not, whether you have a high likelihood of closing them, then you have tools that are action tools, like Drift is an action tool to pop up a chat with an email. You have intelligent to change stuff on your website.

Eric SiU: Got it.

Guillaume C.: That's how I think of my stack. I collect data first, I process the intel data, and then I decide, or the AI decides what happens. And [inaudible 00:14:49] is just one of the intel data sources. There's others, like I've talked on twitter in the past month how I'm now able to collect the data from jet eau crab, which is a sands crit view website and I buy the IP traffic to know, which companies are looking, are reading reviews of my competitors. And I add them to my email flow. So now I know outside of my website, who's doing what. You know? So that's very good intent. I buy data from Bom bora also. So that I know wich companies are reading about administration topics. And so I have intent on who's reading what. I buy data from [inaudible 00:15:34] you know, who's installing the scripts of my competitors, who's removing the scripts of my competitors.
If a company that has a high fit has installed a competitor's tool ten months ago, well, I know there is a renewal in two months, which is a great opportunity for me and my team to go after them, cause the timing is right. There's no point in trying to do that six months ago, cause they just signed. They're not going switch, but now, if the renewal is in two months, we probably got a chance.

Eric SiU: Right. I'm really interested in the methodology about how you're kind of piecing all of this together. Because Madkudu, some of these other tools I haven't heard about before, so how are you discovering all of these nifty tools?

Guillaume C.: I think all of us, we all receive, we see a lot of stuff. I think the difference is that I usually try them. I play fair game, and so a lot of those small start ups, they know that if they come to me, I'll give them an avenue to try their data and test their tool. Then give them feedback. It's random stuff. I'm now a customer of a company called Predict Leads that's in eastern Europe and that scrapes job boards to tell you wich companies are hiring what kind of role. Which means I have a data source on companies about hiring Dimension people. Head of Dimension, which is a great signal. But it means you got to be open to some people from Estonia telling you that they have a good data source. And that you've got to dive into and that's just what I do. I spend time do that. I spend time evaluating. I work with a company called Monkey Man and they are in Auriga. And they do machine learning to apply to response emails. So I know the response from my perspective customer is positive or not.
So I think it's creating a personal demand that is open to guessing these small companies and I always say that, if it works, I'll be public about it. And so I give back. That's my way of giving back.

Eric SiU: I love it. Okay, well this stack that you built around a segment, so we have Clear Bit, Drift, MadKudu and whatever else. How much does that end up costing if somebody is listening to this and they're like, I want to get started.

Guillaume C.: Yeah, that's a great question. It does cost a lot of money, however let's look it that way. When I left Segment, at the end the stack was about 200,000 dollars per Euro. Total, however the revenue was under 4 million dollars. About 20 times. Which means if you count in world day, about 31 days in a month, it means I was more or less paying back the entire stack the first day of every month.

Eric SiU: Wow.

Guillaume C.: So think of it that way, if you know about the metrics, online demension, usually for start ups, for every one dollar invested you are going to get between 0.8 and 1.2 dollars back. But that's usually what you do with most demension channels. In my case I have a bout 20 dollars back, so is it expensive? Yes. However, you'll see that cost is independent from the volume. That's the great thing. It's marginal costs afterward. After you just buy the accessing tool, because I buy all the traffic from [inaudible 00:19:06] so I've got a real big deal with them. Is it expensive? Sure, but I'm not paying per traffic as you would on ad words. And I'm not competing with other people so the price isn't going up. Which means that its up to me to maximize the revenue of that channel and the cost is fixed at the beginning of the contract.
Which means I can draw out the- [inaudible 00:19:29] is a good example where I can get high quality leads, companies that I know obviously like five, six, seven pages. I know there's eight people from the same company visiting seven different pages or eight, ten different pages. Now in a short period of time, great intent, great company. And that's when I look at the volume I look at the yeah or the cost, that's going to cost me one or two dollars. And you can look at a company like [inaudible 00:19:54] is your average customer value is going to be in the couple hundreds of monthly revenue. If your lead is one or two dollars, you are going to be in the green. You're going to be positive, for sure.

Eric SiU: That's incredible. You talked about, I guess at the enterprise level, let's say Delta, two to four percent is kind of what you saw. A 100 percent increase, right, so I'm guessing for a segment, like a V to V stats company, when you set this work flow up, how many qualified appointments, or leads or whatever, how did that shoot up? Cause there's a lot of V to V stats companies here.

Guillaume C.: So, there's a lot of examples. Look at all what the growth thing was doing in front of all that stack and creating opportunities from that stack. That represented in the end about thirty percent of segments new opportunities every month, and about fifty percent of the growth opportunities multiple months.

Eric SiU: Wow. Okay.

Guillaume C.: So it's significant, so we're talking, segment is about 200 people now, and there's like 50 sales people. So yeah, to able to say we generate 50 percent of the growth in opportunities every month is pretty significant.

Eric SiU: Drift is huge. There's also some services people that listen to this, so I have the thing where I've dried drift on our Sass product then I also have a marking agency, so on the services side, and we use Drift on there. So, the problem I have right now is the site gets about 70,000 visits a month right now.

Guillaume C.: Wow, congrats.

Eric SiU: But I think it's decent for an agency, thank you certainly, but what's going on right now, we have a lot of conversations happening, but not many appointments are being booked, and at the same time we have clear bit and rich and I'm looking at the leads right now. Audi, Mercedes, Ernst and Young and all these. It seems like nothing, you know we have drift but something's just not working we have all these unqualified conversations.

Guillaume C.: Correct, so what you need to do is you need to, so new and [inaudible 00:22:02] to this, reach out to me and you're gonna ask for the private access to the Madkudu integration. Madkudu is an integration [inaudible 00:22:14] right now it's in private mode. It's not public. But you can ask me, I'll give you access and that will enable you to create playbooks specifically for those companies. You know, and playbook's a convert mode because you can change the context. I'll give you an example. How we did that in the past and for your power trip is if Audi and Mercedes comes to your site, you're going to be able to inject in the playbook, in the message, their industry information there in the auto industry. And if you have a customer in there, which is what I did in the past, you can name that customer. Say hey, we can help Audi optimize it's buying flow, whatever service you do, you know?
We've helped companies like, I don't know if Tesla is a customer of yours.

Eric SiU: That would be nice.

Guillaume C.: We've helped Tesla, you know? And that's what I've done. Why is that important? Because it creates relevance. Cause, immediately, you message, like, holy moly they have one of my largest competitors, like a brand I recognize, it means that other brand is something I should pay attention. That's all you need. I've injected relevant companies, I've injected technologies I know they use. For example, at segment, we pulled data from [inaudible 00:23:30] darkly into the chat and say hey, we see that you are using this technology, this supported by segment. So making the message relevant on the spot by injecting that diamond data in the message, makes that conversion rate double or triple.

Eric SiU: Got it. I love it. If somebody visits my site let's say they're using Marketo or Hove spot, it's like hey, we noticed you are using Marketo, we can help you build a marketing fund, or something like that.

Guillaume C.: Exactly. And you could say, because you know the industry, you could say, we've helped companies like, name of company, optimized their market of flow. Do you want to do the same. Then what do they see? They see a tool they use, and they see a competitor or company they know in the market. There are two points of relevance. It's insane.

Eric SiU: Yeah, that's incredible. So what we have right now, just so people know, there's a lot of different options, people have different playbooks set up, but we use the account based marketing feature on Drift, so we just load in all these companies we want to work with and it is working, even with this current feature, but the thing is we have to put in 5000 companies or a couple thousand.

Guillaume C.: Yeah. And this approach often is the opposite approach. You don't need to upload a list, you just apply Madkudu which will then work for any company.

Eric SiU: Sounds way better.

Guillaume C.: Yeah.

Eric SiU: Yeah, way better. This is why you guys gotta listen to this podcast cause you get these new features in integration.

Guillaume C.: Listent to the podcast.

Eric SiU: Cool man, well how about for the people that are probably wondering right now, man this stuff all sounds good, but I don't have the chops or the time to implement this. Does Drift have or know any consultants that can do it for them.

Guillaume C.: Yeah, reach out to us. We can help you. I think the game is, you choose one strategy, you know? This kind I've built, took me years to build. Of course if you try applicate it all next month, it's gonna be impossible.

Eric SiU: Right.

Guillaume C.: You choose one strategy that is most likely to work and you apply it.

Eric SiU: Right.

Guillaume C.: And then you move on.

Eric SiU: Makes sense. So, it's not often that I get really cool, great marketers on the podcast, so what else is kind of exciting you in the marketing space right now? What kind of tools? I'm sure you're a tools guy.

Guillaume C.: Yeah.

Eric SiU: What else are you excited about?

Guillaume C.: What we've talked about is the ability to understand who's coming to your site. Who's doing what on the internet. Measuring all that intent and then doing some action, but right now it's very much role based. You can say if this happens, then do a chat. Or if that happens, then email, right? And we're reaching the limits of that. There's just too many conditions. The market is moving interestingly into AI to try and predict what the next best action is. And say, well based on all the intents that people read about your [inaudible 00:26:14] then company desires, then they click on an add. And you know the likelihood to close. Two things are happening. One, we are actually moving closer to being able to move from the fit, to predicting the actual value in dollars. So now you know that company when you apply the close rates from [inaudible 00:26:37] and how much they are going to close at, you know how much you are ready to invest.
Like my boss at Drift tells me we can invest 30 percent of that value, so I know how much to invest. And 2, based on that, you can say, well what am I going to do now? Is email the best course? If I'm allowed to invest 100 bucks, 200 bucks, is email my best first strategy? Probably not. Probably I should do som adds and a gift. Probably should send hand written letter. And all of those things exist through API's, just need to connect them. And as I said, you can't build all of those connections. And so, a lot people, including us are thinking about building an AI that looks at that value, the lead and looks at the cost of each action and chooses the best path to optimize the return investments. The yield. No in dollars.

Eric SiU: Wow.

Guillaume C.: And so I'm building the connection, so right now what I'm working with a company called Sandoval, with whom I've built a campaign that sends coffees, hot drinks automatically to the HQ address to the person that's chatting with us on Drift. I've worked with companies called Letter Friends that sends hand written letters, who's workers use hand letters and send it by the post for five dollars a pop. And I have another team that's working on the machine running the AI part to connect those and that have enough AI that decides what the best investment is to optimize inversion. So, that's the market in probably two years from now. It's removing us as marketers, from that, making the rule based and just actually creating the partnerships, buying the stuff and then the AI will decide ... and doing the message. So if I put that on the calender, two years from now, we have an AI that decided what the next best action is. In four years from now, we have an AI that actually writes the message.

Eric SiU: Wow. I love it. I basically need to have you on this podcast maybe every half year or so just to see what the heck ...

Guillaume C.: Let's do it.

Eric SiU: The new stuff going on.

Guillaume C.: How much progress we making towards that goal?

Eric SiU: Towards what goal?

Guillaume C.: Removing us, you know, I'd like to remove myself. If I could remove myself from my job, I think it would be fantastic. That would be an achievement.

Eric SiU: I think if you can remove yourself, then you are doing your job right.

Guillaume C.: Yah.

Eric SiU: Cool man.

Guillaume C.: Tell my boss that I still want my check.

Eric SiU: This is great man, so a couple more questions as we work towards wrapping up. What's one must read book you would recommend to everyone? Could be anything.

Guillaume C.: Influence by Cialdini. Robert Cialdini. Influence is probably my go to book to understand technology. A lot of what we said about the irrelevant message, injecting that personalization is explained in very good ways in Cialdini's book called Influence.

Eric SiU: Yup, love it. Great marketer's book, of course a marketer would recommend it. Okay, well, how about in terms of, because I know a guy like you is really into learning, what kind of podcast or blogs do you tune into on a daily or weekly basis?

Guillaume C.: We have a podcast called seeking wisdom. Which we don't talk about Drift. We just invite people to talk about their stories. Their learnings and it's a really good one. And the other podcast I like is the one from my friends at Saster podcast is really good also. And I'm pretty active on the community. Reading and writing some stuff there.

Eric SiU: Got it. Love it man. So G, this has been great, what's the best way for people to find you online?

Guillaume C.: Probably twitter and linkden. So Guillaume Cabane.

Eric SiU: Awesome, Guill, thanks so much for joining us.

Guillaume C.: My pleasure. Talk to you soon.

Announcer: Thanks for listening to this episode of growth everywhere. If you loved what you heard, be sure to head back to for today's show notes and ton of additional resources. But before you go, hit the subscribe button to avoid missing out on next week's value packed interview. Enjoy the rest of your week and remember to take action and continue growing.