Damian Bergamaschi: Welcome to the Inevitable Success Podcast sponsored by BuyerGenomics, where our goal is to help you, the marketer, make success inevitable. Each episode will discuss the craft of data-driven marketing, helping you uncover new and profitable ideas. You will also learn what works and what doesn’t work from top marketing professionals and thought leaders. I’m your host Damian Bergamaschi and inevitable success starts here.
So, as normally happens, Stephen and I were discussing various experiences or just thinking about marketing in general and Stephen started to tell me about his experience with Waze and I thought it was so interesting. Waze, if you’re not familiar with it, is a navigation app which uses data, algorithms, and crowdsourcing to figure out how to get somebody from point A to Point B in as efficient a manner as possible. And the thing that was so interesting about it that ties back to marketing is that there is an experience of trust that had to happen where now Stephen uses it all the time when before he didn’t. So why don’t you tell us about your Waze experience?
Stephen: It came out and it was an intriguing product. Well, first of all, it was free. So why not? It was a good value proposition. So I downloaded it and I started comparing that with any other type of navigation tools. The one that came with the car or the Google and then I realized that, oh wait, this is a two-way communication now. I am, as a driver, I’m a contributor to this app. I can say, “Hey there’s a cop hiding behind a tree or there’s traffic, or how slow is it right now?” All those things are my input. So this is a really amazing crowdsourcing mechanism happening in real time, too. And it will help me to avoid traffic reported by somebody else. How noble an idea is that? But in the beginning, I had my doubts. I know I continue using it every day and I’m thinking, okay, Waze is making me taking a left turn and going to the Bronx. I’m going through midtown. Why is it telling me so? And the reason why I bought into this idea is that the cost of me trusting my gut feeling was so high that every time I didn’t listen to Waze I regretted it. It was mostly right and I was mostly wrong. So, therefore, I started trusting the data and the algorithm behind it.
Damian Bergamaschi: So that is basically when you said that, that started this thought process of what would it take for a brand to have that same kind of reaction with a customer, right? Where when a brand sends a message to a customer, this almost “I don’t need to think about it because I’ve already kind of pre-vetted this situation.” This utility is the value proposition. I know this is the best.
Stephen: That’s right. Right.
Damian Bergamaschi: So how do we get there? One of the ideas that we had when we were thinking about it was you have to build trust and ways to use data to really build trust — specifically by avoiding being wrong or avoiding sending out communication that would send me on an errant mission.
Stephen: Let’s reverse engineer what they do.
Damian Bergamaschi: Sure.
Stephen: First they collect the data. Traffic data come from people who point out, “hey the road you know is there already.” They’re not going to build new roads, but if I point out a road here it updates its database, which is fantastic. That starts with the data. Then there’s an algorithm. Now, an algorithm is complex. In other words you have to build a lot of algorithms so now I’m going from A to B, there are about 20 different options or if you’re going to, you know, go from Manhattan to New Jersey you could have a hundred different options. I mean, deciding whether I should cross Lincoln Tunnel or GW is a big decision isn’t it? Because you could end up being in traffic longer than if you’d have stayed on the GW path.
Damian: Right. And if you have a Tesla then you have to also have a route in a way to get the car charged.
Stephen: And by the way, they also have an option to find a gas station on the way.
Damian: Somehow we like to try to work in a Tesla reference to everything else but they’re not compensating us.
Stephen: The algorithm is important that they’re not out and also important and they have to present the way. In other words, it has been a useful form for the driver in the car that off he’s telling me that I’m going to make a left and a million ways without waste. And how do I get that message? They’ve got to have that. So these all these things come together really nicely and they even have a feedback routine. They know they’re measuring every move that I make. And when there’s traffic I’m going to report it, too. This is all a little bit scary for the future of humankind.
Imagine a situation where there’s really bad traffic on Road A. And Waze wants to send most people to Road B. If you do it too much then Road B gets congested. If the machine is smart enough they don’t send everybody to Road B. So you know what that means? The machine is making a decision for you. We’re literally forfeiting our ability to think at this point. One of the reasons we’re talking about this today is how can we do this for retail? How do we reverse engineer this process? The retailers can do the same thing. I mean, I read an article just this morning that AI is not far away. It’s not just reserved for, you know, Amazon and those guys. It should be for everybody because you know why? If you don’t do these things not only are you going to lose out you may go out of business in the world where everything is suggested by the machines. And let’s think about what the gut feelings are for the retailers now, not the buyer, but the retailers.
Why did you order a certain set of products to sell next season? Because your gut feeling told you so? Because your designer says so? That’s a brain-powered one. What if you can have access to a lot of data? Past and present shopping behavior of all these people, not just by you but a lot of other places like Waze does when all the drivers use it, right? Would you trust your feeling and your thoughts or algorithms backed by the data? More and more people are doing it with data backed by the algorithm so therefore you cannot just avoid this topic because this is happening. And I’m not saying that this is all good for humankind by the way. I’m not saying that at all.
Right now I’m going to just hold my judgment about that because this may result in mass loss of jobs. But if you flip it around then if you’re a small retailer who’s not Amazon, so they have no chance to lay you off because you’re not part of Amazon’s system or even if something is replaced by the machines, maybe it’s an opportunity for you. If you’re small or medium size, and you’re thinking that you never had access to such technology before. And you thought that you should hire a lot of people to do this. Now you tell me I don’t need people because if there’s an algorithm backed by that data, then I could do this? Then it’s an opportunity if you think differently.
Damian: There’s a lot of retailing business models that are popping up recently that are very conducive to this whole idea of almost suggesting something for people. For example, I think of Stitch Fix or I think what Amazon Wardrobe may be trying to do where it has like a similar flavor to this whole Waze thing. That if it’s consistently suggested things that weren’t what you wanted, you’d pretty much tune that out. But if you got it right and because you had some sort of real-time data coming in, whether it was crowdsourcing or looking at transaction behavior or things it sold effectively, you’d be able to serve the right thing to people. And over time you’d just kind of defer to this decision maker.
Stephen: That’s right. If you build that trust if you’re right most of the time.
Damian: And you know in some ways I feel like the cost of being wrong is probably, you know, almost more of a thing to avoid than being right. Like the saying, you know, it takes like a lifetime to build a reputation. It takes a moment to ruin it. We talk about batch and blast (email campaigns) a lot, and ways to get it right.
Stephen: People don’t think about probability. It’s like this. Even the clock stopped completely is totally correct twice a day. So you’re not just measuring accuracy, how right are you, because the clock is right twice a day. It’s not how right we get. How wrong are you? I say this in the analytics business all the time. Sometimes you don’t know 100 percent. But the cost of being wrong should be lower. Now, why is batching bad? The idea is that is one way to get everybody, and out of everybody, somebody is going to buy. But what’s the cost of it? The cost of it is a ton of people I’m training to ignore my messages because I’m sending emails with the same message every day. And everybody knows I’m sending the same messages and irrelevance shows them that, “Hey, this is not relevant to me. Why am I getting this? I buy a lot of other things from you. Why are you showing me all this women’s wear?” So we’ve got to really think about the cost of being wrong and to improve the probability of being right. You’ve got to rely on data and algorithms. We cannot avoid those things anymore.
Damian: Yeah I completely agree.
Stephen: So what we want to talk about is that then. Okay, so fine, Amazon is doing this, Google’s doing it, all the big boys are doing it and finding things that those guys don’t even call big data, they don’t even use those words. How do retailers catch up with that, aside from how you can catch up with those guys? I see a lot of articles where you should just buy into this, but with AI there are some prerequisites that you have to get ready for. So if you think about the process here, and this is why I broke down the whole situation, you collect the data mine through crowdsourcing and have a preexisting match. Whatever it is, you have to have a kick-ass algorithm so you are right most of the time. If the algorithm isn’t wrong by the way then I may have to go to Queens by way of Long Island.
Damian: This is where I’m going to interrupt you for a second because avoiding being wrong and focusing on being right are actually a little bit different.
Stephen: Of course.
Damian: I’ll give you one example of a company who’s a master at this: Google.
Damian: Google is very, very focused on making sure that wrong things do not show up for the search intentions, more so than they are about serving the best particular thing. Right? So that is actually kind of like almost a flip in thought. Like, for example, if you turn to the trusting one that we can all understand, right? Always showing up to where you need to be is par for the course. When you don’t show up that’s when the problem occurs. Right? Because now you’re like the guy who doesn’t show up. You blew it.
Stephen: I am saying that for most retailers the bar is pretty, pretty low. Let’s not try to be Google overnight.
Damian: I mean, give yourself a week….
Stephen: Yeah the reason why I’m saying this is that most retailers just batch and blast. They don’t even target the price. Just different products. And by the way, for the batch and blast people, fine, maybe they see that holding out somebody is a lost opportunity, but they are not thinking about the risk of turning people away. Right? If you’re sending emails to everybody all the time, fine, great, do it. But are you at least pampering your recipients in a different way depending on who they are? Are you sharing different offers? Because we do this all the time at BuyerGenomics: If somebody is new make them feel excited. If they’re the most valuable buyer then pamper them and don’t give them discounts easily. They love you already but if they’re far gone they’re fading away then you do something else. One little thing changes a lot of things.
Now let’s extend it further. We can do things like different product suggestions or use different channels. So, therefore, the bar is kind of low because you’re coming from a world of batch-and-blast and just sending everything and let them think what they like. Instead, I’m going to do some suggestions in my emails and then see if I’m right and increase the probability of conversion, if not guaranteeing the conversion. And Google, by the way, guarantees that my search engine is always right.
Damian: Also I think we have to institute almost like an anti-conversion metric. So let’s put it this way: I love analogies to kind of make things more clear. Let’s say you had a medication that you were testing on a sample and you only literally measured like the way that we measure batch-and-blast campaigns, then the analog is how many people got better? But you only measure how many people got better, well, then sending a whole list is going to give you good numbers. If you don’t look at the context of how people got better along with how many people got worse, you can put yourself in dangerous situations and erode trust So I mean it’s such an important concept and it’s a metric that we have to kind of build into it, like what’s a conversion rate? I’m going to give it a name: the anti-conversion rate.
Stephen: You’re saying that everybody should raise their bar. In fact one of my mentors that’s the Wonderman recently passed away rest his soul, but he said this long time ago he’s the father of direct marketing for heaven’s sake. And what he said long time ago that yeah we developed a lot of techniques so that we can realize out of a mailing of like 2 percent conversion rate which is very high in our online world, If you do the conversion rate in the right way which is based on how many millions you blasted I see numbers like 23 percent. Okay. Do you know what that means if you achieve 2 percent conversion rate? You are 98 percent wrong. And we rejoice because you recover the cost of mailing? You know what? The bar should go up, especially in the world where you do batch blasting and your conversion rate is hovering at 1.3 percent. And by the way, let’s not do that based on some clickthrough rate. I’m saying if you look at how many emails you’ve blasted you’re training people to ignore you.
Damian: Right. I think that 98 percent, a lot of that is just ignored or thrown away and ignoring a mailing is different, but in the email situation without a good anti-conversion metric it doesn’t capture all of what you could know. You know I’m kind of done reading this guy’s e-mail and by the way, it’s an effort, isn’t it?
Stephen: For a person to go in that direction, then what you’re doing is you’re really really wrong. So what I wanted to leave behind is just the thought that you should really think about this and raise the bar. You don’t have to be Amazon. I have to think differently. And to do so let’s not forget that it starts with the data and algorithm. When the data and algorithm come together it has some exclusive value. As we saw in Google, as we saw in other ways for example. Now when you do that then the trust goes up because you’re mostly right. And that is what we want to do. You want to increase the probability of conversion if not guaranteeing.
Damian: Yeah. And I think in closing, what more virtuous way to use data than to build trust?
Stephen: That’s a really good way to think about it. Whether you do it for marketing or supply chain and analytics or marketing optimization work. In fact, I recommend a book called The Numerati. It was written a long time ago, but it’s a very interesting way of looking at the data itself. They use examples from real life and don’t use any statistical terms, so please don’t be scared. And they have chapter names such as shoppers, lovers, terrorists, workers, and such things. There are all those behaviors that we leave behind and these leave some kind of a data trail. And if you analyze them right you could predict who’s going to fall in love with you. How do you think all those dating apps work by the way? It’s data and knowing how to read it. Why do you trust that dating app on the Internet? Same reason why you trust all the data.
Damian: All right. Well, thanks again. Till next time.
Stephen: Thank you.
Damian: If you enjoyed today’s episode, we ask you please leave a rating and write a review. Or better yet, share it with another marketer. Be sure to subscribe to the podcast for new episodes. Also, check out the show description for complete show notes and links to all resources covered in today’s episode. If you’d like to speak to someone about any topics covered in today’s episode
please visit BuyerGenomics.com and start a chat with the BG team today.