Make sense of your marketing signals with artificial intelligence
It doesn’t feel like too long since IT and telecoms were very separate entities, writes Carl-Erik Kjaersgaard, the chief executive and co-founder of artificial intelligence and media analytics company, Blackwood Seven.
Just over a decade ago new digital technologies ripped down the boundaries between these industries and the mobile phone emerged as the device that could synergise the strengths of both areas seamlessly. While the convergence of these industries proved massively profitable to businesses and beneficial to consumers, it changed the nature of the telecommunications market forever.
As such, digitalisation and innovation have characterised the industry over the past decade. The emergence of the smartphone, and brands such as Apple in particular, have left telco providers in a highly competitive environment as they try to sell the best contracts for the newest devices. As smartphone brands try and maximise their own product launches this issue is only compounded, with vast swathes of consumers frequently whipped into a frenzy to switch to the latest tech as quickly as possible. In response we see 24-month contracts being reduced to 12, doubling the pace of potential attrition as providers gamble this against giving customers the flexibility they crave.
The pace of change in hardware is also being chased by rapidly progressing software – streaming, and mobile games for example – leaving handset producers struggling to produce a battery life long enough to sustain all our mobile activities. For telco providers this means contracts have become less about voice calls and increasingly became about text and images. As quickly as the technology changed consumer habits, it changed telco marketing too – packages offering unlimited minutes have been quickly replaced by those offering data, or even free streaming on services such as Netflix.
As such it’s not an easy job being the CMO of a telco. On the one side you are faced with a market that is as dynamic as it is competitive, with both products and consumer behaviours evolving at unprecedented rates. Consequently, this fast-moving environment means that for businesses to stay on top, vast amounts of investment must be poured into marketing and media.
This is the other side of the pressure facing CMOs. The pacey nature of the telco market means that quarterly metrics and traditional marketing success indicators are less effective at providing CMOs with an accurate guide for plotting the success of their marketing strategies. And on top of all this, CMOs in every industry are under increasing pressure to be more accountable for their marketing spend and demonstrate clear ROI.
For most media channels, demonstrating true marketing effectiveness and a directly calculable ROI is extremely difficult. For example – if you were attempting to calculate how many new customer acquisitions a TV advert generated, it’s tempting to simply compare the sales during the campaign’s active periods to another benchmark. However, correlation analyses should be treated with caution – such practices are never truly reflective and fail to consider the wider, encompassing factors that contribute to these metrics.
But we live in exciting times. The ever-increasing wealth of data and computational power available to us means that technologies such as AI and machine learning are becoming powerful solutions to the problems faced by telcos and their CMOs. At Blackwood Seven we found that a media plan requires roughly 5,000 decisions – so it’s easy to see why AI is an attractive option for taking over that heavy lifting.
Systems that use intelligent machine learning can absorb masses of data and identify what factors contributed the most to a certain metric at any given time, allowing marketers to properly analyse the ‘true’ effects of their advertisements.
These models also give marketers a strong, fact-based foundation for discussions around performance. For example, one metric that is often vaguely scapegoated is the weather, but machine learning models can accurately quantify the effects of weather on campaign performance, as well as other external factors such as subscription prices, seasonality, product and price changes in competitors, macro-economic factors, holidays, and more.
A fast moving, dynamic market like the telco industry requires an intelligent, dynamic system that can keep pace. Machine learning models give marketers a more granular, detailed understanding of their marketing effectiveness. This allows marketers to employ different media mixes for different purposes, such one for retaining current customers and another for obtaining new ones, enabling them to keep accurate pulse on their effectiveness.