Is your organization too afraid of trying new things? Do you have an idea for an experiment but are not sure how to pitch it to your boss?
Let’s change that.
My guest today is David Arnoux, Co-founder, and Head of Growth at GrowthTribe. In this episode, we talk about developing an experimental mindset, why visual storytelling is vital to get buy-ins, and how to run marketing and growth experiments to find huge wins.
Listen to the episode:
- The symptoms of companies that are not experimenting enough
- What companies need to have in place before starting experimenting
- The danger of “growth for the sake of growth”
- The type of companies that David won’t train or work with
- Why experiment culture should begin with the mindset
- How to change the mindset in an organization
- Why visual storytelling is essential in presenting your experiments
- The single data source to help you pick the right ideas to run with
- A few examples of crazy experiments David ran
- David’s advice to aspiring marketers
- What David thinks marketers should learn for the next 5,10, and 50 years
- Claire Suellentrop – Hot to Use Jobs To Be Done to Read Your Customers’ Minds
- Asia Matos – 5 Steps to Create a Go-to-market Strategy
- Stealing Fire by Stephen Kotler
- Liar’s Poker by Michael Lewis
- Jeff Bezos’ 2018 Letters to Shareholders
- Smartcuts by Shane Snow
- Takeru Kobayashi (The Hot Dog Eating Contest Champion)
- Lean Analytics by Alistair Croll and Benjamin Yoskovitz
- Dave McClure Pirate Metrics
- The GROWS Process by GrowthTribe
- The Brass Framework
- Occam’s Razor
- Wait But Why by Tim Urban
- OCEAN model / Big Five Personality Traits
- Sandra Matz
- Deep Work by Cal Newport
- Edward Bernays
- a16z podcast
- Lectures subreddit
- Growth Tribe Youtube channel
- David’s LinkedIn
- GrowthTribe Instagram
David: Thanks. Thanks a lot. Thanks for the intro.
Louis: Tell me, what are the symptoms, would you say, of a company that are not using rapid experimentation, who are not experimenting enough?
David: The symptoms, that’s a funny one. I think that the one that we see is usually that they’re growing slower than the competition. Then they have slower, let’s say, velocity of experimentation. There’s a very gut-driven mindset within that organization. Funnily enough, when I started… when we started training… Like you said, I have extensive startup background, more in SaaS products or eCommerce, so very short-term sales cycle. Since we started this training company, this digital partner company, we found out that actually…
David: We thought it would be the larger organizations that would suffer from these symptoms. Actually, it has nothing to do with the size of the organization. We find that there’s very small, nimble, post-product-market fit startups that have exactly the same problem as some of the larger corporates out there. It still amazes me to this day, with the amount of tools and possibilities that there are, that people are still not adopting this way of working and this process.
Louis: It’s interesting, because as you said, it’s not based on size, it’s not based on demographic.
Louis: Is it a psychographic element? Do you think it’s highly emotional, the fact that certain companies go with this, “Let’s make mistake and let’s try new shit,” while others are very afraid of that? What do you think is the common trait between companies who do not do this? What do you think is the common denominator between all of those?
David: There’s a few. The first one is age. I think I hate to compare corporates with startups. I prefer to compare digital natives with a little bit older organizations. At the moment, my neighbor is booking.com. Since day one, they’ve had this process, this mindset of experimenting. On the other side, I have Oracle, which since day one has not. I think another thing has really got to do with the business model. I think a lot of people talk about mindset and culture of organizations. What we found is that your mindset and your culture is very oftentimes dependent on the business model of your organization.
David: I’ve worked a lot with eCommerce, for example, so business model is eCommerce, or with affiliates. It is natural for those organizations to have to experiment. Whereas, when you’re tapping into things maybe more around mobility or around health services, around financial services, with a little bit longer sale cycle, then you have less of this need to experiment since day one. I would say the age of the organization, the business model of the organization and then, of course, normal variables. Like, what is the DNA of the co-founders? What is the DNA of the senior management? Also, has the company been incredibly lucky?
David: I think that a lot of the companies that we train or that I’ve worked with, that have… had pain in the past, it’s forced them to learn to experiment, because nothing’s come easy. Whereas, other ones where everything has worked since day one, have less of this culture of experimenting. I think I’m obsessed with two companies, which are PayPal and Amazon. It’s been beautiful to look at the history of PayPal, where it was just a struggle from day one. It’s forced them to learn to be humble, to look at the data, to pivot regularly and to experiment with different things.
Louis: Right. Okay. Now, I think, we’ve nailed the profiles of people who are not using it or not doing properly, and the profiles of people that are doing it quite well. Before we go into the practical step-by-step on how to set up your own process and making sure to experiment and get good wins, there’s something I want to ask you. As you know, this podcast is really about marketing foundations, things that won’t change in the future, are not likely to change, focusing on the customer. We’re not going to go into details about this. Before setting up any process like experimentation and growth, what do you think are the foundations that those company must have before that? What are the things that they have to have in place?
David: Yeah. Well, the foundations are obvious. The first one being product-market fit or close to having product-market fit, which means that there’s a real pain to solve. You have the solution to solve that pain. You know what the market is. You know who your target personas are. I think that’s something that’s very, very rarely looked into clearly. Who are we actually trying to solve the pain for? What are the jobs to be done, of your product, to solve that pain? I think maybe one that’s a little bit less obvious is just what we call the go-to-market strategy, having at least a strategy for how we’re going to address the customer. Is it short-sale cycle, long-sale cycles? Is this a discovery type product, or is this more an intent-based product?
Louis: Okay. We’re not going to touch on that in this interview. We’ve talked about it a lot in past episodes, on persona jobs to be done, go-to-market. We’ve done it all. Folks, if you’re listening to this right now, just go back to previous episode if you want to have this information ready. Now let’s… We’re going to assume, in the next few minutes, that you do have those foundations. You do have the problem we’re solving, the solution, that we actually do have a way to address customers. You know your customer inside out and you have those foundations.
Louis: Before, I want to… Before talking about this step by step, there’s one last item I want to talk on… talk about. I feel like in the marketing world and the business world, there is a lot of people struggling a mental health issue, a lot of people struggling at work, with the pace at which things are moving. I do put some responsibility to the mindset of certain companies and certain people who think that growth is the only thing that matters for a business, instead of just there are people and making sure that there’s a balance. What are your thoughts on this? Where do you stand in terms of growth? Is it growth at all costs? Is it growth for the sake of growth? What is it?
David: Do you know that Edward Abbey quote? I think it’s, “Growth for the sake of growth is the ideology of the cancer cell.” That one really resonates with me and makes sense. I don’t know. I think I haven’t seen so many problems with mental issues, to be honest, at least not recently. I would recommend to anybody who struggles with this or their employees, or whatever, to read the book… It’s a pretty recent book. It’s called Stealing Fire. I think it’s by Kotler. They talk about the state of ecstasis, the state of flow. I think that’s one of the best ways to fight this mental breakdown that’s possible with all of the notifications that happen on a regular basis. Trying to reach this state of flow through either things like meditation or things like knowing how to focus, working on deep work.
David: Your original question was, is growth for the sake of growth healthy? No, definitely not. I do think that it depends a little bit on the business model that you have and whether you’re in a company that has real vision, a real mission. What’s funny to see is a lot of companies reverse engineer their mission and their vision later on. They start out just wanting to grow for the sake of growth, being a little bit sleazy, a little bit sneaky in the early days, because they can. Then two years in, when they’re a little bit larger organization with a bigger target, let’s say, then they start building in that mission and that vision. I don’t know if that’s a good answer to your question, but I do think that there needs to be a reason for the growth.
Louis: Right. You don’t grow for the sake of it. You have an ultimate motive, something a bit bigger than the business itself, that forces you to like, “I want more people to be aware of this. I want more people to use the product, because I believe in a world where this should be more popular.”
David: There’s extensive literature that researches the fact that successful founders tend to be driven by a real mission and a vision, not simply growing for the sake growth. Except maybe traders, except maybe on financial markets, where it’s cash for the sake of cash. Just read The Bonfire of the Vanities or read any book about, Liar’s Poker, about the world of traders. It might be one of the [inaudible 00:09:56] areas.
Louis: Were there big companies that you would you say no to, that you wouldn’t train?
David: Yeah. Sure. We have a whole list of guidelines on companies that we are willing to do or not. Example, we’ve been contacted by the arms’ industry. That’s an easy one. We refused that one. Then it becomes really difficult to be ethically responsible, because some of the worst companies don’t actually look like the worst companies. It’s easier, sometimes, it’s simple to shout at a, I don’t know, a petrol chemical company, like an energy company. It’s really easy to criticize them.
David: Then you’ve got, maybe, a retailer that’s wasting a lot, or you’ve got a clothing manufacturer that’s burning a lot of excess stock. That’s a really difficult one. It’s easy to judge a company by its cover. We do, do very strong analysis of the companies that we train. We’re not perfect. Average age here is 25 and a half, really purpose driven, really want to make the world a better place. We basically have no choice [inaudible 00:11:04].
Louis: A few years ago, I had an agency. One of the belief was we wanted to work with good clients that we’d be proud of. I remember we had the debate around should we work with porn companies, because we had a few porn companies reaching out. I remember the debate about it. Like yeah, but is it good or bad? It just went… You go to this rabbit hole of, actually, as you said, even on the surface, it might look like a bad choice or a good choice. When you start investigating, everyone is connected to everyone. It’s getting very difficult to understand whether a business is doing good or not [crosstalk 00:11:40].
David: Let me… I could give you one clear example, so gambling companies, gambling organizations. On the face of it, you would say that it’s not good to work with a gambling organization. They’re addictive, they make people lose their money, et cetera. Then when you study gambling a little bit, it seems that there’s a correlation between people who like to gamble and people who are quite successful in life, and people who like to gamble and people who tend to not default on their credit payments.
David: I think there’s a company called… I think it’s Health IQ. I can’t remember exactly what the company is. Basically, it’s a lending service. It’s a financial service. What they do is they correlate how likely you are to repay your loan, based on how you use your mobile phone. They’re trying to find your psychographic and your personality, based on how you use your mobile phone. You accept to share everything that’s happening on your mobile phone, [inaudible 00:12:28] encrypted and on the other hand, they’re going to give you certain interest rates, whether they’ll give you a loan or not. It turns out that people who have one or multiple gambling apps on their phone, are more likely to pay back their loans and less likely to default.
David: When you actually start looking at the social science behind it, et cetera, is it so bad to actually trade or to work with a gambling company, especially one that has limits in place? In that case, I can’t train Candy Crush either, and we can’t work with the Fortnite, or we can’t work with most video games, or anything that might be addictive. A lot of eCommerce companies… We have data from a lot of eCommerce companies. Some of them have their customers coming six times a day to browse for over, in total, 45 or 50 minutes. Is that a form of addiction? Like you said, it’s definitely a rabbit hole. I think it’s a very difficult question to answer. Like any question, the first step is ask yourself that actual question. As long as you have a radar, an ethical radar, that’s already a good start, I would say.
Louis: Right. Let’s go to the practical side of things. Thanks for answering this question. I know it’s a tricky topic. I think you’ve shared a lot of interesting data. Need to dig into this, the gambling side, a bit more as well. Let’s consider… We have company. You might want to take a real example, if you wish, or few examples. We have a company that have reached product-market fit. They understand who their person are. They have a way to reach out to those people. They don’t have this way to constantly experiment, to go fast and try and all of that. What is the first step? What do we do to go from this point, to a point where we can rapidly test and win?
David: I think the first step is a little bit some of the first questions you asked. It really… We focused a lot on the tactics and even the process for a very long time. We forgot to focus a little bit on the mindset. What’s interesting to me is that experimentation is actually at the core of many, many things in our lives. It’s almost like since the beginning of time, random experiments have determined what survives, what dies, what adapts, what dominates. It’s what we call natural selection, survival of the fittest. It’s also feedback loops. The law of the jungle is feedback loops. It takes billions of years. It’s more luck than controlled experiments.
David: I think humans, they began to discover this, it’s debatable, but the scientific method was developed by… discovered by Francis Bacon and also by Galileo, in 17th-century Europe. That was the first time in our common history where we looked towards facts rather than just faith. It was increasingly recognized that facts… The result of a series of trials and errors, of tests and experiments, the scientific method was born. We owe some of the world’s greatest invention to this; like antibiotics, silicon conductors, spaceships, all that stuff.
David: Now what’s weird is if the fittest species have embraced this and if science has understood this, why do an overwhelming majority of people and organizations still stick to fate, uncertainty, when taking decisions, when placing bets on ideas? Why hasn’t everybody embraced this trial and error mindset of controlled selection? Some companies have. Company like Amazon, I try to study absolutely everything that Amazon does. Because if you read one my favorite business books, it’s not a business book, but it’s just reading the letters to shareholders, from Jeff Bezos to shareholders. One recurring theme is that their success at Amazon is a function of how many experiments they run per year, per month, per week and per day.
David: That’s one question that I really like to ask founders or companies, “How many experiments did you run this year? How many did you run this month? How many did you run this week? Or how many did you run today, for example?” Of course, depending on how much traffic you have, how much resources you have, you can’t run as many, but how many are you running? Every company runs experiments. Some call them projects. Some call them campaigns. Some call them hiring. This mindset of trying to experiment and do it rapidly is probably the first step.
David: One story that I do like to tell, it’s a story that was in the book, what’s the name of the book, Smartcuts by Shane Snow. It’s the story of this 1.76-meter Japanese man, who happens to be 62 kilograms, and who became the world international champion of the Coney Island Hot Dog Eating Contest. This is a guy who was up against people who were genetically born to be the Coney Island Hot Dog Eating Contest champion. The world record was at 25 hot dogs in 12 minutes. Just imagine eating 25 hot dogs in 12 minutes.
David: This guy, the first time he participated, he ate 32 hot dogs, so he… Sorry. He ate 52 hot dogs in the same amount of time. He doubled the world record. Yet, he’s the small Japanese man who doesn’t seem physically built. What did he do? Actually, everyone focuses on the tactics that he used to win. He experimented, basically, for six months in his apartment in Nagano. How can I eat hot dogs faster than the competition? He did 70 experiments in total. Turns out that two of those experiments actually worked out. He changed the paradigm of hot dog eating.
David: The first thing he did is he separated the sausage from the bread. It turns out, if you separate the sausage from the bread, you can cram down sausages really fast down your throat, rather than having this mushy substance of the sausage and the bread. The second thing that he did was that he found a hack, a legal hack, a compliance hack. You’re allowed to use any liquid that you wanted during the contest. Everybody was drinking the water, the liquid. He started predigesting the bread by dunking it into the water, so then he could just cram it down his throat as well. He obliterated the world record.
David: It’s really about the mindsets. Everyone else was following the same paradigm. They were all eating hot dogs in the same way. Whereas he came along, and he’s like, “I’m going to apply some lateral thinking. I’m going to experiment and maybe I’ll find a better, faster way of eating it. If you’re in business nowadays, there’s so much competition, you better learn to eat hot dogs really fast or, unfortunately, and I know we don’t want to create too much anxiety, you are going to get beaten by a company that goes faster, I believe.
Louis: You make me hungry, Jesus.
Louis: Beautiful story, no, I really like this story. It’s a really good example of how eating applies to everything, as you said. The obvious next question then is… Okay, that’s all well and good, we need to have this mindset. How the fuck do we change our mindset, if we don’t have this mindset? If our CEO is too scared of running one experiment a month, how do we convince them to change?
David: 100%. I think that what’s important is there’s experimenters in every company. They’re just sometimes hidden or they’re not allowed to express themselves. We need to try to change the mindset. Now, we’re in the example where there’s pushback, the mindset isn’t there. I can promise you that, I don’t know, 20/30/40% of your company at least has the mindset. They’re just not allowed to do that. The culture doesn’t allow it. Here, we’re talking about change management. We’re trying to change the mindset of the organization. There’s ways of doing that. After a while, if it doesn’t work, very bluntly, change company.
David: The way to do it is usually to start with a use case. We like to look for the typical cliché, low-hanging fruit. Which basically means it’s something that is important on the roadmap, very important on the roadmap, so you understand what the strategy of the company is. Usually it’s linked to retention of users or increase in revenue, important on the roadmap. It’s quite easy to fix. There, what we do is we try to build a pilot case, try to build a first team. It can be one-person team, maybe one-and-a-half people. Then we try to hit those low-hanging fruit to prove, through data, that by experimenting a little bit, we were able to improve specific metrics, faster than anybody else has or quite rapidly, let’s say.
David: It’s a lot easier said than done. It’s about identifying what those possibilities are. It’s about actually executing it yourself and getting a little bit of buy-in from stakeholders. To be totally honest, if you don’t have buy-in from stakeholders, just do it under the radar. There’s that famous story at Facebook where Mark Zuckerberg was convinced that videos would never work, that they shouldn’t dive into videos. They just did it under the radar. They didn’t tell him about it. Fast-forward six months later and Facebook videos was the second-biggest video streaming platform in the world.
David: We’re looking for that sort of use case. Then, you’d use that use case. You showcase it within the organization. If it still doesn’t work, then sorry, you’re in the wrong place.
Louis: All right. Okay. I like this approach because we are starting slow. That’s one of the core user psychology facts, like the foot-in-the-door principle. You start with something quite small and to have something quite big afterwards. You don’t start with trying to say, “Your company is shit. Your mindset is shit. Let’s change everything right now.” You start with something small, that you can prove, that has actually a business impact that directly relates to what people care about. You move on from there. If you can’t do that, and I like your idea, if you can’t do that, do it under the radar. Don’t be… Be careful not be fired though. Do it in a way that, you know what, fuck it, let’s try that. If it doesn’t work, it doesn’t work. If it works, it’s a good story to tell.
David: Fired. What’s a job? A job is an experiment in your personal life as well. Maybe this is your experiment to see if this is the right place for you to be.
Louis: Exactly. When you start thinking this way, everything is an experiment, isn’t it? How do you then advise…? We’ll go into the, like, what quick wins to pick and how do we select the right ideas? But how do you advise to present such use case, once you have it? Let’s say, in an ideal scenario, let’s say something, it’s working. How do you advise to present it, especially to people who you know don’t have necessarily the right mindset, might come from a place where they’re not super into it just yet? How do you present it?
David: You mean like physically, how do you present it?
Louis: Yeah. Let’s say if you have the possibility to be in the same room or in line or whatever, how do you show that it worked? What’s the best way?
David: We use a bunch of templates for this. It’s basically doing a little pitch of what the use case was. Basically we say, “This was the assumption. This was how I designed my experiment. This is the experiment that they executed. These were the metrics that we were focusing on. These are the results that we had.” What we found is that visual storytelling is extremely important for this. Nobody wants to read a piece of paper, let’s say. It usually needs to be a little bit rich, a rich story that’s told.
David: One thing that works really well is to record. Basically, I’m very practical. If you have the person in the room, you just pitch it one PowerPoint for like 10/15 minutes. If you don’t, if it’s harder and if you want to spread this within your organization, you’d like to make small videos. You open QuickTime on your Mac. You put your PowerPoints or your Keynotes. Then you present it. You do the audio on top. Then you share that on Slack or via email. We also think it’s very agile way of working, similar to Scrum, to have demos. When there’s a team that’s working on experiments, they have their [inaudible 00:23:39] on board, of experiments, that’s easily visible for everybody and a biweekly demo of the experiments that they’ve been working on.
David: I would argue that you don’t even need to get wins yet. You can actually motivate people to work in this way simple with learnings. Although I’m not a big fan of running experiments just for learnings, but already having learnings about the customer, about the product, about the product-market fit, about the customer journey, is already interesting. “We’ve learned four things about our customer journey in the past two weeks.” That can also be interesting. Just a side note, I think 90% of people do adhere…. would adhere to this way of working. It does make sense. It is common sense. The scientific method does quite make sense. We don’t usually get that much pushback. I can talk later about what the main blockers are, what the biggest pushback… where the biggest pushbacks come from. It has nothing… It’s not so… We don’t see it often.
Louis: Okay. That’s really good. It’s really about visual storytelling. It’s about telling a story from the perspective of customer. It’s not about sharing a spreadsheet that is ugly as fuck, that no one can understand. You tell a story and you focus on the user. Okay. Now, let’s say, we have the right mindset in the company, people are already for it anyway, but we don’t have the process. What is step two then? What do we do to start really firing experiments left, right and center?
David: Yeah. There’s two ways to do it. There’s the cookie-cutter approach and then there’s the [inaudible 00:25:15] approach. I would say, if you’re not in a team, if you’re on your own, go with gut feeling. Focus on one metric and try something else. Tell yourself, “Can I try four things? Can I try four different experiments on the same metric, in the next two weeks? Is that possible?” Usually, we say this. That applies to maybe one-fourth of people, because it’s difficult.
David: Then we have actually taken the scientific method, all the way back to Francis Bacon and then broken it down. We have a video about this on YouTube. It’s quite simple. There’s these five steps that you should follow. The first one is to focus on one metric at a time. There’s a guy, I can’t remember who he is. He interviewed 1,500 successful entrepreneurs who had built at least one-million-dollar companies, which is semi-successful entrepreneurs. The one thing he found that they all had in common was focus. They were able to focus. The scarcest resource that we have is time. I would say the first part is to really identify what Alistair Croll calls in his book, Lean Analytics, what’s the one metric that really matters for me right now.
David: You look at your customer journey. You need to have… One thing I forgot to tell you earlier on when you asked me, “What do you need to have in place?” some sort of data tracking, some sort of tracking of what’s your conversion funnel, what it looks like. Let’s say we’d look at your conversion funnel, your user funnel, your customer journey funnel. Where am I bleeding customers? Where am I bleeding users? What’s the real big pain points that I have right now? It usually is revenue, but is there one I can fix before revenue? If nobody’s coming to my website, my product, my service, there’s no point in trying to experiment on revenue. I need to find ways to drive traffic to my website.
David: We use Dave McClure’s Pirate Funnel. I think it’s our favorite customer journey frame. We look at six metrics. We look at awareness, which is driving people to my website, product or service. We look at acquisition. How many of these people don’t actually drop off when they visit my website, my product or my service? We look at activation. What percentage of people reach the, “wow, I understand the value of my product” moment? We look at retention. How many people are actually coming back daily, weekly, monthly, yearly? We look at referrals? What’s my viral coefficient? Finally, revenue, how many people are actually paying for it?
David: You try to paint a picture of this funnel. Once you’ve identified that funnel, so, we try to identify what is the one metric I want to focus on right now, the OMTM, the one metric that matters. There’s usually one or two metrics. Typically, it’s either a retention metrics, if you’re a subscription product. Or it’s a conversion metric, if you’re a transactional product. If you’re a media product, if you’re a media entity, it’s usually something a bit [vanity 00:27:51] like number of page views or number of ad views. Focus on one metric. I think if we were to just stop the podcast right one, that would already be 50% of the value, try to focus on metric at a time. Once you’ve identified that metric-
Louis: Let me go back. Let me go back, because you said a lot of important stuff and I want to make sure we cover them. You said something you forgot to say was you need some sort of tracking. I don’t want to go too much into the tools right now. Because I hope that this episode can still be relevant in five years or ten years, and those tools might have gone already. Basically, what you’re saying here is you want, in an ideal world, you want to know how many people go through each step of this funnel. You want to know whether or not there’s a huge drop-off in one particular stage. At a big… At a high level, this is what you want to know, right?
David: Exactly. Where’s my most expensive drop-off? Exactly. Where am I the most…? Yeah.
Louis: Your most expensive drop-off. It does not necessarily mean that it’s the place, in the funnel, where you have the highest drop-off in terms of percentage.
David: No. It’s really… Let’s use a few scenarios. It could be, if you’re running ad campaigns, that’s expensive if people… very small number of people are converting on those ad campaigns. If you need to test for retention, that means that you need cohorts of users coming through your websites or to your service, on a weekly basis. What’s killing you right now is that you’re not driving enough traffic to be able to measure retention, what your retention rate is like. That would be the most expensive.
David: If you’re a transactional website, if you’re selling something, you’re doing drop shipping, something like this, if you’re a consulting firm… Well, let’s say drop shipping, you need to be converting more. You need to be increasing your conversion rates on your website. If you’re like a consulting firm, it’s product your deal size. That’s your most expensive metric, is, our customer lifetime value or deal sizes are 10K at the moment. How do we get them up to 100/150K?
Louis: The place in the funnel where you have the highest, the most expensive drop-off and the one metric that matters, are connected, right?
David: Yes, 100%. I would say… There’s this Sean Ellis, came up with the notion of the North Star metric. That’s the most important metric for your business, on the very long term. One metric that matters is the most important metric for your next two weeks, short term. Next two weeks, what’s the one metric that you need to fix? Like, you’re in pain right now and you need that Band-Aid, you need to fix it right away.
Louis: From experience, I suppose it the same for you, when you look at your funnel like this, there will be something screaming at you. It’s not going to be a point where you’re like, “There’s five things. I don’t know where to start.” There’s one that is bigger than the other, usually.
David: Eight times out of ten, you don’t even need data analytics tool right now. I should not experiment. You already know where it is. No one is visiting my website. People are dropping off after one week. We’re not making enough money. They tend to be really simple. All this stuff is actually quite simple. That’s why we stopped consulting on this stuff and we started training people on it, because it’s actually quite simple.
Louis: Interesting. Before we go to the next step, what do you mean by this? You’ve stopped consulting on it, because it was too simple. You felt there was no value in the consulting, more value in training people to do it?
David: Yeah, exactly. The thing is, consultants or agencies, we work with agencies and consultants. They try to sometimes complexify things in order to increase the price. There’s a saying like this. I can’t remember what it is. It’s complexify to something -fy. I can’t remember what it is. Before starting Growth Tribe, there was this one project where I trained three interns on my process; analytics implementation, how to run experiments, what tools to use, et cetera. To be totally honest, they did it better than I did, after a month and a half.
David: When I used to do consulting projects, six months later, I would give them a call, “How is it going? Are you still growing?” “No, you left.” Whereas where I train people, those three interns inside a team, call two months later, “Hey, how’s it going? You’re still growing?” “Yeah, 100%. By the way, I’m head of growth now and we’re running experiments you want. By the way, we’ve done this, this, this.” Like, “Oh wow, I just learned something.”
David: We found that, actually, growth should be embedded within the DNA of the company. I still don’t understand that we internalize product development, yet so many companies still outsource their growth and their distribution. They’re outsourcing their tone of voice. That’s crazy in a world where distribution is maybe your biggest competitive advantage. In that sense, educating people inside the organization is more cost-effective. You have longer-term impact on the organization, embedding that growth into the DNA.
Louis: Interesting. This is one of our biggest marketing pet peeves, this addiction to name and coin new terms and create new processes on top of processes, to make yourself look smart, instead of simplifying to the core. Because as you said, this thing, even though it’s hard to do, it’s not complex, it’s not difficult. It’s just simple, but it takes hard work, right?
David: Yeah. I think the difficulty grows with the maturity of the organization. The first steps are quite easy. Your first experiments will be quite easy. Then you get more mature and you grow in maturity. Then your tool-set will become difficult, your tech stack will become more difficult, types of experiments you’re running will become more difficult, the metrics you’re going after will become more difficult.
David: We always advise people, for their first experiments, to tap into the top of the funnel, the tofu. That’s usually where you have the least blockers to experimentation. It’s usually where you have the most traffic, so you can get the biggest numbers, if you need statistical significance. It’s usually where you have the least stakeholders. It’s usually where you need the least support.
David: The deeper you go into the product journey, the more that you’re stepping on people’s toes; operations, sales, marketing, developers. If you’re working on a SaaS product, subscription product, and you need to make changes to the features, then you need to get buy-in form very scarce researchers and developers. We always say, usually, the easiest one tend to be at the top of the funnel.
Louis: Right. We have our one metric that matters now. What’s next?
David: We like it when people remember this. We call this the GROWS loop and G-R-O-W-S. Basically, first, you’re going to gather as many ideas as possible. You know how developers have a feature backlog? Growth team, they have an ideas backlog, which is basically a dump of all the ideas. We like it to be a democratic idea backlog. We’ve actually developed a tool called the growth tool, where people can drop ideas, whenever they want, into that tool. One thing that’s really interesting is we found that most of the best ideas that have worked or the most creative ones, don’t happen during working hours. That’s an interesting fact.
David: It’s basically a super fancy UI, for people to drop ideas in there. The ideas, usually the best ones, come from the best internal and external quantitative and qualitative data. This is one of the reasons why larger organizations that have business intelligence units, they tend to be better at coming up with successful experiments. Whereas, a smaller startup, a smaller organization who’s doing a lot of guessing, they can run more experiments, but they have less successful ones. The first part is to just do a brain dump of all the ideas that you have, based on data analytics, based on user testing, based on user intelligence, business intelligence and just based on your gut feeling, for example. We call that gathering as many ideas as possible.
Louis: Based on user testing, based on analytics, based on a gut feeling, from your experience, what it the best source? If you have to select one source of data, and try to be as specific as you can, what is the best source of data that you can rely to truly come up with good ideas?
David: Let me give you two. I’ll do two. For top of the funnel, for campaigns, your competition, just copy, reverse engineer. Reversing engineer campaigns from competition. There’s a million tools out there that allow you to do this. They are probably using a tactic that’s going to work for you, really good one.
David: For deeper in the funnel, for things like retention metrics or revenue metrics, it’s actually just really hard, cold, raw data from your analytics software, usually by doing a little bit of correlation analysis. There, you want to have something nice set up, like an [amplitude 00:36:18] or a mixed panel or just whatever, to be able to look at the data. That’s why, also, those experiments on retention tend to be a little bit more complicated than the ones on marketing campaign. It’s the hard data but also the qualitative data linked to that.
Louis: What you mean by correlation analysis would be, for example, when this user used this tool and this tool together, they tend to retail more than when they don’t, for example?
David: Yeah. I’ll give you practical example. We were helping out a task management app. They found out that, when we ran correlation analysis… Of course, correlation is not causation. Just because two things are linked doesn’t mean that one caused the other. We found that people who had integrated their calender into the app were more likely to be retained longer. The calendar integration was lost somewhere really deep in the app. It wasn’t part of the onboarding. Then we did an A/B test, causal and test, where actually the onboarding was only three steps. The second step was integrating your calendar.
David: Yes, in fact, when run that test, we were able to… I can’t remember what the uplift was, but more people were actually retained on the platform, once they integrated the calendar. No one would ever have thought of that feature as helping with the retention. The second one that we found was customizing the color of the left menu. It’s stupid, but being able to customize the color of the left menu actually meant the people felt like they own part of it. It had less of an uplift, but it also had an uplift there.
Louis: What do you advise people who don’t necessarily have the resource to do that? What other source should they use? You said competitor. Is there’s something else they can replace it with, that is maybe not as good for sure, but maybe good enough?
David: Yeah. Sure. You mean for deeper metrics, that are deeper in the funnel?
Louis: Yeah, to have another source to rely on, to generate ideas.
David: Yeah. This is one of the stuff we teach as well. This one’s pretty straightforward. You use… You do the same thing, but with qualitative data. I call it Sunday data science. Basically, what your going to do is you’re going to take three batches of customers. Ones that never use your product, they dropped off right away; ones that are still using your product, and ones that started using your product but stopped. Or people who didn’t buy, versus people who almost bought and people who actually bought. Then you fight like crazy to try to have an interview with them or to talk to them. Then you find out, what is the difference between the ones that stayed, the ones that didn’t stay.
David: Now if you don’t like talking to customers, of course, you can use tools like Hotjar. Hotjar user recordings are fantastic for this. They’re like… I used to do this. I would watch between a hundred to a thousand user recordings at five times speed. Wait. Is that a thousand? I’m exaggerating. It was more like a hundred. I would look at a hundred and with Hot… This has become a sales pitch for Hotjar. I would look at a hundred user recordings of people who we retain on our platform, at five times speed, on Hotjar. Look at a hundred user recordings of people who were almost onboarded and almost started using the product, but dropped off. My brain would do the correlation analysis. I would start to spot some patterns. This is the number one tactic that most people use, who are too lazy to set up good analytics. It works. It just works.
Louis: I’m glad you mentioned this tool. I don’t mean Hotjar. I mean your brain. Because as you said, your brain is actually pretty good at doing correlation analysis. It’s not… It’s definitely fallible. There’s a lot of mistake there. You have a lot of the assumption you want to validate. As you said, if you look at the behavior of people who didn’t convert, versus the one who did convert or the ones who almost didn’t, you can start seeing patterns. We’re pretty good at doing pattern analysis. You see patterns, differences. This is when you can start seeing, those ones who didn’t convert, they didn’t understand that you could close this window, while the ones who did… You start seeing those stuff. That really should make your brain run and come up with shit and other ideas as well.
David: Yeah. I think what’s beautiful with the digital world is you can have access to this in one or two days for a very little price. This used to take focus groups, which are bullshit anyway, because people act differently when they’re under the microscope. If you look at bacteria, in a normal state, bacteria doesn’t move. As soon as you shine a light on bacteria and look at it through a microscope, it starts moving. I think it’s called the [inaudible 00:40:38] effect or something like this.
David: You have… Nowadays, with all of these beautiful tools that are available, I’m able to launch a five-second test on my website In a matter of 15 minutes, I can understand qualitative data based on Hotjar user recordings, in about a week, for like 19 bucks. I think it might even be more powerful than conducting customer interviews, because customer interviews, you have the… There’s always the problem of, if I only interview 10 people, maybe it’s up to luck that I can see a pattern in those 10 people. We have access to so much data, thanks to these tools. It’s exciting. There’s almost no excuse to not be using it. I think the excuse might be lack of knowing how to Google, and a little bit of laziness maybe, or lack of drive.
Louis: All right. At this stage, we gather ideas. Anyone can pitch in. You said it’s a democratic process, meaning people vote?
David: Yeah. That’s the next part. That’s ranking the ideas. We use some ways to rank the ideas. You want to try to have as many brains as possible in the room, so typically, a creative person, analytical person and then a salesperson. A hacker, a host or a hipster, or a developer, a marketeer and maybe somebody who’s customer-facing. Sales is customer-facing, of course, but like customer success. Everybody always says you should talk to your customers, by the way. I find the best shortcut for doing that is to talk to sales or to talk to customer support, for example, because they talk to the customer every single day.
David: We will take all of the ideas. Then we rank those ideas based on one or two frameworks. If it’s a marketing strategy idea, we give a framework called the BRASS framework. It’s a mix of gut feeling, blink. R is relevance, how relevant is this channel for me? I won’t go into all of the details. Basically, there’s two frameworks. A little bit of gut feeling, a little bit of how big of an impact will this idea have, and how easy is it to implement this experiment?
David: I’m fine for you to redo your whole onboarding or redo the whole navigation on the website, but can you actually execute on that experiment? We try to look for what’s called the low-hanging fruit, is something that has a high probability of succeeding, based on your gut feeling, that will have a high impact on the metric that you’re trying to improve, and it’s just easy to pull off. Once you start to be a bit more mature, you can then filter these depending on, do you want a quick win now or do you want a really impactful win.
David: We’re training, for example, a large pension fund at the moment. They’re actually running, now, 15 experiments every two weeks. They’ve reached the maturity where they’re going for the high-impact experiments, because everything is set up. We try to get together, rank those ideas, and then what happens is, organically, you have always three/four ideas, five or six that really stand out. We’re going to put those aside, say, “We’re going to do those.” Then we have two Takeru Kobayashi ideas, usually, like really radical ones. That’s the hot dog eater I was talking about before. Paradigm-shifting, not so obvious ones, where the probability is not so high.
David: Then somebody usually puts a veto and says, “I want to do this crazy idea.” We try to have 20% of experiments to be a bit wacky, not based on data, just based on, “I really want to carry out this experiment.” I think one thing that’s really important is an experiment will never succeed if somebody doesn’t truly believe in it, if there’s not an experiment owner. Because usually, there’s one or two strikes the first time you launch it. Somebody does really need to believe in and be the leader of that idea.
Louis: Give me an example of a crazy experiment that was actually successful and no really felt it would be.
David: That no one actually thought it will? I think saying no one thought it would be, would be maybe a-
Louis: Maybe just a few. It was crazy enough idea to say, “Shit, if it works, wow. If it doesn’t, yeah, no surprise.”
David: Let’s use a personal example. For us, it was, for example, giving stuff for free that was usually extremely expensive and thinking that we would actually be able to convert people after it. Our products have a high value, so the price is sometimes quite high. We thought, could we repackage this into something that’s a little bit smaller, a little bit more of a teaser, but that actually we don’t make people pay for? For example, a power session. Would it be possible, then, to actually convert these people, even though they already had all of the value from the course? I think another… The one I gave before, about the calendar, nobody believed that the causation was true. Everybody thought it’s just correlated. People are integrating the calendar because they’re using the tool. They’re not using the tool, because they’re integrating the calendar. It turns out it was exactly the opposite.
David: I do have to say that these radical ideas though, 90% of the time, they actually don’t work. Don’t do only those. Do… I think people should… That’s actually a problem. People do tend to focus on these radical paradigm-shifting ideas. Actually, the easy stuff, the really obvious stuff, it’s like Occam’s razor, the solution is usually quite simple. You were talking about basics before. Nine out of ten companies or teams are just not doing the basics yet. They’re not. Usually it’s the simplest stuff that actually ends up working. I have one example, one more, if you want, or we can…
Louis: Yeah, let’s go.
David: This was for a consumer electronics company. They were launching this big, new razor. Somebody’s idea was… One of the ideas was the big PR firm came in and they said, “You should start a movement and there should be the copywriting,” et cetera. One of the interns that was in there room said, “Why don’t we try something crazy? Let’s just call it This is Not a Razor.” It was actually a razor. “Let’s try the campaign, This is Not a Razor.” Nine… There was seven people in that room out of eight, didn’t believe that that would work. We tried it anyway and it just exploded all of the other campaigns. That one was pretty fun.
Louis: Nice. To summarize what you said around prioritization, you basically have two frameworks that you use. There’s the ICE framework, like the impact, confidence and effort that it takes. You can rank that pretty well. The second one you mentioned is what, the BRASS? BRASS, you said?
David: Yeah, just google it. It’s called BRASS framework, B-R-A-S-S, so it’s blink… This is only for traffic generation ideas. Blink, what’s my gut feeling? R, relevance, do I have product-channel fit? If I’m GoPro on YouTube, fantastic. GoPro on LinkedIn… Do I have product-channel fit? Availability, do I have the resources in-house to do this? If it’s videos, do I have video software? Do I have people who can actually make video? If I want to do content, do I actually have somebody who can make content?
David: Then important one that we sometimes forget, it’s the S, so B-R-A-S, scalability. Can I crank this up? SEO is sometimes hard to scale. PR stunts, they’re hard to scale. Conferences, they’re hard to scale. If we want something that really allows… We have very, very strict lead generation targets for next quarter and we need something that’s easy to scale, then we will also maybe give a more important score to the scalability, hopefully a very scalable channel works, like a social channel or a search channel.
Louis: Right. As you said, at the end of this exercise, we have a bunch of experiments we know we want to run. You mentioned six. You don’t have to state the six, I suppose. What do you do then? What’s next?
David: Yeah. Now you actually have to design the thing. Design the experiment in very precise details. For that, we just use experiment card. Basically, we believe that by doing this, by doing X, we can improve this metric. To verify that, we’re going to… Then you describe your experiment. To verify that, we’re going to do this and this for a period of one week. We are right, if we hit these numbers. We are right if…
David: Let’s say we’re doing, I don’t know, like a campaign. I have no idea. Let’s just say we’re doing a campaign. We believe that by launching this campaign, we’re going to be able to increase sales. To verify that, we will create 20 different assets for this campaign. We will launch them on this platform, this platform. We will let the ads run for a period of one week. We will spend 100 euros per ad asset. We are right, if we get at least this much increase on our one metric that matters. For example, we sell 25% more than we usually do.
David: That’s the hardest one, the “we are right if…” putting a line in the sand, to see if you’re right or not. I actually shouldn’t say this, but it’s also okay not to put a line in the sand. Just run your experiment and see what happens, and see if it increases the yes or no. That’s basically you design your experience with excruciating detail. Then you reevaluate your eScore, once you’ve actually designed your experiment, because it turns out you forgot a lot of stuff when you were actually doing the eScore. Then it’s good to have those five/six. That’s basically it. You design your experiment with excruciating detail, as much detail as possible. We call it experiment.
David: We have a theory at Growth Tribe, that 90% of experiments should last two weeks maximum, from the thinking about it, to the setting it up, to the letting it loose into the wild and analyzing the information. If it lasts more than two weeks, you need to break it down. Even if this is a big strategic partnership, even if this is a big company you want to launch or a big new feature release, if you’re smart about how you design your experiment, you actually test the riskiest assumptions of that big campaign or that big feature release, with smaller experiments.
David: I want to have a strategic partnership with Virgin Media. I have no idea. Well, okay, that will probably take six months to test. Okay, well, let’s break it down. Would it be possible for me to already…? Can I even find 10 contacts of the people I want? That’s one experiment. Probably takes me like, I don’t know, an hour. If I reach out to them, can I get two of them to contact me back. That probably takes a day or two. You can always break these things down. You’re basically trying to hedge your bets. You’re trying to limit the risk as much as possible, by placing as many small bets as possible.
Louis: Why the focus on only one week or only two weeks, especially when you get started?
David: Yeah. It’s just that we did this… You know Tim [Auburn’s 00:51:00] blog? He looks like how… He turns your life into boxes, how much time you have. We realize that there’s not that many two-week periods in a year. If you include sick days, holidays, weekends, all that stuff, there’s not that many two-week periods. If you want to run as many experiments as possible, you’re going to want to shorten it. You’re going to want to time box yourself as much as possible. It’s just to increase the quantity of the experiment.
David: I used to be so focused on pushing people on the quality of experiments. Then I realized that actually people already run quality experiments. No one’s running stupid experiments. Everybody talks about changing the color of the button on landing page. Nobody does that. Nobody really does that. Now I really try to focus on the quantity of experiments. I trust you, as a founder, as a head of growth, as a marketeer, to run quality experiments. I trust you to understand your customers. I just want you to run as many as possible, as fast as possible.
Louis: Okay. I think we’ve covered the basics in quite a lot of detail. Thanks for going through that with me. Really appreciate it. I think people really got a lot of value out it. I get quite a lot of emails from people who are either lost in marketing, they haven’t started. They want to use marketing, but they don’t know were to start. Or they want to become marketers, and they don’t know where to start. I’m curious, from your perspective, you’ve worked with multiple companies, you founded multiple companies, you have yet another successful company under your belt, what would you be your biggest advice for them, to get start in the marketing world?
David: Yeah. I think, look at your personality type. Maybe look at the OCEAN model; openness, conscientiousness, extroversion, agreeableness, neuroticism. Based on that, you can find what out what type of marketeer do you want to be. Do you want to be a number cruncher? Do you want to be consumer-facing? Do you want consumer products or do you want B2B products? The best way to do this always to find a marketeer you look up to, somebody that you really like, and try to reverse engineer what process that person went through. If possible, even contact them and ask them, “What were your first…” those steps.
David: There’s three/four types of marketeers nowadays. There’s still the brand marketeer, which is very good at storytelling, very good at conceptualizing, very good at creativity. Then you have direct marketing, which can basically be pushed into two. One is the pure hardcore performance marketeer, all about conversion optimization and short-sale cycle. Then you’ve got the data-driven marketeer on longer-sale cycles. Then maybe you have the content marketeer, that’s much more about generating content, building content.
David: I would say, look at your personality. Try to find someone who has a similar personality, who you really look up to. Try to reverse engineer what they do and which skills they actually have. The best way is just experiments. Just like Luke, our head of digital, the way he got into marketing was he was going too slow in his previous company. He just started play around with fake drop shipping, just to experiment with what he’d actually like. Then he found what he actually liked there. It’s experimenting at home, basically.
Louis: How do you…? Repeat for me, the different… the letter… the process to found out your personality. What is it again?
David: This is one of the most researched… It’s called the big five or the OCEAN model. I also think that this is very important for marketeers to understand. There’s a beautiful use case of behavioral psychology done by Sandra [Matt 00:54:28]. We’re not bad people as marketeers. The research she did, it really contrasts to decades of research reporting that there was a weak relationship between buying products, consumption, and happiness. It turns out that’s not completely true. Her findings suggest that you can increase happiness, if you buy the right things, based on your OCEAN personality profile. What they did is they used more than 76,000 bank transactions. Then they found that individuals who spend more on products that much their personality, tend to actually be happier in life.
David: I’ll give you an example. If you’re extremely high on conscientiousness, very long-term thinking, and you’re an introvert, you’re actually going to get life satisfaction from investing in accounting software or in legal advice. If you’re very much of an extrovert, a little bit less conscientious, you’ll get a lot of happiness and life… not happiness, life satisfaction, from buying experiences, from going on holiday or from going to the movies, for example.
Louis: Well, okay, so quite a good answer. I hadn’t heard of this action process and hadn’t heard of this research. Thanks so much. I’m pretty sure a lot of listeners will get value from that as well. What do you think marketers should learn today, that will help them in the next 10 years, 20 years, 50 years?
David: That’s a tough one. From a purely skilled point of view, we write a lot about this stuff and I still get lost in it. I think the data visualization is powerful. UX, also very powerful. Data science is definitely interesting, developing intuitive understanding of data science, but in 10 years… What were the timeframes? Sorry, I forgot.
Louis: 10 years, 20 years, 50 years, long time.
David: Honestly, I have no clue, in 10 years. I would say, more importantly, I would say it’s learning to focus and enter the state of flow. I think entering in the state of flow is more important now than ever. We’re so distracted. Knowing how to do deep work has become competitive advantage. Learning to read, learning to learn. Your competitive advantage is going to become deep work. Not many people know how to do deep, meaningful work anymore, because we’re…
David: It’s like trying to get deep sleep. You can’t do deep sleep if you keep getting woken up by somebody patting you. If you can get into this deep work, deep flow, then you can actually build something that’s more powerful than other than people are. That would actually be an edge, learning to learn, yeah, learning to learn. I would really recommend diving into this. There’s a book, I forget who it’s by, it’s called Deep Work. That’s interesting. Then there’s that book I recommended in the beginning, it’s called Stealing Fire, as well; and staying extremely curious.
David: Then on the hard skills, you’ll always need UX and you’ll always need something with regards to the customer, whether that’s through data and data science and analytics or maybe something more psychological, behavioral psychology. At the end of the day, marketing is a lot of behavioral psychology. The person who invented PR was Edward Bernays. He basically invented propaganda.
Louis: I knew you’d find something good for this question, so thanks as well. Maybe on the back of that, as you said, you shared a lot of resources on this episode already, but perhaps you could come up with three others that you would recommend our listeners today. It could be anything from podcasts, books, conferences, anything you want.
David: Yeah, actually I prepared this one, one second. Podcast, I would recommend the a16z Podcast. It’s the Andreessen Horowitz podcast. It’s amazing. It’s like having some of the best BC’s in the world as your mentor. The internet is a beautiful place. The second one is random. It’s a subreddit called Lectures. It’s basically watching… I like to watch random lectures on subjects that I didn’t know I would be interested in. It allows you to develop your lateral thinking. It’s a subreddit called Lectures. You’ll found out about the history of mathematics or the mating call of some weird bird in South Africa, really interesting stuff.
David: Then, of course, there’s… I need to do a little shameless plug. There’s our YouTube channel, where we share… It’s Growth Tribe YouTube channel. Just type Growth Tribe YouTube. We share a lot of sort of these tools, these tactics. We have specifically one called Growth Insights, where every month or every three weeks, we share the latest tools, latest tactics and resources, in a condensed, very fast manner. Those would be the top three. I’m not going to reference any more books, because I know you guys aren’t going to read them anyway.
Louis: David, once again, thanks for your time, really, really insightful episode. I really appreciate, especially the fact that you shared a lot of resources from different angles, different stories. I think people listening really got a lot of value. I genuinely mean it. How can they connect with you and learn more from you?
David: Yeah. Connect with me on LinkedIn. I’m really, really focused on LinkedIn at the moment. You can just connect with us, Growth Tribe, on Instagram as well. We’re also hiring, so come and check out Growth Tribe, growthtribe.io. Come and check out the website. We have like 22 positions open at the moment. If you’re smart, come over.
Louis: Nice one. Thank you so much.
David: Thanks. This has been great.
I’m a no-fluff marketer living in Dublin, Ireland (but yeah, I’m French).
I believe you can treat people the way you’d like to be treated and still generate results without using sleazy, aggressive, hack-y marketing. This is why I’ve started Everyone Hates Marketers – a no-fluff, actionable marketing podcast – as a side project in April 2017.
I’m also the Content Lead at Hotjar – a powerful way to analyse people’s behaviour on your website or app and understand how you can improve their experience.