Scott Graham is the Vice President of Marketing and UX at Carjojo. He is also the founder of Mimmer and Top Growth Hacker. His cross-discipline strategies utilize engineering, behavioral, and cognitive design, and innovative marketing techniques to increase sales, customers, and viral referrals. He also has a passion for bacon and bourbon.

Follow Scott on Twitter


Episode Transcript

J Cornelius: When's the last time you bought a new car? Did you enjoy the experience? Probably not, 87% of people don't. Silicon Valley startup Carjojo is aiming to make it better and we're going to talk to their VP of Experience about how.

Intro: This is Design Driven, the podcast about using design thinking to build great products and lasting companies. Whether you're running a startup or trying something new inside a Fortune 1000, the tools, methods, and insights we talk about will help you create things people love. And now, your host, J.

J: Hey Scott, welcome to Design Driven, we're happy you're on the show today. How are things where you are, out in sunny Silicon Valley?

Scott Graham: Well, they're a little cloudy right now, but you know, it's pleasure to be talking with you and I hope Atlanta's warmer than it is here right now.

J: It's actually pretty cold here, we were in the 30s last night, believe it or not.

Scott: Okay, so you guys win.

J: Well, not exactly the contest I'd like to win, but I'm sure we'll improve here shortly. So, for those of you who aren't familiar with Scott and his work, he is the VP of marketing at Carjojo, a cool car company or car purchasing company out in the Valley. Scott, tell us a little bit more about what Carjojo does and who they do it for.

Scott: So, Carjojo is essentially for new cars. We pull in about 3.9 million new car inventory every night, analyze that in real time, and understand what is the lowest possible price that we think a dealer will accept for any particular car that you're looking at.

J: Is that every car in the country?

Scott: 99.8% of all new cars currently on lots in America, we have in our inventory.

J: Wow, that's pretty impressive. So, we've seen some other car related or other car purchasing related startups happening, Carvana comes to mind, so how is Carjojo different?

Scott: Well, in a couple of different ways. So, first, all of the companies like Carvana, Edmond's, or True Car or even Kelley, all work with the dealers, so they all plug into the back end of the dealer data systems. So, they end up only having certain partnerships with certain dealers. They're inventory level for new cars usually hover between 15 and 20% of the national inventory. And the prices that they are offering are lock, step, and key with the dealer, so in fact, many of our competitors offer prices that are above even what the dealers own internet prices are. Carjojo does is a completely different way, we don't work with the dealers, we work for the car buyers. Meaning that, when you come to Carjojo, you're getting unbiased information about the marketplace and about our true value of what we think the car will sell at any particular time.

J: That's pretty interesting. So, what has been the challenge for you in helping to communicate that to end user. Like, what's their experience like and what's your process of thinking through that and can you talk a little bit about how you understand the user's needs and make sure that you're meeting them?

Scott: Right, well the first and foremost piece of data that we saw was that 87% of people hate the car buying experience, they love cars-

J: - But that number seems low, only 87?

Scott: I think I saw some data that said that people trust cars salesmen less than they trust lawyers and bond salesmen and loan officers. So, it's an industry, it's well deserved, most of the marketing practices are terrible. The experiences that we hear from dealers can be quite troubling at times. And you also have this thing where you only see in a car market where you aren't really sure what the real price should be. You have MSRP's, you have invoice pricing, you have what the dealer says they can sell it for, then when you come in they say oh, we have this deal and then, on top of that, you have all these rebates that are coming from the manufacturer or from the dealer and how does that all play together. So it's a really confusing environment for the consumer and it's one of the reasons that it is so loathed. So, we're entering a behavioral environment where at the very least, we know people don't like the current process. So, now it's a matter of us to look at that current process and say, how do we improve upon that substantially, how do we create user flow paths and a real business model that fits into a way that people can one, understand, and two, feel good about, to feel confident in their purchase decision, feel confident in the price they paid for the car.

J: Yeah, that confidence is important. I know that one of the things that we look at in our work is helping people on the product design side understand the mental model and kind of the position that the consumer, the user, has when they're entering a process and how they look at the process and have some empathy around that. So, what kind of tools are you guys using to figure out what people want and how they're feeling at different points of the process and then how are you adapting the things that you're building to compensate for that or to address it?

Scott: Right. So, we have a number of tools, both on the product side, and on the marketing side. On the marketing side, we have launched a number of focus groups where you sat down with real customers, people in market, and showed them, not only our product, but competitors’ products and asked them what were the first takeaways, what was the message they got across, what was the feeling they got across from going to these sites. And sort of worked that thought processing into our product and our messaging strategy. On the product side, from the moment you see an ad all the way through conversion, we're testing multiple types of messaging, trying to craft one to one messaging as much as we can throughout the flow and always AV testing it. So, when someone comes and sees an ad based upon they're looking for a certain car. Our landing page has multiple different variance on it that they see next, it says, "Go see this Honda CRV." And we break out those headlines, those calls to actions, those incentives for continuing, one both to get people to come march down the funnel, but two, also, try to educate them in kind of little breadcrumbs. At Carjojo, because we are a new company and a new way of perceiving the car buying process, we have to educate the customer while also not diverting them from where we want the end goal to be.

J: Right.

Scott: So, it's a continual learning process for us and measuring where we do good and where people tend to be falling off. And taking that into account by looking at our tools, we have a pretty robust Google task manager set up so we can see pretty much any computer action on the front end and quarter back from where they came from originally on the ad side. To understand what is this user behavior and how do we optimize our sales to get better every day.

J: Yeah, that's interesting. So, by looking at behavior on the site, that's helping drive design and experience decisions?

Scott: Exactly. In fact, on our backend database, we track people, not just in single sessions, but their entire lifetime across multiple devises. So we see, throughout that customer buying experience, how many times they're coming back to the site, what causes them to come back to the site, what kinds of content are they viewing when they come back to the site. And on average, our active engaged users, during their car buying experience, are spending 47 minutes on average on our site across multiple sessions. We’ve seen huge increases as we've tailored this messaging more and more down to buyer level. In fact, a couple of months ago, we had one individual who broke the record and spent 90 hours on the site in just over six weeks.

J: Wow. That's a lot of site time.

Scott: He did become a paying customer though, so it paid off.

J: Yeah, it paid off, wow. So, you know he didn't just forget the browser tab and leave it open.

Scott: Correct.

J: So, when you tailoring it down to the customer, how detailed are you going so each individual customer has their own tailored experience?

Scott: So, that's our end goal and we're working our way towards that. You know, as a startup, you can't go and deliver one to one very easily right off the bat. But knowing where we want to go is helping us craft how to get there. So, we know that a buyer in Texas looking for a Ford F150 is a much different type of person than a person looking to buy a Prius in San Francisco. So, how do we tailor our brand, our experience, our messaging to each of those types of personas and how do we do it in an authentic way where we're not just spamming people to get an end sale, but to actually help them through the process both on our technology platform and in just their own buyer's journey.

So, whether that's saying hey we saw you were looking at a Ford F150's, just so you know Chevy Silverados are having a really killer rebate right now, if you're interested and open to checking out different types of trucks, that might be a good way to save a few thousand dollars. So, we're building in those experiences, both on our platform, we're implementing and installing Marketo in a couple of weeks, that's going to help us engage that traffic both across social and mobile and of course, email automation. So, we're bringing in a lot of extra tool sets to deliver that one to one marketing messaging and that unique customer flow that's never the same across any one particular person.

J: Yeah, that's super interesting. I'm glad you mentioned personas because that's a really powerful tool in helping people understand across both design, marketing development, like who you're building something for. Can you talk a little bit about what your process is in getting those different disciplines within Carjojo to line up and everyone kind of pointed towards the same goal?

Scott: Right, so, at Carjojo, the front end of the design product experience and the marketing team are one in the same. So, it helps us really practice strategy end to end across the entire customer experience that is tailored towards making sure that this person has an unbelievable experience with the product and the brand. And when we're kind of thinking through the personas and building those out, first we started with big buckets. So, we said, okay, when is it when you're buying a car in your life? There's, well, your first car maybe when you're 16, that's moving later of course now with millennials choosing Uber over drivers licenses, but okay. So you have your first car, then you might have your college car, then you have the first car you buy once you're in your first job, you have the car you get when you get married, you have the car when you have kids. All these are different life moments and they really tailor back to when you're buying a car and what type of car you're buying, you know. When you're 22 and you're single, you might buying a smaller car, a more fuel efficient car, a convertible. If you're working in areas that require moving lots of materials, it might be a big truck.

But we can start to tailor down those personas and block them off into, first in big chunks and say, okay, now we have an idea of people tend to buy a big car when they have their first kids, what is it that makes that person buy either a minivan or a SUV. Is it gender, is it relative age, is it income? So, we start to break out what factors of that persona cause them to lean one way or another? And the more fine tuned we can get, the better messaging and better product experience we can have overall for that customer 'cause we're providing the most value possible tailored towards what our data says they're probably looking for.

J: Yeah and so that gets into another thing that we use a lot is psychographics, not just demographics and behavioral data, but the psychographics of decision making and how this person's situation in life and they're all those other data points, fuel the way that they make decisions around buying a car or where to go for dinner or those types of things.

Scott: Right.

J: So, it's interesting to see you're using that in a way to assist them actually finding the right vehicle at the right price.

Scott: Exactly, I think you hit the nail on the head. So, it's not just about getting the best price, it's really about finding what is the perfect vehicle with all the options that you need and that you're looking for and your type of lifestyle that's going to fit you best. And when we're looking at our competitors and our competitive landscape, the fact that they have to bounce between four or five different big websites and then go also to the dealer websites, you can start to begin to cover the inventory that's available to you and seeing the cars that you could potentially get into, was an immense problem and that's why just pulling that information into one location and allowing people to search across nearly every new car is addressing a big pain point.

J: Pretty powerful. So, with all the complexity with trying to customize the experience with each individual person, what is your process like for deciding what to change on the website and then actually getting live and then measuring if that change did what you wanted it to?

Scott: Right, so, great question. We are actually in the process of moving from ... we're launching our second iteration of our product, a really big refresh here in just about ten days. So, a lot of it was collecting the insights from the original product and saying what's the user feedback we're getting, where are people getting stuck, what are they getting confused about. And we started to implement those thoughts and ideas and feedback into the new version of the product. And throughout everything we do, we're always collecting data on the behavioral aspects of where people are getting trapped. And again, it was utilizing things google task manager and our systems allow us to see that. And when we decide, hey, let's try different button copy or let's try different headlines here, let's try changing the way people checkout. At some point, we had testing going from checking out with three cars to checking out with one car and we found a really interesting result on there, we said, oh, that actually made it so the shopping experience was less robust. So, you find these surprising things by kind of just A/B testing the small things and the big things and working with the technology team to find the best way to implement those tests and pull those levers so we can measure it in real time.

J: Right, so it sounds like you're able to take an idea from the discussion at the conference table to actually being live and tested on the website within a short period of time.

Scott: Yeah, that's all just working closely with our CTO and our technology team. In any sort of technology enterprise, startup or big, it's all about having great ideas, but also having a team that can cement them and execute them and provide the feedback that the non-programmers of us need to receive. It's a partnership towards a better end goal. It's a partnership towards success and that really just involves how do you create a great team and how do you work together to achieve the end result.

J: Yeah, I agree completely. In that process, are you doing any prototyping? Are you doing paper prototypes or you using prototyping or wire framing tool? Where do those decisions get made?

Scott: So, we have a couple of different way that we get prototypes. We have high def prototypes we'll go in Adobe Photoshop or Adobe Illustrator and make really fine-tuned copies. And we'll take those, whether they're in iPads uploaded to our vision app where we can do click throughs or just there's one page and we're trying to get feedback, where are people's eyes going, what's the first thing they take away. And we're taking those two focus groups, whether those be formal focus groups that we use firms for to gather people for us or sometimes it's just, hey, let's go down to Starbucks and meet with some people and see what they think and give them a $5 Starbucks gift card and get out of our heads for a little bit. It's always good to have the formal things, but formal focus groups take a long time. They can heed the quickness that you want to develop. Sometimes you've got to take it into your own hands and say ... it may not be exactly scientific, those focus groups that I'm doing at Starbucks down the street, but it's better than marching ahead without consulting potential customers, consulting people outside your business walls.

J: I'm glad to hear that you go to the coffee shop and test these with real people because that's something that a lot of development teams and product teams just completely fail to do, they just don't get out of their own office. There's the old adage of get out of the building, right? That's the way to figure out if what you're building is going to be on track or not, so I'm glad to hear that you're doing that.

Scott: Right. Coffee shops are great way to do it, bars not so much because you end up just staying around the rest of the day.

J: And the feedback you get is obviously skewed.

Scott: It better be real easy, right?

J: Exactly. So, actually there was a guy some time ago that was, I think he called it or something, you paid him $20 and he would get drunk and test your website and tell you if he could use it drunk.

Scott: That sounds like a business model I could be a part of.

J: I could see some use cases for it.

Scott: Oh yeah.

J: So, when you're going through those prototypes, what kind of decisions do you try to make at different levels of fidelity? Like, are you trying to make certain usability decisions at low fidelity or what is that process? What's kind of your approach or methodology there?

Scott: So, with this product in particular, what I found is, it's better to have a little bit higher fidelity in your prototypes. You can do blocks and you can do just gray boxes, but I don't think the average-

J: - Do you mean for internal or for testing with customers?

Scott: For testing with customers. It's better to have a little bit more high def for the customer test with this product in particular, primarily because just having boxes or stand in components that say, insert graph here, it doesn't have the same resonance or impact with their buying decision. Buying a car or like buying a house is a very ... it's one of those few buying decisions where people are looking for a lot of data on the consumer side. They want to see what the other prices are, they want to see that they're making a good choice. So, we try to do is make sure that we are providing that data and that it looks like credible data.

The other obstacle that we hit all the time in this industry is that credibility of any automotive website is, especially one you've never heard of, is usually pretty terrible and that's because most automotive websites, you give them your email, they start spamming you or they know the price that they're looking at isn't the real price, the real price is only when you go into the dealer. So, it's building that credibility and that authenticity with customers that's important. We really try to reflect that in the mock ups that we create and that they feel comfortable with what they're looking at and so that also we can compare apples and apples with what's out there today, what's out there on the Kelley's and the Edmond's of the world.

J: Yeah, right. So, have you been tracking any of the metrics around how quickly you can get an idea into the market and how that's impacted the growth of business or how's that impacted maybe your efficiency or just ultimately increase the bottom line of running the shop?

Scott: Yeah, so, I would say on average, depending on the complexity of the test or the new feature, we're usually going from concept to deployment in under six weeks.

J: That's pretty quick.

Scott: Yeah, we run pretty quick, the only out light of that is the complete product refresh. We're really taking a lot of different feedback. But other than that, it's been really quick development schedules and again, our CTO does a great job of one, helping us concept and two, deploying that really quickly. So, getting that quick learning is imperative to being able to be a great startup. It's the ability to grow fast, learn fast, and fail fast. The last thing you want to do is invest three months and then fail.

J: Right.

Scott: Or fail really hard. So, as much as we can, we try to bite off small chunks as we go, see what's working, see what's not working, and then, continue to improve. And on the result side we had, we've seen phenomenal results. Our cost per a converted customer is down something like 70% in the past six months. Just in doing this type of testing. So, making sure our messaging is right, making sure the flow patterns are right, making sure that how people are using the product, aligns with the end result we're trying to get them to do. And a lot of that again, we've taken a lot of that feedback and there was somethings that just didn't make sense to try to change without changing two or three other things and that's why these bigger refreshes have to happen. But we did so in a way that we were really tackling a lot of different parts of consumer feedback all in one go. So, conglomerate it would actually be less than six weeks for a big feedback item that we get in this major product refresh.

J: Right, and when you add it all up, it obviously takes a little bit longer.

Scott: Yes.

J: Cool. So what would you say, if you're talking to another product manager or VP of product somewhere, what would you say is your most valuable advice, what would you tell a younger person in your own shoes about how to use some of this methodology to grow their business?

Scott: Well, I think the first thing that's really carried us, not just in Carjojo, in terms of the success that we've had over the past six months, but I think in my entire career, which is always think creatively and always think in a way that if somebody's already doing that, you can learn from it, but it's unlikely that you'll ever overtake them by repeating the same process.

J: Right because they're ahead of you already.

Scott: They're already ahead of you. So, it's not about just looking at what other people have done and implementing it yourself. Real success comes from taking an original idea and making it ... going through this process of human centered design and then discovering that golden nugget, that diamond underneath all the things that you're working towards, that ends up creating results that are a multiple of what you're doing before. And you never know when it's going to happen, you're never sure this is going to be the paper cut that drives us from a two person shop in a garage to Google. But enough of those conglomerate of kind of big wins that you uncover are how you get there.

J: Right, they add up.

Scott: Yeah, again, I think it's about being aggressive, being confident in yourself, and be willing to test things that aren't happening and somewhat going with your gut a little bit and somewhat listening to customers and finding what their pain points are.

J: Yeah, it really comes back to just listening to customers and understanding their pain and getting a really good idea of how you can alleviate that pain or how you can bring them a win and working with advocates for the customer.

Scott: Correct.

J: No, I agree 100%. So, if someone wants to get in touch with you and learn a little bit more about Carjojo or just kind of pick your brain about stuff, what's the best way to get in touch with you?

Scott: Great question, anybody can reach out to me directly it's I'm also at Twitter @scottgraham9 and I have a pretty large following on LinkedIn. So, whatever your poison is, I'm probably there.

J: That's great. Scott, thanks again for coming on the show today, I really appreciate your insights, good to hear what's going on out there in the Valley and we'll have you on the show again in a little and hear how that V2 of the website works out for you.

Scott: Looking forward to it and thanks for having me.

Outro: That's it for today, thanks for listening to Design Driven, we're glad you enjoyed the show. Have comments, questions, or an idea that you'd like us to cover? Point your browser to and click "contact us" on the top of your screen, we'd love to hear from you. Tell your friends and colleagues about the Design Driven pod, post on Facebook, Twitter, LinkedIn, or send them an email and tell them to go to or wherever they find their podcasts. Until next time, remember what Thomas Watson, founder of IBM said, "Good design is good business."