Mapping the Future: TomTom’s Evolution in Mobile Navigation and Location Data

[00:00:20.950] - Olivier Destrebecq
Welcome to the show, everybody. Today, I have Pim Spaanderman and Kees Oostinjen. They're hard-to-pronounce names for me because they're from the Netherlands, and they're from TomTom, and they're with us live today. How are you guys?

[00:00:33.110] - Pim Spaanderman
Doing good. Thank you. Thanks for having us.

[00:00:35.290] - Olivier Destrebecq
Welcome to the show. I have Jeff, as usual. How are you doing, Jeff?

[00:00:37.920] - Jeff Grang
I'm doing great. Very happy to have TomTom around the table today. Two people, I think that's your first one, Olivier.

[00:00:43.560] - Olivier Destrebecq
It is a first. It is a first. We'll see how this goes, and maybe we'll do it again. We'll see. Both of you have been with TomTom for over 10 years, from what I see, so you probably know the ins and outs. I'm going to start with you, Pim. You have a background in e-commerce. You came from the telco industry originally. For the last two and a half years, you've been the Managing Director of the Consumer Business Unit. Can you tell us a bit more about what that entails?

[00:01:09.170] - Pim Spaanderman
Basically, I'm responsible for the consumer business of TomTom. The business where TomTom originally started off with selling satnavs, bringing those to the market. Now basically the business also moved on into different territories like automated driving and traffic information for cars and all those things. It's really truly a location data company, how we call it right now.

[00:01:35.400] - Pim Spaanderman
But there's still a consumer business, and the consumer business still entails satnavs that we still sell approximately a thousand per day. It's still a business, especially in the truck market, for instance, where there's a lot of PNDs, how we call them being sold. But also app business. Also, our apps that basically nobody really knew about a couple of years ago, we started pushing towards our consumers as well.

[00:02:02.010] - Olivier Destrebecq
Awesome. Kees, you actually have been with TomTom longer than Pim. I read 12 years here.

[00:02:07.540] - Kees Oostinjen
Yeah, just a little.

[00:02:08.520] - Olivier Destrebecq
Congratulations. You do have a background in sales operation analytics. You moved to the apps at TomTom about 5 years ago, if I'm not mistaken. I'm curious, what were you doing before the apps?

[00:02:19.430] - Kees Oostinjen
Before I started working on the app, I was in sales, sales operations and sales analytics. I joined TomTom 12 years ago, straight out of university, basically. My first job was as a sales analyst. I was analysing the markets and then our position within those markets and how our position was developing, et cetera.

[00:02:38.690] - Kees Oostinjen
From there, I moved on to Sales Operations Manager, where I became a key contact between the headquarters and the local teams responsible for forecasting, line-ups. From there, I moved to the apps business, basically, which was around 4 or 5 years ago when we started looking at how we're going to position ourselves as an app, a navigation app developer, and what are the opportunities for us within that market.

[00:03:02.830] - Olivier Destrebecq
Cool. Before we moved to the apps, actually, I've known TomTom as a brand for, I don't know, maybe 20 years. I want to say that when I was just starting to drive, all my friends were buying GPSs, and they were TomTom's, obviously. Could you give us a little history of TomTom over the years and up to moving into the app space?

[00:03:25.630] - Pim Spaanderman
Yeah, a lot happened. The company was originally founded in 1991. Basically, the forefounders constructed software for handhelds, PDAs, if you still remember those.

[00:03:38.520] - Olivier Destrebecq
We do.

[00:03:39.210] - Pim Spaanderman
Before the smartphone territory, we had the personal… I don't even know what is the PDAs, that I don't know. But remember those handhelds with the agenda in there, et cetera, and you can crop-

[00:03:49.820] - Olivier Destrebecq
Personal Digital Assistant.

[00:03:51.640] - Pim Spaanderman
That's where the company originated from. Then they thought, let's bring navigation software to those PDAs. Then they made the leap, let's make a standalone device hardware, specifically only for that software, for the navigation software. That was the TomTom Go that came to the market in 2004, which was the market mover. They made the market, basically.

[00:04:14.930] - Pim Spaanderman
At that point, people call basically the industry also the brand name. People talked about a TomTom being a navigation device, and a navigation device is a TomTom. That's how strong basically the company was at that stage. If your brand name is the category, then you're really strong. The founders always thought that it's an in-between solution. In the end, this software needs to go into the car.

[00:04:41.140] - Pim Spaanderman
Already, when years passed by when they first introduced that PND, how we call it, the personal navigation device, they already started investing into onboard systems and being inside the car, talking to the automotive industry, et cetera. We now have a flourishing business in the automotive industry where we sell our maps, but also traffic information or all kinds of components, technological software components, towards the automotive industry, but we also sell it in the enterprise business these days. Companies like Uber. All the Uber apps basically are built on our technology as well. They're a big partner of ours.

[00:05:20.640] - Pim Spaanderman
We're in various corners of the world. We do business. We're a company now of approximately 4,000 people, and really a software company. We were the first big data company, big tech company in this space. We were one of the three big companies that built maps. You got Google, you got HERE, and you got TomTom. We build the digital maps out there that are being used by all the companies. We're taking a leading position there.

[00:05:51.190] - Olivier Destrebecq
Awesome. I got one related question to Uber being a partner with you guys on the map. This is not a sales call, so you don't have to give me the numbers. But what's the differentiating factor between using the TomTom map in a mobile app and, for example, using the Google Maps?

[00:06:09.360] - Pim Spaanderman
There's all kinds of different factors. Maps are continuously moving element. We used to build a map, and we think that it's a static thing. We printed it and that's the truth. But everybody knows that continuously there's things moving around us, new buildings, buildings going down, construction happening, et cetera. As a company, as a map builder, you want to move towards a real-time map. Nobody is even close yet to being the provider of a real-time map. But TomTom is trying to construct to getting the most narrow towards that point that you're a real-time map and that you're always able to provide the most accurate data also at the most accurate time for your customers.

[00:06:53.390] - Jeff Grang
Moving from a specific PND device to a mobile app where there's a lot of competition today, and especially a lot of free competition. You said maps, but there's Google Maps, Apple Maps, Waze, and all these free alternatives. Why would someone, a regular consumer, pay for TomTom?

[00:07:06.670] - Kees Oostinjen
Yeah, that's a fair and understandable question, of course. If you look at the navigation app landscape, then basically that's a market of billions. Every mobile phone has a navigation app installed on their phone. Most of the time, that's obviously a free app.

[00:07:20.800] - Kees Oostinjen
But I think what we are doing with our app, GO Navigation, is we're playing in a niche of that market because we are targeting… Our solution is a premium navigation solution, and for that reason, we're targeting a niche for users that are power users on the road a lot. That's also what we see in our user base. You see that we have a large share of our user base being road warriors, what we call road warriors. These people are on the road a lot often for work.

[00:07:48.610] - Kees Oostinjen
They're looking for not a one-size-fits-all solution like Apple Maps or Google Maps, but they're looking for a fit-for-purpose solution like GO Navigation, which is built for driving. I think that's what sets us apart from the rest within this market.

[00:08:05.160] - Olivier Destrebecq
Awesome. It sounds like you guys have really found the right niche for TomTom, and I'm curious what your acquisition strategies are in those niches.

[00:08:13.490] - Kees Oostinjen
If we talk about acquisition, then for us, there are a couple of focus areas. One is obviously that we're building upon the history that we have. We have a large user base that we've built over the years from the satnav users. Many of those satnav users, they have got used to the way we visualise navigation.

[00:08:33.690] - Kees Oostinjen
When I talk about visualising navigation, that is, for instance, the way we show people when and where to take a turn on the highways, or in our app, there's a specific feature which is called Route Bar, where you very clearly see all the traffic or petrol stations or parking spots or whatever is coming up on your route.

[00:08:54.550] - Kees Oostinjen
These are very specific features that our users become used to from the past. That was, of course, one source of acquisition for us to just target those users and ask them, "Do you still want to go for the satnav?" Or maybe, "Do you want to go for the app solution that we have?" That's, I think, the first target audience that we had.

[00:09:12.120] - Kees Oostinjen
Another important source of acquisition, which really fuelled our growth over the past couple of years is our performance advertising. Our performance advertisement has been mostly on Meta, GA, and ASA, where we very strongly are steering up on cost per acquisition. What's important for us is that we target users who are willing to try out our app. As long as they start a free trial, then we know that we are in a spot where we can convince them.

[00:09:37.850] - Kees Oostinjen
If you look at our trial-to-paid conversions, and you compare them against industry benchmarks, then they're very high, they're very good. We know that that's the point where we need to get the user to. If they're there, the product will convince them, and we'll make sure that they retain. That's what we steer upon, cost per acquisition, meaning the cost of a user starting a trial. That has been very successful for us over the last couple of years and has really fuelled our growth.

[00:10:01.780] - Kees Oostinjen
What we did next to that was our awareness campaigns, where we tried to fuel basically the performance of our bottom-funnel campaigns. We've been advertising on, again, Meta, but also on streaming services. For brand awareness and also to really get a message out. TomTom is not only the satnav manufacturer any more, but we're also into apps. If you're looking for an app, we're there for you as well.

[00:10:24.940] - Pim Spaanderman
Because that was one of the main challenges when we had with marketing this app. Initially, when we started pushing the app with really performance advertising to trigger downloads, to start doing a trial, et cetera, at a certain point, you start seeing that you're hitting a ceiling, that your CPA is increasing too much and that you're topping out on your activities. That's when we saw that we need to change the ball game.

[00:10:50.310] - Pim Spaanderman
The majority of people still didn't know we have an app, and that needed to change. We started indeed doing advertising on, for instance, streaming channels to show people like, "Hey, remember this, the PND, the satnav? It now turned into this," and that's the app. Really showing that there's something else to try out. Then we pushed those advertising in parallel of our activities that we were running to do the performance advertising in the downloads and getting the trials in to see what effect was there to increase that ceiling again, so you can spend more basically on your bottom-funnel activities.

[00:11:25.860] - Olivier Destrebecq
Interesting. There's two things that struck me in your responses. One was the consumer getting used to the visualisation of the navigation and then people being aware that there was an app from TomTom. I'm wondering, how much is that navigation visualisation a barrier to entry to your potential customer, but also a barrier to exit for your potential customer. I know I hate my car navigation system because I'm so used to the one on my phone. I'm curious, is it just me or is there really a thing there?

[00:11:56.680] - Kees Oostinjen
I'm not sure if I completely understand your question. Because for us, I think the UI is definitely not a barrier in our case. I think what TomTom has been really good at from the beginning is making a very user-friendly product. It's super simple to operate, and it's super simple to understand. I think that's what you still see in our products today, which is why we have such a long retention of our users, whether it is through satnavs or through apps.

[00:12:22.560] - Olivier Destrebecq
It's not so much on the ease of use as you mentioned, when do you guys display the, "Oh, you need to turn now or turn in 300 meters", and how you represent the road that you're following on the screen and people are getting used to those. I know personally, I guess I'm fairly sensitive to it, and I'm like, "No, I really like the way the nav that I use represent it and switching is hard for me for some reason." I was curious if that's something that you've seen or not.

[00:12:45.690] - Kees Oostinjen
You mean switching from a competitor to us and the user might already be used to the competitor? I think that's why it's so important for us to get this user to start a trial with us. Because once the user starts a trial, then we know this user will stay with us. Once they've tried it, there's a very high likelihood of this user converting and retaining. That's what we also try to do with our acquisition tactics. We really try to, definitely for our bottom-funnel to really try to visualise for the user in what aspects GO Navigation is better than the competition.

[00:13:22.180] - Kees Oostinjen
A couple of those examples I mentioned, like what we call the Moving Lane Guidance or the Route Bar. But also, for instance, traffic information is one that is very highly rated within the industry. That's the message that we try to bring across.

[00:13:34.370] - Kees Oostinjen
For people that are on the road a lot and that drive a lot, they experience that difference, and they notice the difference between a fit-for-purpose app like GO Navigation, which is built for drivers or a multi-use app like Google Maps, which is really good at what it does, of course, if you need pedestrian navigation or if you need public transport information, or even maybe if you need to go for a short drive. But if you have to go for a long drive or if you drive a lot, then yeah.

[00:14:02.500] - Jeff Grang
You were mentioning a lot of the user needs to try your app once they try it, they love it, et cetera. I believe that you're doing a lot of testing in your app to be able to make people jump to that ah-ha moment. What have you been trying, reverse trial, free trial, length, duration, pricing? I'm very curious to know what were the tests that you run in your app.

[00:14:21.390] - Pim Spaanderman
To really pinpoint for us, we did a lot of research to see, "Okay, what's moving the needle during a trial?" We have a 7-day trial. At a certain point, we saw that people that did a first drive within that 7 days, that they had a higher conversion. Of course, this is logical because they've driven with the app. Of course, you would say, "Yeah, then that works."

[00:14:41.490] - Pim Spaanderman
But still, if you look at those metrics, and you know that that's true, then what do you start doing? You start to push those people that don't drive with their car during the trial. You try to move them to start behaving like that. Because on one hand, you do an advertising campaign on the phone and people download the app, et cetera, then they're not in the car yet. You need to remind them continuously and trigger them, put the right triggers out there to jump in the car, and when they're in the car, to use our app. That's what we try to do with continuous CRM campaign. We build up a lot of messaging during that 7-day trial to really stimulate people to go for a drive.

[00:15:19.980] - Pim Spaanderman
Once they did, we saw that eventually boosted the numbers of the conversions towards the paid subscription, which we wanted in the end. That's what we did the trial. But a lot of stuff happened also before that. Because the fact is that in the end, this is a paid app, and it's a big gate. If you download the app, you go through the onboarding, you hit a paywall, and you need to go into a trial, and that's where you lose the majority of the users.

[00:15:47.690] - Pim Spaanderman
There's a lot of people who are open towards downloading the app and taking that effort. But everybody knows if you have this set up in your app, you're going to lose really a lot of people in that onboarding flow, and once they hit that paywall. Knowing that, of course, throughout the last couple of years, we did a lot of thorough testing in that onboarding flow to really see what moves the needle. One of the biggest thing that impacts that is having the right pricing and having also the right trial time length.

[00:16:19.400] - Pim Spaanderman
When we started with the project 5 years ago, there was already GO Navigation app as it was. However, there was not really a commercial team on it. It was built by the engineers, and it was run by the engineers. Basically, it felt like a bit of a basement project within the organisation before we incorporated the app in the Consumer Business Unit. Then we started looking with my background in e-commerce. I started looking at all the activities we could possibly do in terms of testing, optimisation, et cetera.

[00:16:46.910] - Pim Spaanderman
We started defining where the big gaps are and the quality user research as well. For one, people thought our product was really great, and they didn't even know what they were paying. When they heard about what they were paying, they were shocked because it was so cheap.

[00:17:03.640] - Pim Spaanderman
It was a bit of a mess because the product was also not at the profitable state, but it was not well commercially managed. If your customer is saying the product is great, and they are able to, and they want to pay more for it and the product is not profitable, you start optimising price, of course. We did a lot of price testing there as well, and that had made a significant leap also in terms of the growth that we did from a revenue point of view for the app.

[00:17:30.200] - Pim Spaanderman
I think the product was costing around, I think, 11.99 per year, €12 per year, so €1 per month on a yearly subscription. We started really looking at different price for it. It's €20 per year, €30 per year, 40, 50, et cetera. In the end, we only moved to €19.99, which sounds like a small, but in the end, it's doubling. We made that choice.

[00:17:57.280] - Pim Spaanderman
It wasn't the best choice in terms of profitability because the best choice was more around, I think, around €40 at that point per year. However, we lost a lot of volume. Since this was the first thing that we were working on, we thought, "Well, we don't want to lose all that volume", because if we keep all the volume in, we make a big leap forward in terms of pricing and a big step-up, then we can still optimise if we can keep… I think we kept around 90% of the volume with only going to €20.

[00:18:26.180] - Olivier Destrebecq
That's pretty good.

[00:18:26.950] - Pim Spaanderman
We lost a massive amount of volume going to like 50, €40. Although it was more profitable, the volume was gone, and we couldn't optimise any more. We basically went for the €20, and that was the best choice. Then we continued to optimise. Also, there was a trial 30 days, and after that trial, basically nothing happened. It didn't go automatically in a paid subscription, which is also more common these days.

[00:18:51.470] - Pim Spaanderman
We shortened that to 7 days, and we put in the triggers also, the adjust triggers, et cetera, to optimise our campaign, so we can optimise the campaigns based on also on the trials that lapsed and went into paid, et cetera. We got more data in. The time period was shorter. If it was 30 days, you cannot optimise your campaigns based on those events any more.

[00:19:11.570] - Pim Spaanderman
That went back to 7 days, and we made it a binding trial, meaning that if you go into the trial, you already sign up, of course, for the paid subscription, but you can jump out of the trial, and you don't have to pay if you don't want to within those 7 days. Those things in the beginning that we did that made a big leap forward in terms of the growth of GO Navigation.

[00:19:31.730] - Jeff Grang
Do you know how much we're talking about? It's 10, 15, 20, 30% more?

[00:19:35.640] - Kees Oostinjen
Yeah. There were a couple of things that were happening around the same time. If you look at the last example that Pim gave where we changed from a 30-day non-auto-renewing trial to a 7-day auto-renewing trial, that really changes your traffic flow. What happened there is that we saw that we had a lot of less trial starts, obviously, because there was some commitment now before you went into the trial. That was costing us more or less half of our traffic that were going into a trial.

[00:20:05.590] - Kees Oostinjen
But then what you did see happening is that your trial-to-paid volume was increasing by a lot. That one increased by 30% point, actually. It went from 15 to 45%. In the end, because of that, although you lose more volume higher up in the funnel, you end up with more subscription sales. That one went up by over 30%, actually. That was for the change in dynamics of the onboarding, basically. If you look at the price increase, I think there in the end, the uplift was 23%. We're just optimising the price. All of these things had a big impact.

[00:20:39.500] - Jeff Grang
They compound.

[00:20:40.190] - Pim Spaanderman
Yeah, and then they compound in it. In isolation, it's a nice win. But if you add them all up together, then it's a huge leap forward.

[00:20:48.280] - Jeff Grang
Yeah, impressive.

[00:20:49.050] - Olivier Destrebecq
Nice. You mentioned a lot of tests that you guys have done, and some good numbers are going up. I'm curious if there were some of those tests, our test results that were more surprising to you where you say, "I bet the house on this, that is going to go to A," and it ended up going to B, essentially. Is there any test like that that's stuck in your head?

[00:21:07.890] - Pim Spaanderman
I think a lot.

[00:21:08.800] - Olivier Destrebecq
We have time. Go ahead.

[00:21:11.880] - Pim Spaanderman
The amount of tests that we do is significant. At a weekly basis, so what's-

[00:21:18.550] - Kees Oostinjen
At a weekly basis, I think it's around three tests a week that we're at. We're testing from small to big. It can be copy changes in the acquisition that we do up until trial length or price testing. It's a wide span of testing that we actually do. But it's in our culture. Everything we do, we start small, we try it out, we test, we measure, and only then we decide to scale.

[00:21:42.820] - Kees Oostinjen
Maybe on your question, what was a very surprising test or insight that you guys got that you didn't expect? For me, it was the truck plans that we introduced on iOS. Is GO Navigation offers navigation for both car drivers as well as truck drivers? Well, I think from the beginning, we already had a good understanding of the truck potential for Android. We knew from our user base on the satnavs that Android was a large platform for them and the truck navigation on Android.

[00:22:14.330] - Pim Spaanderman
Basically, all the satnav users that have a truck satnav, they were driving with Android phones. Almost nobody had an iPhone, which was really interesting to us. Okay, they're not driving with iPhones.

[00:22:26.900] - Kees Oostinjen
We introduced those plans, and it was going very well. It was going very well, I think, slightly better than we planned, actually, which was why we started debating truck plans for iOS again. Then we started investigating. The research was basically we asked our users of our iOS user base, "Guys, a truck plan, would that be beneficial for you?"

[00:22:49.960] - Kees Oostinjen
Actually, by just asking that question, we already found, "Hey, actually, we might have it wrong here. There might be users that are on iOS as well." Then we're always very close to the reviews, understanding how the users are perceiving our product. We saw more and more users asking for a truck feature on iOS in those reviews.

[00:23:08.430] - Kees Oostinjen
When we looked at all the acquisition that we were doing, then for truck, we were only acquiring users for Android, obviously. But when you do something on YouTube or wherever, it's very difficult to only target Android users and not get any iOS users in there. Also, there, by monitoring the comments that we got, we saw a lot of actually iOS users asking for this feature. That actually triggered us to look at it again, build the business case for iOS, and we made the call to also start developing it on iOS.

[00:23:38.690] - Kees Oostinjen
Actually, then it turned out even better than we thought. Now, iOS is like, if you look at our acquisition, it's almost half of Android, actually. That was one that really surprised us.

[00:23:51.610] - Pim Spaanderman
Yeah, and looking back, we could have known because basically, if we looked at the truck satnav sales in Europe, it didn't match the truck subscription sales of GO Nav. We saw that the territories were not overlapping. There was hardly any interest, for instance, in UK and in Germany, when it came to the truck GO Nav subscription. Well, we do sell really a lot of truck satnavs in those countries. Once we launched the iOS solution, it levelled out again. Basically, we see that the distribution of the countries is similar on mobile and satnav again, and it's matching the audience. We could have known.

[00:24:32.580] - Jeff Grang
That's monetisation blind spot.

[00:24:34.300] - Olivier Destrebecq
Hindsight is always 2020, so it's okay. I'm curious because we've been talking about the truck plan for TomTom. I'm curious what's different between a regular subscription and a truck subscription, just the highlights.

[00:24:49.490] - Pim Spaanderman
One thing is important. A trucker cannot drive, for instance, with Google Maps. You see those funny pictures now and then when you see a truck trying to go below something, and it hits the top. Then he's driving with Google Maps because they cannot enter the dimensions of the vehicle and therefore, they go into areas where they're not allowed to drive.

[00:25:09.070] - Olivier Destrebecq
Do you have an ad like this where side-by-side driver, Google driver and TomTom driver?

[00:25:14.440] - Pim Spaanderman
Yeah, we use them technically in our ads sometimes. But it's a bit tricky. But yes, it happens. There's a differentiating factor. For a truck driver, you need to put in dimensions, but you also have, for instance, these days, you have got the low emission zones where truckers are not allowed to drive, but it's also about weight, it's about cargo, which is not allowed in certain areas. All those things we have covered in our maps, and we can basically filter them out when a truck driver needs to plan a route. Therefore, he requires a custom navigation, basically, which we offer.

[00:25:51.690] - Jeff Grang
You shared with us some great recipes about what you've done and great successes and growth that you had. Is there something that every app maker should copy from you? What would that be?

[00:26:02.210] - Kees Oostinjen
If you look at commercially, what had the biggest impact for us over the last couple of years, then it's really studying that paywall and your price points. You really have to find that fair price point. That made a huge difference for us. But you can only do that, I think, if you're really close to the user, and you really understand the user needs, which is the other thing that we're doing. Like I just explained with the truck plan that we introduced.

[00:26:28.750] - Kees Oostinjen
In the end, we found out because we're studying all that feedback that we get in either through our customer care department or through the reviews that we see on the stores or through the comments that people leave on our ads. If you're that close, you'll find things in the end. You'll be able to connect the dots, you'll be able to understand, "Okay, this is what our users are looking for, and this is how we can improve their user experience." I think those two together, they've worked for us.

[00:26:54.180] - Pim Spaanderman
Customer research is crucial. You always need to start from there. If you're big or small, that It doesn't really matter. Customer research should always be on one, and you need to follow that. Basically, it's your leading indicator towards everything that you do. Where, of course, we have an app with a lot of users and a lot of data. However, GO Nav is also a paid app. In the world where there's a lot of freemium models, et cetera, being a paid app is on one hand also a tricky thing. Pricing is therefore 70% of the impact of what you do. If you're paid, then your pricing needs to be right, and you need to fully focus on that because everything else is optimising on a 30%, and 70% is the paid thing that you need to optimise. That's a big thing to know.

[00:27:38.830] - Jeff Grang
You were mentioning three tests per week, right?

[00:27:41.450] - Pim Spaanderman
Yeah.

[00:27:41.850] - Jeff Grang
Three tests per week, I guess you also need to have an internal organisation that is able to plan, run, analyse, and do that all over again at the same time. Could you share some insights about this culture of testing, maybe?

[00:27:54.830] - Kees Oostinjen
Some of the tests that we run is just in our culture. Take, for instance, on acquisition. What we always do, we take certain regions where we first start testing our new creatives and our new copy. We look at how it performs, we look at our benchmark, and from there we decide to either scale it or not.

[00:28:12.300] - Kees Oostinjen
The way we analyse that, that needs structure. Because if you don't apply structure, then you'll be looking at the numbers differently every time. For that, we have certain structures in place where we can basically pull the data in, and it tells us, "Okay, this is good", or "This is bad". It's never an exact science.

[00:28:29.520] - Kees Oostinjen
There's always noise in there which you have to take into account. It can be because of different regions. It can be because of seasonality, can be because of certain holidays or whatever have you, which can always influence your results, which you need to try to account for, but at some point, it stops, and you need to take a step back and say, "Okay, if we would do this, is it going to be better than what we're currently having? Yes or no?" "Is the data pointing us in that direction? Yes or no?" Then you need to go for it.

[00:28:59.400] - Kees Oostinjen
Once you then decide to go for it, it's important to have a really good understanding of your run rates. Your test can be a success, but you only know it's going to be a success once you've scaled it. In order to understand if something is a success once you've scaled it, you need to have a really good understanding of your run rate. Because if you don't know your run rate, you also don't know what you've changed that's really made an impact, yes or no.

[00:29:19.620] - Jeff Grang
Which KPIs do you analyse? Is it only one or two KPIs, or do you look at overall one KPI having an impact on the other? Another question related to that is, do you go back to the test you ran 6 months ago to be able to check if the results are still that good? Because we know that on subscriptions, there can be great numbers in the beginning, but at the end, you might be losing. After a year, the test might not turn into being a successful one.

[00:29:44.920] - Kees Oostinjen
On your first question on which KPIs do we track, we're tracking the full funnel. Depending on what we do, that defines the KPI. But if you look at the commercialisation, so many of the examples that we gave earlier, then, of course, the KPI is revenue.

[00:29:58.940] - Kees Oostinjen
But the run rate we measure over the full funnel, and we really try to understand, okay, once we change this in our acquisition strategy or that in our onboarding flow or this in our price points, what happens? Are we going to see a drop in volume that comes in? Or are we going to see a drop in trial starts, or are we going to see a drop in conversion to paid? Or where is the difference happening, and how can we actually make sure that that difference is how we want it to be? That's on the KPIs.

[00:30:28.080] - Kees Oostinjen
On the testing, and then if we look back, definitely, I think that's also very important. Keeping a backlog or documentation of all the tests that you've been running is essential. Definitely when you're running a lot of tests, because what you don't want to is that you're running tests that you've already done some time ago.

[00:30:45.980] - Pim Spaanderman
Or when you need to, then you need… Sometimes you need to run them again.

[00:30:54.860] - Kees Oostinjen
Sometimes you need to run them again, but then there's a reason for it.

[00:30:55.230] - Jeff Grang
We had some people in the show telling us they ran their test a year after and the result turning too different because people change and habits are changing also.

[00:31:03.620] - Kees Oostinjen
No, I think that's also fair. Your traffic can change, right? Once that changes over the years, then that also gives a new opportunity to run a test that you run in the past to run it again. But then you have a reason for it. As long as you understand why you're running something again, then it makes definitely sense.

[00:31:22.500] - Jeff Grang
Some sanity checks.

[00:31:24.400] - Kees Oostinjen
The whole idea is that you build one test upon the other. You learn from one and based upon that one, you can say, "Okay, guys, we see when we change the paywall, what we see happening is that we sell more family plans, but we also see that we're selling more monthly plans," for instance. What do we like about it? What don't we like about it? Well, we like it that we're selling more family plans because that's actually a more expensive plan for us where we see a higher retention. But the monthlies is not our preferred plan to sell.

[00:31:54.630] - Kees Oostinjen
What can we do in order to counter that? Then we say, "Okay, let's take the same paywall again, but let's move the monthly plan to the background." In the end, then you see that with the first test, it wasn't making a real difference yet on the bottom line. But then with the second test, you see, "Hey, now something is starting to change, and it's starting to move in the right direction." I think that's why it is important to keep documentation, to always draw your conclusions from the tests that you've done, to always write out the next steps that you're going to do, and to make sure that you build one test upon the other.

[00:32:28.130] - Jeff Grang
Excellent.

[00:32:29.780] - Olivier Destrebecq
Thanks for that turn-by-turn navigation success on testing, that's awesome. You gave us a lot of insights, a lot of great answers. If people want to learn more about what you guys are doing individually in TomTom, where can they go?

[00:32:41.590] - Pim Spaanderman
I think they can just reach out to us.

[00:32:42.820] - Olivier Destrebecq
All right. What's the best place to do that?

[00:32:46.140] - Kees Oostinjen
You can find me on LinkedIn.

[00:32:47.840] - Olivier Destrebecq
Awesome. We'll put the link in the show notes. Awesome. Well, thank you very much. It was really great to have you, Pim and Kees on the show. We really appreciate you taking the time to talk with us today.

[00:32:58.090] - Kees Oostinjen
Thank you, guys. Thank you so much.

[00:32:59.150] - Pim Spaanderman
Thank you.

[00:32:59.530] - Kees Oostinjen
Nice to be here.