Teragence Mobile Network Crowdsourcing and Geo-Analytics

Christian Rouffaert, had a talk with Geomob’s Ed Freyfogle discussing how Teragence has carved a niche in providing granular network mobile coverage insights through the use of mobile network crowdsourcing and Geo-Analytics. Christian shared his journey from corporate warrior to entrepreneur, explaining how Teragence uses geo-analytics to meet the needs of various customers.

This podcast allowed us to showcase our mobile coverage insight solutions, but we also explain why we have chosen crowdsourcing and geo-analytics as our methods. This podcast offers a behind-the-scenes look at our strategic decisions and the real-world impact of our work. Join us to gain a clearer understanding of the value and innovation Teragence brings to the table.

We share below the original podcast and the transcript, grouped by thematic sections for ease of navigation and to enhance your understanding of the key points discussed:

• Introducing Geo-analytics and Teragence’s Unique Approach
• “Mobile Network Coverage Maps on Steroids”
• Teragence’s Telecom Data Analytics Expansion and Strategic Shifts
• From Telecoms to Geo-Analytics: The Entrepreneurial Journey of Teragence’s Founder
• Embracing the Startup Life After Corporate
• Mobile Network Data Analytics: Teragence’s Customer-Centric Approach

Happy listening!

Introducing Geo-Data Network Analytics Teragence’s Unique Approach

Welcome to the Geo Mob podcast where we discuss Geo innovation in any and all forms, be it for fun or profit. Welcome back, everyone. Today my guest is Christian Rouffaert. Christian spoke at Geo Mob, London all the way back in September 2016. He is the founder of a London-based company called Teragence that does mobile network analytics and I wanted to have Christian on the show.

There are a couple of reasons I thought this might be very interesting for the Geomob audience. So first of all, as a as a startup founder, in some ways, Christian doesn’t exactly fit the mold of your typical startup founder, so I think there’s some things we can explore there. But also what it is that Teragence does. Because while Christian is a Geomob regular and many of you may have seen him about Geovation, they don’t present themselves at all as a geo company even though they use a lot of geo stuff behind the scenes.

So I think that could be very interesting . So without further ado, Christian, welcome to the podcast. Give us a little introduction and tell us what exactly it is that you do and what’s geo about it. 

Christian Rouffaert
Thanks, Ed. Well, my name is Christian Rouffaert. I am the CEO and founder of Teragence. And we describe Teragence as a network crowdsourcing and geo-analytics business, which is quite a mouthful. I’m still working to get that elevator pitch a bit more crisp, but that’s what it is.

Basically we work with location enabled apps to collect location-specific data about the state of mobile networks worldwide. We do that collection in a data privacy respecting way. So we all know all about GDPR and walking on the right side of the line of GDPR. In short every time that a location enabled app connects to the mobile network, it collects some information about the state of that mobile network.

We collect and aggregate that data at scale. We clean it, we standardize it, and then we crunch it to create mobile data insight products and we sell those products to everybody who essentially depends on mobile networks to make their business work. The Geo angle in our business comes from the fact that a lot of crowdsourcing companies will issue reports saying the best network for 4G in the UK is XYZ.

We believe that’s a much too coarse way of looking at things. We believe that you should say “the best network for mobile connectivity on the corner of Street A&B is network X”, so very location specific and that location specificity got us into the whole geo angle of things. 

Ed Freyfogle
OK, fantastic, Christian. Lots of pieces to explore, so let’s get right into the first thing that springs to mind: anytime I hear of someone doing crowdsouring of data in a privacy respecting way under GDPR. But how exactly do you do that? So how does this data collection work and can you go into any details there? 

Christian Rouffaert
We work with location enabled apps to collect that location specific network information. The one thing that we are very, very careful about is that we do not collect any permanent ID associated with those measurements.

Basically,  if you have a permanent ID associated with a number of location measurements, you do not need a lot of location measurements to work your way back into who that person might be. Whether that ID is “anonymized”, or is a lot of cryptic symbols doesn’t really matter. We are very careful not to collect any permanent IDs associated with our data. You cannot find anybody in our data and we are very careful that that’s never the case. And that’s a great comfort for our partners and we make sure we always stay on the right side of GDPR. 

Mobile Network Coverage Map and Insights On Steroids

Ed Freyfogle
Very nice.  At this point now you’ve got the data and then presumably you do some kind of processing on it to create these products. But what are those products and who are the customers, what are the use cases?

Christian Rouffaert
Yes, we often get that question: “It’s quite cool, but what do you do with it?”. It actually turns out one can do quite a bit. The first product  we have is what we call our Mobile Signal Checker.

It’s essentially a mobile network coverage map but on steroids . We think that the operator provided mobile coverage maps are essentially marketing tools, with very little operational relevance. They allow you to ascertain mobile coverage in aggregate, but not on the corner of Street A and Street B. And we provide exactly the detail that says “On the corner of Street A and B, what is the expected signal strength for operator X?”

Basically that is a service that’s directed at everybody that wants to deploy a wireless device in a specific location and wants to make sure it’s going to work.

And generally speaking, networks work 80 to 90% of the time, but in 10 to 20% of the time, for example in Internet of Things businesses, it doesn’t work. And when it doesn’t work, it causes a disproportionate amount of grief. We basically help people estimate  how much grief they’re going to get.

Another use case, is with parcel delivery companies, courier companies. When they have to deliver to rural locations, they have two problems:

#1 the address, and its match with the location is a bit vague, so drivers end up running around finding where is “The Big Tree Cottage” exactly, but more importantly in those areas there’s often bad mobile coverage. As a delivery driver you’re then running around trying to find this specific cottage and you cannot make a call and you’re wasting a lot of time.

And so again, our service slots into that use case. It gives people the heads up that there might be an issue which enables you to shave off a few percentage points in your delivery efficiency, which actually translates into quite a few dollars for those companies. So that’s one use case.

Another product is our Cell Coverage product, which basically enables us to describe in great detail. footprint of every individual cell tower. We say if you connect to this cell tower, we know that that cell tower covers this block of streets and that has a use case in in coarse location assessments.

So every time that you want to find the location of an object, but you cannot use GPS,  we can use cell based location assessment which is not as precise, but it’s good enough and in many cases good enough is just that- “good enough”. We sell that to a variety of location based players, as well as what are called MVNO’s- mobile virtual network operators.

And then the third aspect is basically any matter of custom analytics, AI, machine learning, whatever you want to call it that we can do on our data to find specific insights.  Let’s just say bespoke questions and bespoke insights for our customers.

For example, we’re doing work with a specialized real estate player to map out indoor signal quality across a number of buildings in major cities. So that’s what we do. That’s who we serve. 

Teragence’s Telecom Data Analytics Expansion and Strategic Shifts

Ed Freyfogle 
Are you doing this primarily in the UK or are you doing it across multiple markets? Or where?

Christian Rouffaert 
We’re doing it mainly now across Europe and North America, but that’s just a consequence of our customer base. But there is no reason that we couldn’t do it for the rest of the world.

Ed Freyfogle 
A couple of things immediately jump out at me. First of all, that must be an absolute immense amount of data that you’re. Shuffling around.

Christian Rouffaert 
It is indeed yes.. We are quite mature about that.

Ed Freyfogle 
But your target market is basically the mobile operators, right? There aren’t hundreds of mobile operators. So how do you kind of grapple with that ? 

Also, from the broader point, speaking from my own experience with start-ups, trying as a tiny startup  to work with these big companies  like telecoms or large organisations in general: they just move at such a different schedule and such a different pace. It’s an absolute nightmare. And then even when, even if people want to work with you, the amount of time it takes to get into the accounting department and all this kind of thing and actually get paid. So how do you grapple with that? I’m sure that’ll be of interest for others.

Christian Rouffaert 
Yes, that’s one of the interesting learnings for us as a company. As you said in your introduction, I’m not your classical fresh out of college entrepreneur. I actually had a significant corporate career before I started this business. When we started , we thought “Oh my God, this  data is going to be so cool.

We can do so many interesting things and the obvious market is big. The mastodont of a market opportunity is the mobile operators. So we’re going to target the mobile operators.” That did not turn out so well – for a number of reasons, one is that with corporates and especially mobile operators are a very special breed of organisation. You can have many, many good meetings. And you can meet people that are very interested in your proposition. But what you learn very quickly is that in corporations good meetings lead to other good meetings, which lead to other good meetings, because that’s the way a corporation works. Many people need to be involved. Many people need to see it.

There is essentially a common understanding that needs to be built, a consensus , that buying a certain service, especially if it is from a small unknown brand is a good idea. That process takes an enormous amount of time. When it is then decided that that is a good idea, there is an entire process of admin buyer registration, accounting, all kinds of forms that you have to fill in. Then typically the buying format is some kind of competitive tender, which means that you fill in even more forms and you get squeezed on the price. And we found that process really hard.

Hard and draining, both psychologically and on the physical resources. And we were not very successful. In the end we pivoted away and took a very conscious decision not to sell to the big corporates. We decided to  target smaller companies. Smaller companies in our world means turnover anywhere between, let’s say  £1 million and a billion pounds – it can be still quite large.

Creating a targeted solution for those players in those markets enables us to collapse the decision timeline and get to sales very quickly. We’re not hunting these big mammoths. We’re hunting, what I call, squirrel and deer. But you know you can bring them in at a regular pace. Also more interestingly, we discovered that there’s actually a whole universe outside of big telco. We call it “small telco and associated services”. One does not think of them at first, but this is a very significant segment.

For example, in every country you have three-four major mobile operators, but there is also a multitude of  what are called mobile virtual network operators. They each are businesses of several 10s of millions of pounds or dollars in their own right. They are looking for creative solutions. So that’s a very interesting market for us. Similarly there are operators are specialized just in indoor cell coverage.

There are many more than the traditional mobile operators, again smaller, making faster decision. We found that tier-2 or adjacent markets actually to be quite large and much easier to sell to . And that’s what we focused on. That that helped us to become successful. 

From Telecoms to Geo-Analytics: The Entrepreneurial Journey of Teragence’s Founder

Ed Freyfogle
That’s interesting, because actually at my own company, OpenCage, we went through a similar process. We got contacted by one of the big operators in the world. It seemed like they liked our product and were ready to purchase. After six months of going back and forth doing calls, we  basically had to stop the process because we could not spend so much time on this. At that point we realised they don’t even make the purchase themselves.

They used an outsourced purchasing company. And I was trying to deal with these people who had no idea what they were purchasing or why they were purchasing it. And it’s at some point I just got so fed up and just said “I can’t do it”. It’s interesting that you’ve discovered this kind of niche of what you call the small mobile. And  that’s an interesting terrain to hunt in.

Your own background is coming from telecoms. Tell us a bit about that: how you made the jump to entrepreneurship , and how that process was.

Christian Rouffaert
Yes, that’s an interesting story.  I  worked in telecoms business in various guises for the first 15, maybe even 20 years in my of my career . And then I worked as a management consultant for Accenture Consulting, which I really enjoyed.

But one of my bosses once said to me: “If the company was a brass band and was striking up a tune, you fall in line with the tune, but you are the guy playing the piano and find jazz improvisations on the main theme” .  When I hit my mid 40s, I kind of came to a point where I had to decide whether I would really join the brass band or start my own jazz band – proverbially speaking. And I often joke and say, well, I had my midlife crisis and instead of buying myself an open top Ferrari and reliving my youth, I decided that it was time to make the jump and get my own start-up.

Ed Freyfogle
And to clarify, did you have any geo background at all when at any of these previous roles or purely more on the telco side?

Christian Rouffaert
I had zero geo background. We came to this idea of crowdsourcing and analytics and software defined networks which are very big themes in the industry. Our thinking was “We know the world is moving to data. Telecoms is moving to data, so let’s do analytics on these crowdsourced data and create some valuable products around that”.

Geo did not really come into it. But as we started to peel back the layers of the problem statement or the market opportunity, we realized that geography and location mattered . As I said, we wanted to give the indication of what is the signal on this street corner versus that street corner. That gets you into all kinds of location things and we realized that with the vast amounts of data that we had, that not only geo is an important dimension, it’s also a dimension that you cannot be treated as any other information point in your database. It has peculiarities that make it sometimes very difficult to aggregate and to process. So by hook and by crook, almost forced by circumstance, we had to dig into what is probably described as the world of GIS.

We brought geo in as a dimension, not because we were geo obsessed, but because we were business obsessed and that was an important dimension. We started to use tools like QGIS and Esri and to understand all the geo libraries and tools available in Python . We’re quite proficient in those, but it’s always from an angle of the business first and the geo dimension later. For example, when we address the market, we don’t talk about GIS .

We use the term “geo analytics”, which is basically analytics in a geographical context. It makes it less scary and more business-focused.  It’s something that we see all the time. We also use AI style algorithms, but we don’t say we are an AI company. Because that betrays a fascination with the technology, rather than the solution. We work around a lot of geo, we have to understand a lot of geo use cases, but we’re not geo-obsessed people.

Embracing the Startup Life After Corporate

Ed Freyfogle
Well, first of all, you must be the only startup in in the world that’s not claiming AI. But congratulations on this transition. That’s great and frankly, for me personally, of course I welcome all the slightly nerdy GIS people at Geomob as well, but I love it when people come kind of from other industries and bring that knowledge and are applying Geo to it. I think that’s really where the interesting things can really happen. So tell us some more about this. As you as you struck out on your own awith your improvisational jazz. So what what has been kind of the good, the bad and the ugly of it, the transition? From a more traditional corporate world, cannot have been easy.

Christian Rouffaert
I came from corporate life. I had an MBA under my belt and I thought start up life or entrepreneurial life will be like corporate life, but just slightly different. And I quickly learned that it is not the case. It’s a totally different beast at so many different levels. I think the first one ,and I think is one of the most fundamental, at least for me, is on a personal level. No matter how much you profess to dislike corporate life, it does have a significant dimension of personal validation.

You have a job, you have a job title, you have a boss that wants to hear from you from time to time, you have underlings that you have to direct and that want to hear from you as well. In a start up business all you have is some good ideas and the phone is not necessarily going to ring very often. And when you ring up other people, they’re not going to say ”Oh, yeah. I, you know, you are backed up by this great brand. I I’d love to hear what you have to say”.

A lot of that kind of external scaffolding of your social persona falls away. And you depend on that to a lesser or a bigger extent. In startup life you have strip that back and build it up again and you have to hustle to do that. You have to drum up business and to find customers. You have to tell your story and to convince them that it will work, without the brand of your employer behind you. Instead, you have to build your personal brand. That was quite an interesting experience.

Ed Freyfogle
I think many, many people underestimate that. Dealing with that ambiguity and not having an easy answer when someone asks “So what do you do?”. I think a lot of people struggle with that. It’s harder than people might realize.

Christian Rouffaert
Yes, exactly. A friend of mine says all transformation is painful, which sounds horrifically cliche, but it is and it is not to be underestimated. You know when you have to go and say, “I run this business. I am the brave independent wayfairer”. But that’s not easy.

The second one is the overall support structure. I in a corporate world, people do things for you. Accounting takes care of the costs, the billing takes care of the billing. In consulting land somebody will make your PowerPoint for you. Those things all happen automatically.

They don’t cost anything. One of my big initial was that I would declare that something needed to be  built. And then you go well, where is the person that does that? And that person does not exist.

Ed Freyfogle
Yeah, I do that too. I just announced the plans and I just clapped and hopefully it appears. Somehow it doesn’t always work.

Christian Rouffaert
Yeah, it goes to all things big and small. And for all things big and small, you then have to make a decision:  I could probably buy or pay somebody to solve that problem. That does not mean that you have to hire an employee. You can hire somebody through platforms like Upwork, but that that cost you money. And as a small startup, you don’t necessarily have loads of money.

So you have to decide: do I spend the money, or will I spend my own time, trying to get my head around this, to learn a bit of coding,to create my own design. But that’s a lot of personal drain or effort to create something that’s actually quite mundane. You have to make that continuous decision between  DIY  or paying somebody which actually costs real money as opposed to corporate money. In the beginning I had to teach myself GIS tools and the Python coding, and I would run around in the house really grumpy because it wasn’t progressing.

But the flip side of it is when you then master these domains or you can create something that then has value or is perceived to have value in the outside world – there’s no greater kick than that. You came up with the idea. You made the idea happen and somebody’s paying money for that. That’s the best kick in in life.

Mobile Network Data Analytics: Teragence’s Customer-Centric Approach

Ed Freyfogle
Well, congratulations. That’s an impressive story. And my hope is that there’s someone out there listening right now with their mouse hovering over the purchase button on the convertible Ferrari. And then they pull away and say, you know what, maybe a start up might be a  better way to  go. I think it’s really impressive, Christian, you should be pleased.  And I guess as part of this journey, of course, you eventually along the way discovered Geomob at some point and and you’ve been a regular attendee over the last couple of years. So any one, one standard question that we like to have for people who have been that have any favorite memories or talks that inspired you?

Christian Rouffaert
Well, thank you. I don’t remember the person who made the talk, but I remember where somebody presented different ways of visualizing data in a Geo context and he illustrated that visualisation is not neutral. The way one presents data in a geographical context always reflects one’s own predilections, or political or moral opinions. One can play around with them, and one can bring out certain conclusions or de-emphasise others. Very simple tricks, the same information presented slightly differently, can make a big impact. That always stuck with me.

Ed Freyfogle
I think several people have made that point, but actually I don’t know if you listen to the most recent episode that’s come out with the new book by Ken Field. He takes the US election data from 2016 and makes 50 different maps. The same data set, he presents in 50 different ways. And he shows how you can, depending on which message you want to emphasize or how you want to kind of manipulate the conversation, you can do that with the data. That is a very powerful example. I think it is something everyone needs to be aware of as they’re as they’re presenting data because it is easy to kind of force the narrative in one way or another.

Christian Rouffaert
I I think people have have a lot of faith in data analytics analytics because it sounds scientific. But there is quite a bit of interpretation and projection and direction in there. It’s good to be aware of that.

Ed Freyfogle
Agree. So Christian, thank you for coming on the show and telling your story. I think it is very inspirational, hopefully for others. What is the best way, if this sparks someone’s interest and maybe they want to become a customer of Teragence or want to learn more about it. How can they best get in touch with?

Christian Rouffaert
Well, do check us out on our website Tergance.com or drop me an e-mail. We’re not really on Twitter, Facebook or Instagram. We are on LinkedIn. So reach out to me via LinkedIn and I will definitely respond.

Ed Freyfogle
OK, we’ll make sure we get all that into the show notes. So thanks again, appreciate you coming on the show. And telling your story.

Christian Rouffaert
OK. Thanks very much, Ed.

Ed Freyfogle
Thanks everyone for joining us today and listening to the Geomob podcast. Hopefully you’ve enjoyed the discussion. Please don’t hesitate if you have any feedback for us or any suggestions for topics that we should cover in the future. You can get the show notes over on the website, which is at geomob.com.[…] Bye.