ExchangeWire Research’s weekly roundup brings you up-to-date research findings from around the world, with additional insight provided by Rebecca Muir, head of research and analysis, ExchangeWire. In this week’s edition: App-solute domination; Social media ad growth; Broadcasters top completion rates; and Machine learning disruptions.

App-solute domination

Apps have dominated mobile time almost since their inception and reached an important milestone in 2016, accounting for over 50% of the UK’s total digital minutes, according to comScore.

Growth in UK digital time is led by mobile internet consumption but, more specifically, mobile apps are dominating. Total minutes per month on desktop were down in September 2016, decreasing more than 12% year-on-year. Mobile app consumption is the real driver behind this ascent – compared with March 2016, total minutes spent on mobile apps are up more than 14%, representing 82% of the total time spent on mobile and more than half of all digital minutes.

The combined UK app minutes in September 2016 totalled more than 86 million days. On a per-user basis, average app time is up 14% in the last six months; and we now spend, on average, almost 63 hours a month on mobile apps, versus 14 on mobile web. In September 2016, in the UK, more than 51% of total app time was spent on one of the top-20 apps.

Social media ad growth

Global advertising expenditure in social media will grow 72% between 2016 and 2019, rising from USD$29bn (£23bn) to USD$50bn (£40bn), according to Zenith’s Advertising Expenditure Forecasts.

Social media advertising will account for 20% of all internet advertising in 2019, up from 16% in 2016. Social media advertising is growing at 20% a year; and by 2019 will be just 1% smaller than newspaper advertising (USD$50.2bn/£39.9bn for social media compared to USD$50.7bn/£40.3bn for newspapers). By 2020, social media will be comfortably ahead.

Online video advertising is growing almost as quickly as social media, at 18% a year; and by 2019 it will total USD$35.4bn (£28.1bn) across the world, fractionally ahead of the amount spent on radio advertising (USD$35bn/£27.8bn).

Zenith forecast global advertising expenditure to grow 4.4% in 2017, the same rate as they estimate for 2016. This is a strong performance, given that the unexpected results of the UK’s referendum on EU membership and the US presidential election have increased political uncertainty and raised the risks of restrictions to international trade.

Broadcasters top completion rates

Broadcasters saw the highest completion rates for pre-roll ads, finds Ooyala’s Q3 2016 Global Video Index. This group averages better than 85% for pre-rolls on computers, smartphones, and tablets, and even reached 94% on connected TVs.

Mobile pre-roll ad impressions for broadcasters Q3 spiked to nearly 29% in the quarter. Combined with tablet pre-roll ad impression (15%), the two delivered 44% of all broadcaster ad impressions, on par with the share of impressions delivered on desktop and laptop computers (44%). Connected TVs delivered the remaining 12% of impressions.

Mid-roll ads are taking share from pre-roll ads. By the time viewers move several minutes into a video, they’re more likely to tolerate an advertisement. Tablet devices delivered nearly 38% of mid-roll ad impressions for broadcasters in Q3, more than three times the number of ad impressions on smartphones (11.2%). Publishers, meanwhile, generated more mid-roll ad impressions on computers (68%) than on smartphones (30%) or tablets (2%).

Machine learning disruptions

The big data market will grow from $1.7bn (£1.4bn) in 2016 to $9.4bn by 2020, comprising 10% of the overall market for information management tooling, research by Ovum suggests. While the volume of sales is, and will continue to be, North America-dominated, the rest of world will also be posting marginally higher growth rates.

The study finds big data posts growth rates approaching 50%, compared to 5-10% for more mature areas such as business intelligence, data management, enterprise content management, and enterprise performance management.

For upper management, it is all about big data that can enable them to deliver new products and services, engage more intimately with their customers, or operate with superior risk mitigation and security.

Machine learning will be the big disruptor. Many enterprises are already taking advantage of machine learning and artificial intelligence capabilities, but do not know it. Only 38% reported using those capabilities to automate routine tasks, yet 88% said they used analytic tools that incorporated predictive analytics, automated written reporting and communications, and voice recognition and response.