Many retailers have adopted automation and AI in recent years to streamline processes, reduce costs and create a better customer experience. However, there are limits to what any company’s automation and AI can accomplish without an equal focus on developing their people and their culture. Now, we’re seeing many companies shift from the automation-focused initiatives of the early 2000s to recognise the role of human intelligence in building smarter, more innovative automations and AI.
Getting the most from AI and automation requires the evaluation skills, expertise and creativity that only humans can deliver. That’s true now, and it will be even clearer and more urgent in the years ahead as digital channels become the arena where retailers compete most fiercely.
AI needs humans to exist and to get smarter, faster
AI relies on humans in two important ways. The first and most obvious is that people – whether data scientists, software developers, analysts or statisticians – are always behind the development of AI. The second reason AI needs humans is to provide feedback and nuance. Without that feedback, AI often delivers experiences that aren’t quite what users want or need.
For example, AI can personalise customer experiences if it’s done in a way that reflects each customer’s journey. However, Gartner has predicted that 80 per cent of marketers will drop personalisation efforts by 2025 “due to lack of ROI, the perils of customer data management or both.” In other words, automation alone is unlikely to meet customers’ needs.
At ClearSale, we often consult with prospective customers who face a similar challenge after fully automating their e-commerce order screening for fraud. Orders that conform to the algorithm’s rules are approved. Those that don’t are rejected.
This sounds like an ideal use-case for full automation. However, the rules that screen out fraud also reject some legitimate orders. When those customers are falsely declined, they often don’t come back. The fully automated system can cost the merchant more in lost revenue and customer lifetime value than it saves by preventing fraud.
ClearSale solves this problem by using human intelligence to evaluate AI-flagged orders to look for legitimate ones. Then, the analysts feed their findings into the algorithm to teach the AI to read these situations better in the future. More good orders are approved, and the retailer keeps more customers.
This combination of human insight with the power of AI allows retailers to create a better experience for their customers while eliminating the financial impact of fraud. This kind of human feedback serves all kinds of AI systems, such as:
Streaming audio playlists. Apps such as Spotify have human playlist editors who curate music and create specialised lists. Yes, playlists can be created by AI that analyses your listening habits. But humans also link certain songs with moods and activities in ways that AI can’t yet.
Social media content screening. Facebook has invested years in developing AI tools to flag offensive content, but the company still needs input from human content moderators in order to keep up with the rapid evolution of memes and new topics such as pandemic-related hate speech.
The evolution of AI object recognition. Google’s Recaptcha AI is a tool that sites can use to make sure bots aren’t filling out their forms. It’s also an object recognition algorithm with uses that include digitising millions of books and newspapers, refining Google Image Search results and helping driverless cars avoid collisions. How? Every time a human user identifies all the traffic lights or street signs in a Recaptcha, the AI gets a little bit smarter.
Clearly, human intelligence is a key ingredient in artificial intelligence. To cultivate the human intelligence that drives innovation, technology companies must put people at the centre of their culture.
Technology companies need to cultivate a people-first culture to innovate better
Retail, like other industries, can fall into a trap that business consultant Clayton Christensen has called the “innovator’s dilemma”. This dilemma arises when actions and experiments that lead to innovation are discouraged or neglected in favour of technology that works right now, or that worked in the past. When this happens, companies get stale, then fall behind when newer competitors deliver fresh approaches.
Here again, the solution is a combination of the human element and technology. The people who build AI systems – the data scientists and developers mentioned earlier – need the freedom to keep innovating rather than becoming locked into one way of doing things. ClearSale CEO Bernardo Lustosa advocates creating a company culture of learning and experimentation based on two major pillars.
One is a structure that frees employees to experiment with new technology and processes, even if they fail, while maintaining the resources to deliver the experience customers expect now. The other is leadership that’s prepared to pivot to meet changing customer needs and market demands during times of disruption.
In 2020 especially, we’re witnessing a shift in retail that demands innovation and responsiveness. Companies that were already fostering creativity and adaptability are coping better with this challenge.
Balance technology and culture during rapid growth
When your company is growing fast or adapting to disruptive change, you may end up with new roles to fill. The challenge is to find people to support your technology while maintaining your culture.
ClearSale has grown from a couple of dozen employees to more than 1,500 in a decade, while nurturing its culture and improving its technology. One way we’ve accomplished this is by hiring for cultural fit and problem-solving skills rather than strictly for technical knowledge.
By training people in-house and encouraging creative thinking, we’ve been able to deliver services that go beyond solving basic fraud problems to solving the larger, costlier problems of false refusals and lost customers.
This approach can work for other companies that rely on technology and automation but need to stay innovative. For example, McKinsey found that more than half of the most digitised companies build their AI capacity in-house and 42 per cent train or upskill in-house talent for their AI initiatives.
At ClearSale, we see AI-driven automation and human insights as complementary rather than competing, within our own company and at our innovative e-commerce clients. Working together, people and AI can maximise the value of automation, prevent unforeseen negative outcomes such as false declines, drive creative solutions to new problems and pivot quickly to meet rapid changes in the market. Retailers who embrace the human/AI partnership are poised to make the most of the current landscape and whatever comes next.
For more information on why it’s important to have human expertise and AI-driven automation combined to fight fraud, please click here.
Original article at Forbes