100+ million profiled users in 2 years. 2.1 billion location footprints in 4000 cities. The AdNear story
You are heading into a supermarket to buy monthly groceries. After you park the car, you check your phone and see that your chat app shows a message from a friend. You open it to reply, and find that next to the chat, there’s an ad displaying discount deals at another supermarket nearby. You make a mental note to try that grocer next time. Bingo! AdNear has bagged another potential customer for their client, the rival supermarket.
This is just one of the countless ways AdNear, a big data company in the mobile ad tech space, leverages geo-location and other data points to create intelligent targeting capabilities for ads across mobile devices. Today, AdNear has 2.1 billion location footprints across 4000 cities; they serve targeted ads from 40,000 apps to over 100 million users. Currently headquartered in Singapore, and backed by Sequoia Capital and Canaan Partners, they process billions of ads and terabytes of data every month.
Incredible for a 16-month-old startup that grew out of Bangalore? Read on to find out how they did it.
Mobile user profiling
Anil Mathews, Founder and CEO of AdNear – and serial entrepreneur since 1999 – had three successful exits before starting up AdNear in November 2012.
When we met him at his Bangalore office – to understand how this company had taken off into the stratosphere – Anil could barely contain his excitement about the 100 million plus profiled users in the AdNear kitty. “You see, they aren’t generic users. Each one of them is a user to whom we have assigned a unique ID, about whom we have basic information like gender to deeper details like whether she or he is a frequent traveller, a student or a housewife,” he tells us. That’s a gold mine of data for them to tap for their value proposition to premium brands – Audi, Pizza Hut, Google, Titan, P&G, Airtel, Microsoft, MAC, Toblerone, Samsung … their client list sounds like a Global 500. What they offer to these companies is a more effective, targeted ad campaign, instead of the traditional blind, blanket ad which gets a small fraction of the response that AdNear can generate.
To understand how AdNear builds the intelligent user profiles that hold the keys to their success – without actually collecting personally identifiable information like names or even mobile numbers – you have to go back to the beginning to hear their story of serial pivots.
A bend in the river
The year was 2009. Anil and his engineering team were working along with Rediff.com on how to bring location awareness to mobile phones without relying only on GPS. The plan was to license this location data out to companies who could use it for their own services like search, advertising, and so on.
GPS penetration was low – it still is, but was worse then – and many users turn off the battery-draining GPS on their smartphones even if they have it. Mobile operators weren’t ready to provide user location either. “The only option was to explore if we could build a ubiquitous platform to work across phones — smartphones, dumb phones or feature phones — and across operators without depending on GPS,” Anil recalls. For this, they had to get thousands of people to drive across the country with mapping devices to log mobile towers, wi-fi zones and other geographical data. It was tedious, expensive and time-consuming. But eventually, it became one of their chief assets.
Soon enough, they realised that the licensing model doesn’t work. In India, companies wanted the data but didn’t value it enough to pay well for it. This forced Anil to pivot and adopt a new business model. Ajit Balakrishnan, Founder and CEO of Rediff.com, too played his part in this. One day he called Anil and asked: “Do you think you can make this location platform a $100 million business?” Anil replied no. “Then you should come and work for us,” Ajit said. That is when Anil decided he had to change the business model.
Where’s location data used the most, he asked himself. Primarily in two areas: search and advertising. Search was owned by the biggies, so there was too big an entry barrier. Interestingly, at that point of time, nobody was into targeted mobile advertising using location, except for Navteq, owned by Nokia. And Navteq was heavily dependent on GPS. So when Anil approached them with his terrestrial data for a possible partnership, they were thrilled. Thus AdNear was born in November 2012.
Working with Navteq gave Anil and his team a deeper insight into mobile advertising, and the programmatic ways to leverage their location data on it. So when Navteq became Nokia Location & Commerce, and later wound up, the AdNear team could effortlessly move on. “And we became the largest location aware mobile ad tech company in the Asia-Pacific geography, including Australia,” Anil says. They now have a presence in five countries: India, Australia, Singapore, Malaysia, and Indonesia. And they also serve other South East Asian countries like the Philippines and Thailand.
Of geo-fences and location footprints
It is AdNear’s use of real world data on top of virtual data (which is what traditional ad companies use to sketch a user’s profile) that differentiates them from others in the mobile ad tech space.
For example, if you are seen using chat apps frequently, you are likely to fall in the 18-24 age bracket. But then, most of us older people too use chat apps these days. So, to improve the predictability and profiling, AdNear adds real world data on top of the virtual data from your online behaivour – for example, where you have been spending most of your time over the last 30/60/90 days can tell a lot about you. If they saw one of their unique user IDs tracked within a university campus from 9am-5pm every working day, then the probability of the device belonging to a student is very high. (To reiterate, AdNear does not identify the student or mobile number, only tracks a device to which a unique ID is assigned the first time it taps an AdNear ad.)
AdNear has billions of location reference points mapped, which they call their ‘geographical layer’. That data is their strength.
If your phone is found to be in one of the posh localities in the city from 9 pm till the morning, night after night, they use it as a data point to determine you are an affluent. Or, if you are found in a residential area during the day as well, and if you regularly travel to a school and back, it could mean you are a homemaker and a mother. These are just some of various parameters they use to profile you.
Before you think it’s an invasion of privacy, Anil points out regulatory bodies are strict in the mobile space, and AdNear could not collect personally identifiable information even if it wanted to. “We also do not share any data with third parties, other than the analytics info we pass on to our clients,” he adds.
Staying a step ahead of rapid changes in mobile world
The mobile ad industry has been evolving rapidly in the last 12-18 months. With global mobile ad revenue expected to touch $11.4 billion by the end of 2013 (up from $9.6 billion in 2012) and grow five-fold between 2011 and 2016, this space has been a hotbed of innovation.
The nature of the beast has forced AdNear to keep pivoting in sync with the changes in the industry. Some of the companies in this space perished because they didn’t do that, Anil says.
For example, AdNear initially bet on India, where they expected a hockey stick growth trajectory. But when they saw that the market here wasn’t willing to pay for premium features, they spread operations to Singapore and Australia, where brands were faster to appreciate the value of what they were offering. “One of the important things for us as a company has been not to stick on to our original plans.”
They began as a location platform company, became an ad tech company and today, they look upon themselves more as a big data company. “What we do now is to make sense of big data, and it just so happens that we are currently leveraging it in the mobile advertising space.” What they’ve become expert at is to collect vast amounts of granular geographical data from multiple sources, as well as huge masses of mobile user behavioural data, and connecting the dots with intelligent programming.
The next bend in the river is hard to predict.
What do you think are the new, new possibilities? Let us and AdNear know in the comments section below. And here’s a tip: they’re looking for a data wiz.