According to a Salesforce Research paper published last year, 79 percent of customers now expect offers and recommendations from companies to be personalised based on what they’ve already bought.
It’s part of a trend we’ve seen emerging for years in commerce, a trend that is now beginning to reach critical mass. If a brand neglects to give their customers the uniquely tailored experience they’ve come to expect, they’ll simply find it elsewhere, usually with a competitor.
This failure to adapt to a modern customer experience is the downfall of many businesses that would otherwise have been very successful.It’s difficult for sure, but thanks to machine learning and AI technology, it’s getting easier. But what does it all mean, and where should businesses even start?
It’s hard to imagine in 2019, but transactions were once regarded as ‘one and done’ deals and had no deeper meaning or analysis attached to them. There was an exchange of product and money, and that was that. At most, a customer’s details might have been manually entered into a rudimentary CRM suite (Customer Relationship Management) where their file would surely have gathered dust until the business uncovered their email address to drop some random offers into their inbox in an effortto stir up sales.
Back then, this was the extent of the Customer Experience – detached, directionless, and oftentimes annoying. Customers never liked being sold to, and that’s even more true today.
These days, thanks to artificial intelligence and analytical technology, the insights derived from a transaction are almost becoming more valuable to a business than the transaction itself. The Customer Experience has been completely transformed, and if a business can anticipate a customer’s needs and present personalised recommendations based on transactional data, they’re far more likely to become loyal, repeat customers for years to come.
The modern Customer Experience is about removing purchase barriers, reducing friction, and making it easier for the customer to come to you rather than you necessarily going after the customer. For this to work, businesses need to start really getting to know their customers as individuals.
How can a national brand ‘get to know’ thousands, tens of thousands, or even hundreds of thousands of customers as individuals? Transactional data is fantastic, but it only really paints part of the picture.
To truly transform the Customer Experience, we need to know what customers are thinking and feeling. It sounds insurmountable, but it is entirely possible. It begins by extracting qualitative information from customers directly through things like feedback, ratings, and surveys.
This is the easy part. The difficult part is getting that information into a form that’s meaningful, measurable, and actionable.
Fortunately, machine learning is overcoming many of these difficult barriers for us. Using new technology, we can unlock the value in qualitative opinion-based input and apply quantitative traits that can be used to influence and develop services.
Essentially, machine learning helps make the immeasurable, measurable. Combined with natural language processing (sometimes referred to as NPL), machine learning is capable of extracting keywords or phrases from individual reviews, and then applying that same technique to large scale batches.
Once extracted, the context is considered during a ‘sentiment analysis’ which can determine whether items are being talked about positively or negatively. These two elements combined allow businesses to collect detailed feedback at scale, even picking up on certain elements that may have previously been overlooked. The frequency and context surrounding particular items can help steer the attention of the business in the right place, influencing overall strategy and behaviour.
As well as transactional data, which can help shape the individual experience in terms of predictions and buying patterns, machine learning can help shape the customer experience en masse and transform services to better suit the market. Extracting ‘in the moment’ feedback for immediate response is great, but there’s also a lot of value to be gained at other points in the customer journey, away from the transaction itself.
This ability to rapidly analyse thousands of customer touchpoints throughout their time with a brand can help that brand identify particular behaviours or sentiments as they are emerging.This is where data translates into valuable, actionable insight.
Similarly, the qualitative, opinion-based insight from customers could be combined with transactional data, stock levels, product specifications and even a customer’s browsing habits – all to deliver a service that preempts the customer’s needs and makes their decision to purchase completely seamless.
To see this innovative technology in full swing, we can turn our attention to the health insurance market. Some pioneering brands in that industry have embraced the Internet of Things (IoT) to personalise insurance policies and deliver value that specific to each individual.
For example, tracking heart rate data, step count and general activity through wearables can demonstrate that an individual is healthy and active, therefore granting them access to better deals and cheaper insurance. Effectively, it’s a way for health insurers to do what they’ve always done – evaluate risk – but with far more data at their disposal. Factor in other IoT technology such as smartphones and smart appliances, or black boxes on cars, and you begin to see how this tapestry of technology can be woven into a highly personalised and desirable service.
Regardless of industry sector, machine learning and natural language processing has been the missing piece of the jigsaw when it comes to providing a truly unique and customer-driven experience. Combined with data from other channels, customer insight is allowing businesses to learn more about their target market and break it down into very specific and detailed segments.
Not only can this inform business strategy and the development of services, it can shape all aspects of business marketing and give sales teams the insight they need to attract, retain and delight customers. The beating heart of this elaborate process is the customer experience platform – allowing verified customers to engage with brands easily and share their opinions on their experiences in a meaningful and valuable way.
Consumers give brands less and less time to get it right when it comes to Customer Experience.
A survey of 14,000 consumers around the globe published by consulting giant PwC last year found that almost a third (32 percent) would leave a brand they liked after just one bad experience.
That doesn’t leave much margin for error. Brands need to ensure they are using every tool at their disposal to understand what motivates customers and how their demands are changing.Knowing this can make all the difference to securing sales, driving up revenues, recruiting to loyalty programmes or hitting higher customer experience benchmarks.
The picture is complicated by the multiplicity of channels and platforms we increasingly use as consumers. Opus research in 2016 found that customers used an average of 3.5 channels to complete digital shopping and other brand-related tasks.
Retail organisations know that in the age of multiple channels and easy-come-easy-go consumer interactions they must get as close to customers as possible, understanding their changing requirements and responding to their concerns as quickly as possible.
AI turns customer data into a powerful source of insights
Almost every consumer-facing enterprise is now using analytical techniques to extract insights from transaction data, but until the advent of artificial intelligence (AI), it was not possible to analyse customer reviews as effectively. Firstly, they were too numerous and secondly, how do you obtain actionable insights from what customers think?
The answer is that AI uses natural language processing (NLP) and machine learning (ML) and together these two technologies make it possible to automate the analysis of thousands of reviews. This is about more than just picking out particular words such as “camera”. ML applied to NLP detects the sentiment behind the words so that what consumers think about the quality of the camera is revealed as well as particular aspects such as its lens or range of shutter speeds.
Companies with access to a platform that applies these technologies to the insight of real customers, obtain hugely valuable, easily-understood insights into what pleases or irritates customers. AI can spot trends in anything from fashion to festive toy-buying or travel destinations before consumers themselves are aware of them. On a personal level, real-time customer review insights enable customer-care departments to respond quickly to consumers who express their dissatisfaction about some aspect of a product or service, building trust in a brand as a result.
Uncovering the unknown knowns in customer data
Companies can use AI to uncover from customer data what may be eluding them through conventional means. If we take the example of two car dealers selling the same model of four-wheel-drive vehicle, they can deploy AI to analyse the feedback their customer’s leave, to discover why they generate very different levels of satisfaction in their respective sets of customers.
ML and NLP can rapidly drill down into the insight left by customers and blend all the available data to find out the likeliest causes of these problems from hundreds of possible factors. It can achieve in seconds what may take a member of staff many days, without the same guarantee of AI’s accuracy.
Blending customer data for greater insight and personalisation
The capacity of AI to digest masses of information and teach itself about significant patterns, means that customer review data can be blended with the other data that enterprises hold, transaction data being the most obvious. This can be combined to provide more wide-ranging insights that increase efficiency and responsiveness. The range of sources can include stocking and supply chain data, or details of finance deals, after-care packages, warranties, or product specifications.
This also feeds into the necessity for personalisation in customer engagement. With increasing numbers of consumers becoming inured to standard advertising or content delivered directly by brands, personalisation is necessary to cut through the noise. Individual transaction histories along with browsing habits or inquiry logs can be combined with review insights. Using this data, AI automation gives a company the ability to intervene with offers, notifications and discounts just as consumers are hovering close to the point of making a purchase or in danger of dropping out. At these moments, a single timely intervention can be decisive.
AI-driven insight is essential for you to compete with the retail giants
The retail giants such as Amazon are fast developing their capabilities, honing predictive analytics so that recommendations are personalised and based on the individual’s history, rather than on what other customers “also bought”. The rest of the retail world cannot afford to overlook AI when it is sitting on masses of valuable information in the form of customer insight – especially when they are from real, authenticated customers.
AI solutions are fast developing right across retail, such as the rise of conversational commerce through applications such as Google Duplex, or the emergence of visual product searching in which a mere image is sufficient to trigger a wealth of results.
Many AI solutions will struggle to make any impact in the real world of retail, whereas the insights delivered from the feedback of real customers are available now and can become the catalyst for transformed levels of customer engagement.
Customer Experience has become the main battleground of competition and personalisation one of the most effective weapons. No retailer can afford to ignore AI and the insights it delivers from real reviews if it wants to get ahead of the competition.
Can we foresee a world where Artificial Intelligence (AI) recommends who people should have children with, where they should live, and what their next job should be?
This is not the fantasy of an over-stimulated script-writer. AI, fuelled by ever more powerful algorithms, is capable of constantly updating its knowledge of individual consumers’ patterns of behaviour, private lives, hobbies, likes, and dislikes.
Using such information and cross-referencing it with masses of other demographic data it has the predictive capacity to make recommendations with high degrees of confidence. After all, dating apps and websites use algorithms to match their customers as effectively as possible or to give them advice on how they can improve their chances of finding the perfect partner.
In reality, few citizens want their private lives governed by algorithms, however sensitive. But in the decisions they make as consumers about what to buy or where to go on holiday, AI (an umbrella term that includes natural language processing and machine learning) is already a force to be reckoned with and is set to become far more influential.
Analysts at Gartner predict that 20 percent of citizens in developed nations will use artificial intelligence assistants to help them with an array of everyday, operational tasks by the year 2020. Millions of consumers are already enjoying the benefits of using voice-activated devices. In the UK, Ofcom reports that 13 percent of households use an AI-driven smart speaker such as Amazon Alexa.
The growth of consumer-facing AI solutions is putting greater power in the hands of the customer contemplating the purchase of goods and services. He or she can interrogate the solution and rapidly get a comprehensive picture of what is available, and where and what other consumers think.
Consumer-facing businesses will have to respond. They will have to optimise their entire operations by integrating data such as internal sales and customer records, competitive intelligence, trend analysis and social media preferences. Instead of keeping all the data in separate departmental pools, they will need to bring it all together in a data lake where it can be analysed. Then they can create customer profiles or personas that bring them closer to customers before providing hyper-personalised services, recommendations and updates that are intensely relevant to what individual consumers want.
Is word of mouth changing?
One important source of insight is often overlooked, which is customer sentiment. Every day millions of customers leave opinions about services and products that offer a wealth of insights to fellow consumers. That covers almost every aspect of buying, receiving and using electrical goods, clothing or car parts, or the quality of service provided by legal firms or estate agents. Smart, AI-powered, highly agile customer insight platforms are giving consumers rapid access to the accumulated wisdom of thousands of real fellow-customers.
AI is able to filter out the particular aspect of the product or service that most interests them or to provide an accurate insight into sentiment about a product or service. This is a very potent consumer tool. Research among 2,000 UK consumers by Feefo conducted this year found that 94 percent of respondents now turn to online reviews before buying products or services.
Consumers can get the information they want more quickly than ever
AI will rapidly analyse the information to keep consumers informed in real time. Take a national double-glazing company, as one example. Customers in Bristol may suddenly have started experiencing difficulty obtaining quotes or find installation is shoddy, whereas customers of the business elsewhere in the country are happy.If you live in Bristol that information is likely to be critical to your choice of supplier.
Equally, two retailers selling the same model of washing machine may generate very different levels of satisfaction in their respective sets of customers. Consumers can quickly drill down and find if these problems are serious enough to affect whether they go ahead and buy a new appliance with a specific retailer. Gone are the days of relying simply on word-of-mouth.
In the travel industry, AI is giving consumers greater levels of insight that they use before deciding on a significant purchase. Smart insight platforms give the millions of consumers contemplating a weekend away or a month in the Maldives, the ability to extract key information from thousands of reviews about any aspect of a holiday, from whether a particular hotel is suitable for their age group or a villa genuinely has disabled access. The power is at the fingertips of the consumer.
Transforming the service sector
The same forces are at work in the service sector. Smart insight platforms provide job-seekers with accurate and up-to-the-minute feedback on how agencies and individual recruitment consultants perform. Candidates can see for themselves whether an agency is suited to their needs, relying on information that comes from real clients and is on a reputable platform operating nationally across multiple industries. They are more likely to put their faith in such information than in the opinions of a limited number of relatives and friends.
In myriad ways, AI will hand power back to the consumer, enabling them to make smarter decisions more quickly than ever. To respond, any customer-facing business will need the ability to detect trends in sentiment and how they relate to specific aspects of their products or services. Research has found that while 78 percent of businesses monitor the “voice of the consumer”, less than a quarter feel they have access to the insights they need to transform their organisation.
That must change. All businesses and professional services organisations need to take customer insight more seriously and use AI-powered platforms to obtain intelligence about their customers that will give them a major competitive advantage. From supply chain managers to online helpdesks and store assistants, access to this information in real time allows everyone to meet the requirements of a much more demanding market of consumers using AI to shape their decisions.