Roman Postnikov, CEO of artificial intelligence (AI) solutions company oneFactor spoke to Mobile World Live(MWL) about how mobile network operators can best use machine learning (ML) and big data to grow and diversify their sources of revenue.
MWL: What are the biggest challenges operators face globally?
Operators are facing the challenge of next-generation network transformation for 5G deployment. The transformation will not guarantee the kind of revenue growth that was generated from the 3G-to-4G transition. Indeed, the biggest question today for telcos is how to scale their business. Operators around the world are being forced to seek additional revenue sources.
What are some of the ways they can do this?
An effective way to grow is to make use of big data through AI and ML solutions, which can open up commercially-proven opportunities in these promising areas.
Big data is an integral part of a mobile operator’s ecosystem. They understand the great benefits of external monetisation, delivering value-added services to consumers. We see a number of cases where operators start to leave for neighbouring markets driven by big data — banking, retail, advertising, internet services, et cetera.
We have seen operators make huge deals with companies and gain access to massive amounts of data. For example, a few years ago, Verizon acquired Yahoo and America Online, one of the world’s largest internet service providers.
Another example is when Russian operator MegaFon bought a stake in Mail.ru Group, which owns the country’s largest social networks VK.com (more than 500 million accounts) and OK.ru (over 200 million registered users and 45 million daily unique visitors).
However, not everyone succeeds. In fact, most of them have rather poor results. Not many companies can perform a quality geolocation footfall analysis based on pedestrian traffic data. It’s not enough to own data, the technology is the key factor.
What operators can do instead is monetise their own big data and develop innovative products for other industries. They can turn their own data from a source of expense (for storage and data processing) into an asset that delivers additional earnings.
However, they must keep in mind generic AI solutions are not as effective as specialised platforms.
Based on our experience, ready-to-use specialised AI platforms for mobile operators generate up to 4 per cent on top of revenue, through pre-installed sets of B2B value-added services for key industries.
How can AI solutions help operators?
AI services and ML algorithms based on mobile network operator data can be trained to provide solutions around retail network optimisation, recommendations for optimal store locations, credit portfolio management, executing target marketing and communication campaigns for clients.
According to McKinsey research in 2017, up to 49 per cent of all operations will be automated by 2028. Labour automation in key industries can lead to the boost of productivity and growth of GDP in countries, which will ultimately drive their local economies. Mobile operators can play a big role in this, based on the use of data from various industries.
If operators sell or just transfer their own data, they will fail to get added value, staying out of the market like a big empty bucket. To become a key influential player, mobile network operators should turn into a digital ecosystem for the country.
What are the challenges around this?
The development of AI solutions requires investments, R&D, time and labour resources.
Also, any external use of data must of course comply with GDPR and other relevant regulations, and they must consider the impact on customer privacy.
When building data monetisation products, necessary safeguards should be fundamental to the design. In almost every case, operators should sell the products of their analysis rather than the raw data. Operators must meet a real market need with any data monetisation product.
There must be a willing market ready to spend money on the product being proposed. Operators should have a clear idea of how to connect with that market and an understanding of how the market will use the data to design the product appropriately.
Transforming telecom data into a valuable product that can be adopted in different industries is a very difficult task. It’s not the same as applying AI and ML capabilities to solve internal goals.
Mobile operators should remember two simple things. One, it’s impossible for them to solve internal and external big data tasks on the same level of effectiveness. And two, generic AI platforms do not solve specific business tasks.
What work does oneFactor do with operators?
Since 2015, we have on-boarded and generated revenue for several mobile operators, international payment systems and credit history bureaus.
oneFactor works with leading operators to help them monetise their own data, generate new non-telecom revenue streams for banking and online lending, insurance, retail, e-commerce, advertising, et cetera. We make AI available for every business unit, delivering efficiency boosts and increasing organisation revenue
Our solutions help operators to transform their businesses, enabling data collection from different sources and uncovering new business opportunities. For example, using our automated ML platform any mobile operator can manage not only their own subscribers database, but all mobile phone users in the country, earning up to $4 billion annually. The platform is a ready-to-use solution to solve almost any particular task for any industry without special configurations.
At MWC19 Barcelona we have introduced our privacy-respecting Data Brokerage platform, which allows companies to process and combine data from transactional business resources in order to maximise the accuracy of the final AI and ML algorithms. At the same time, the platform protects the owners’ sensitive and confidential information, as it does not require taking the data outside the protected perimeter of the mobile operator.
