Should you bother with real-time data?

Real-time insights are essential to adapt to a changing consumer landscape, but companies that ignore trust and transparency as part of the process are risking it all


The advice that “trust takes years to build, seconds to break and forever to repair” is attributed to an anonymous sage, which is good news for the sage because the dearth of real-time data means they’ll escape an endless stream of personalised ads.

But it’s wisdom that brands would do well to heed. Now more than ever, given that consumer trust is so difficult to earn and easy to lose, and organisations are becoming increasingly reliant on customer data to manage sales.

The Edelman Trust Barometer Special Report, published in late-June, found that, after price, the most critical factor in a customer’s purchasing decision is trust. “If trust is a key consideration for consumers, it must be a key consideration for brands,” says Henk Campher, vice president of corporate marketing at social media management platform Hootsuite.

However, consumer trust has been eroded in the last six months if engagement from brands has been lacking, or tone deaf, according to new Pegasystems research, which reveals the extent of damage the coronavirus pandemic has caused for businesses’ relationships with their customers.

More than a third (36 per cent) of respondents say they lost existing customers during the pandemic due to failings in their communications. And a similar number (37 per cent) admit to communicating at least one message that was poorly received and dented brand reputation.

It’s not easy for brands, though. The January State of the Connected Customer report from Salesforce highlights a rise in consumer expectations, while stressing four in five consumers won’t buy from companies they don’t trust.

Timing is key to real-time data success

The research shows almost three quarters (73 per cent) of customers think companies should understand their needs and 78 per cent expect consistent interactions across departments. And to make that work, real-time data is required.

“Brands that deliver connected, multichannel and personalised experiences will earn consumers’ trust,” says Adam Spearing, Salesforce chief technology officer for Europe, Middle East and Africa.

Personalisation perhaps feeds from trust as much as it drives it

“Having a 360-degree customer view is crucial for enabling brands to have more personal and contextually aware interactions with customers. For retailers, this may be understanding the most appropriate time to offer customers in-store or online discounts. Whereas manufacturers can get ahead of demand based on what customers usually order at a specific time of the year, based on decades of data intelligence.”

And if companies can use real-time data to communicate with customers at particular times, and it feels sincere and authentic, then brownie points will be won. “Brands can build trust through meaningful interactions with their customers, anticipating their needs and delighting them,” says Spearing. As an example, he lauds Premier League football clubs that send personalised messages from star players to supporters on their birthdays.

Personalisation is a risky business

“The more valuable an interaction is for a customer, the more inclined they will be to continue to trust a brand to use their data appropriately,” he says, though warning there is “a fine line” to walk. “Only if brands use the data respectfully will they gain that trust.”

Andrew Hood, chief executive of data analytics consultancy Lynchpin, is equally ambivalent. “Personalisation perhaps feeds from trust as much as it drives it,” he says. “While I might be happier to share my data if I receive a better, more relevant experience in return, if I don’t trust you as a brand with my data in the first place, I might not feel confident enough to make the first move.”

M&C Saatchi’s senior art director Tom Kennedy is treading carefully and acknowledges the risk that comes with data-driven personalisation. “In January, Aviva addressed its whole email base as ‘Michael’, proving that with even the most basic data, mistakes can happen,” he says. “The assumptions, errors and insults will be amplified with each step more personal.”

Increased awareness of data privacy

Hunting for real-time data can be viewed as insidious and creepy, and there are many instances where organisations crossed the line. Cassandra Moons, data privacy officer at navigation technology firm TomTom, recalls how in 2012 American retailer Target supposedly worked out a teenager was pregnant before her parents through data mining. “Knowing intimate details about your customer that they have never told you can make people very uncomfortable,” she says.

More recently, consumer trust has been chipped away by serious data breaches. “Using data to personalise communications could be the tool that destroys people’s trust in advertising if not used smartly and respectfully,” says Megan Jones, senior planner at R/GA London. She points out that record numbers of people are using internet ad blockers and search engines protecting privacy, such as DuckDuckGo.

“This shift is symptomatic of greater public understanding around data due to Cambridge Analytica’s influence in the Vote Leave Brexit campaign, as well as greater awareness of data privacy through the launch of the General Data Protection Regulation two years ago,” says Jones.

Trust second only to price

Don’t rely too heavily on personalisation

Because customers arguably cherish personal data more than before, she questions a market strategy founded on real-time data. “Almost a decade ago, easyJet stopped investing in Google search terms and moved that budget into more traditional media to deliver phenomenal results. The company saved £6 million a year and there was a 95 per cent rise in seat sales,” says Jones.

“Similarly, last year adidas’ econometric analysis showed they’d been relying on ‘personalised’ communications too heavily as it was the broad brand-building communication that got them the majority – around 65 per cent – of their sales. And let’s not forget that Amazon, hailed as an exemplary data company, was the fifth-highest investor in traditional media in the UK in 2019, with a spend of £114 million, £26 million more than the year before.”

Lynchpin’s Hood concludes: “Ultimately, privacy and personalisation, using real-time data, go hand in hand. And brands that are transparent with the former are more likely to be able to deliver on the latter effectively to their, and their customers’, benefit.”

This article was originally published in Raconteur’s Future Customer report in September 2020

Grounding sci-fi ambitions in reality

Ridley Scott’s 1982 cult film Blade Runner, based on Philip K. Dick’s science-fiction classic Do Androids Dream of Electric Sheep?, came of age five months ago: its dystopian futurescape was Los Angeles ablaze in November 2019.

While some elements accurately hit today’s world, now stricken by the coronavirus pandemic, the planet is dangerously warm and computers can be commanded by a human voice for instance, other predictions fall short. High-collar, full-length trench coats are unfashionable, flying cars have failed to take off and, most pertinently, so-called ‘general’ artificial intelligence (AI) does not exist.

UK AI businesses

Sci-fi is increasingly becoming sci-fact, admittedly, but a technology that can replicate a range of highly advanced human characteristics – the basic definition of general AI – does not walk among us, yet. Moreover, the so-called singularity, when machines achieve sentience and technological growth becomes uncontrollable and irreversible, is some distance away, most experts say.

“Think of general AI as HAL from 2001: A Space Odyssey, or Skynet in the Terminator series,” suggests Bernd Greifeneder, founder and chief technology officer of leading automated-software organisation Dynatrace. “We’re currently nowhere near that becoming a reality, with estimates ranging from it being five years to a century away. Some even believe we’ll never see general AI step out of sci-fi and into the real world.”

Arguably that conclusion is good for the longevity of the human race, though not everyone agrees. “Unless humanity takes a wrong turn, general AI is likely to arrive around 2050, perhaps sooner,” says David Wood, chair of London Futurists. “General AI, handled wisely, can enable humanity to enter a profound new era that I call ‘sustainable superabundance’, in which we can transcend many of the cruel limitations of the human condition that we have inherited from our evolutionary background.”

Gorilla warfare in the technological jungle

Wael Elrifai, global vice president of solution engineering at Hitachi Vantara, pleads for greater caution. “When we achieve general AI, it will drastically transform our economy and society in ways we can’t even predict,” he says. “We’ll be faced with what Dr Stuart Russell, a pre-eminent thinker in the field, dubs ‘the gorilla problem’. Namely, human beings will be outmoded by machines in the same way we evolved to dominate our gorilla kin.

“Finding our place in that future isn’t a decision that can be left in the hands of a few. Technologists, educators, psychologists, policymakers and testing experts must put their heads together to consider how we measure human capital, improve human performance and ensure equity in a world where machine intelligence surpasses human capabilities.”

For the moment, though, narrow AI, which is programmed by humans to focus on a niche task, will have to suffice. The hype around AI has calmed recently, in part because business leaders have realised it is neither akin to the general AI of Blade Runner or Terminator nor a silver bullet. Narrow AI, however, is potent if pointed the right way; those who work out what direction to aim at will triumph.

Besides, as Dr Iain Brown, head of data science at SAS in the UK and Ireland, posits: “The machines have already taken over, to some extent, and with little resistance.” Our smartphones, smart speakers and driverless cars all rely on AI. “Self-learning machines are embedded in services or devices used by three quarters of global consumers,” says Brown, “and algorithms choose what news we read and the entertainment we consume.”

Canny members of the C-suite are beginning to realise the true potential of narrow AI. “General AI isn’t a pipe dream, but it is irrelevant,” says leading futurist Tom Cheesewright. “Focusing on it as a business leader is like seeing the wheel for the first time and spending your time dreaming about a Tesla. Make use of the wheel.”

AI adoption challenges

Targeting niche tasks with narrow AI

Indeed, according to Microsoft’s Accelerating Competitive Advantage with AI report, published in October, businesses in the UK already using AI at scale are performing 11.5 per cent better than those who are not, up from 5 per cent in 2018. Further, the study calculates the number of UK companies with an AI strategy has more than doubled, from 11 per cent two years ago to 24 per cent in 2019. The report also finds that more than half of organisations in the UK (56 per cent) are using AI to some extent, including a rise of 11 per cent in machine-learning from the previous year.

“Narrow AI is certainly a more rewarding prospect for businesses in the short term, as it has more specific applications and so can help to overcome the clearly defined challenges that exist today,” says Greifeneder. “It’s also easier to manage the risks and ethical implications associated with it.” As an example of granting too much autonomy to a machine, he points to Microsoft’s infamous AI chatbot Tay, which began tweeting racist and inflammatory remarks in March 2016, after just 24 hours of exposure to users on Twitter. And, like any tool, AI can be used for good or bad.

Focusing on general AI as a leader is like seeing the wheel for the first time and spending your time dreaming about a Tesla. Make use of the wheel

“We don’t need to wait for general AI to experience elements of AI utopia or dystopia,” says Peter van der Putten, assistant professor of AI at Leiden University in the Netherlands and director of decisioning solutions for cloud software company Pegasystems. “AI is used successfully to understand the structure and function of COVID-19 and to mine COVID-19 research articles. But bias has been creeping into models to determine credit card limits, decide who needs to await a court case in jail or who gets selected for preventive care programmes.”

Why general AI and man must work together

There may be justified concerns about algorithmic biases, how the associated technologies might develop and AI displacing human jobs. But it is critical for business leaders to understand what AI can achieve and it’s certainly not for every organisation.

“If you don’t understand what you are trying to solve first, you are carrying a hammer looking for a nail and AI is going to be of no real use,” says Nick Wise, chief executive of OceanMind, a not-for-profit organisation using AI to protect the world’s fisheries.

For now, the realm of sentient computers seems a long way off. And if we humans are prudent, if or perhaps when general AI becomes a reality, man and machine will augment one another. As Brown concludes: “The future belongs to the cyborg: humans working hand in glove with AI, rather than the android alone.”

This article was originally published in Raconteur’s AI for Business report in April 2020