Telecommunications Predictions 2020: Ready for the AI Fiasco?

From financial inclusion to AI failures, here are four things I see happening

“Siri, do you have any thoughts on stuff that might happen in the telecommunications industry in 2020?”

“What, you want me to think for you again?”

“Yes, please”

“Think for yourself, I’m busy answering existential questions”

“Fine, whatevs”

Well, seeing as Siri wasn’t in the mood to help write this post for me, I figured I’d start this piece with a prediction that Apple might have already beaten me to...

A brand-damaging AI project will make headlines

Artificial intelligence has certainly reached top of the hype curve during the last couple of years. However, the deployment and use of AI could well have unintended consequences. As outlined recently by Scott Zoldi and Kate Crawford at FICO World 2019, algorithmic bias can creep into decisions made by AI.

For example, if AI is being used in the hiring process, the algorithm might look at previous employees that have been successful at the company. If the company isn’t very diverse to begin with, this can lead the machine to conclude that it should only look at hiring people that are similar to who has been hired previously. This perpetuates the lack of diversity.

This brings me back to Apple, which launched their credit card in the US earlier this year and began issuing smaller lines of credit to women than men. There was a lack of clarity about how the algorithm worked and how to explain the output.

As we move into the era of smart cities, such as Las Vegas, looking to manage traffic more effectively, one is reminded of apps like Waze which have been accused of creating traffic congestion due to drivers using alternative routes and short cuts that then amplify existing issues or create new ones.

Given the current lack of control and governance around AI, and that even Apple can mess up, it is conceivable that a large CSP will deliver a project that drives unexpected results and adversely impacts customers as an outcome of poor data science rigour.

Financial inclusion initiatives will help accelerate monetisation of GSM data

Telcos have been talking up the potential of monetising their data assets for years, however they have had limited success, in large part due to the data privacy implications of selling personally identifiable information for commercial use.

So, why don’t operators look at it from a different point of view and offer their customers something in return for their data being used? Mobile for Development remains one of the pillars of the GSMA strategy, whether that be for financial or digital inclusion or to reduce the gender gap of mobile ownership and use within developing markets. The ability to offer financial products where there is little to no credit bureau, yet almost ubiquitous phone ownership, offers opportunity for both CSPs, financial services institutions and consumers. (Watch this space for news of a project we’re doing.)

CSPs will start to centralise group-level decision platforms

As the practical use of advanced analytics such as AI becomes more widespread, so will the consideration and adoption of platforms that can execute the outputs of the analytics into the business. These decision management platforms could apply decisions across the customer lifecycle and apply to both OSS and BSS sides of the house. Examples include NFV / SDN, customer onboarding, loyalty and retention and customer care.

Several large CSPs have a number of companies across the group, with many local installations of such platforms. Groups will begin to evaluate how cloud-based scalable decision platforms will enable consistency across the organisation, reduce the overall cost and complexity of vendor management, plus allow centralisation to become a reality. Most CSPs have stated strategies for digital transformation and setting themselves up for future growth. A standardised approach to decision management will facilitate delivery on such strategies.

CSPs will start getting better at the fluffy stuff

How well does your telco know you? The answer should be, “They know me really well, they have a hugely rich set of data about where I go, when I go there, who I interact with, how frequently, what I like to watch….” And so on. Whilst that might seem a little Orwellian to some, if consumers are ok with the likes of Netflix and Amazon serving up relevant suggestions based on what they know about the user, then it is reasonable to expect telcos to at least try to match this experience.

But how well does your operator interact with you? How often do they provide you with personalised communications, that are relevant to you in that point of time, with context that demonstrates they understand your needs? Not that often, right?

In my experience it’s fair to say that not many telcos are brilliant at managing their massive customers bases that effectively. Recently, I received my ‘Spotify Unwrapped’ communication which tells me how I’ve used their service over the year (and compares to previous years). It’s essentially a nice way of engaging with their customer base, encouraging positive sentiment on social platforms and helping users to discover more within the product through relevant suggestions. Would something similar from my CSP give me some insight into how I use their services? Might they be able to persuade me to upgrade or cross-sell me some value-add services based on what they know about me and ‘personalise’ offers to me via a decision platform fuelled by prescriptive analytics?

Do you see something else in the year ahead? Post a comment below.

chevron_leftBlog Home

Related posts

Take the next step

Connect with FICO for answers to all your product and solution questions. Interested in becoming a business partner? Contact us to learn more. We look forward to hearing from you.