AI’s Personalisation Paradox
Originally published by Algorithm-X Lab
By Katie King, author of Using Artificial Intelligence in Marketing: How to Harness AI and Maintain the Competitive Edge, published by Kogan Page
The term Artificial Intelligence is actually a misnomer.
AI enables teams, for example sales and marketing professionals, to provide a personalised experience to a user without being too intrusive. AI is already allowing the marketer to optimise websites, personalising them for different users, for example, serving them tailored messages and designs which resonate, based on their profile and needs. The technology behemoths including Adobe, Verizon, Microsoft, Google and Facebook are driving this trend, one which will prove very lucrative.
Marketing and advertising have been quick to adopt AI into their practices, but we’ve only touched the tip of a massive iceberg, and one which is as relevant to operational and HR teams, as it is to sales and marketing professionals. As AI technology develops, reaching individuals at scale will be possible, and embracing the fact that AI solutions will be taking on the repetitive tasks which will allow savvy individuals to actively focus on creativity and strategy—skills that machines simply cannot replicate, yet!
An enhanced AI experience
Subha Ganesh, the former Global Head of Customer Acquisition at UK fashion brand Reiss, explains: “In a digital world, you’re often thinking more as a transaction rather than as actually an experience. But our aim is to bring the experience of the shop to the online world. Technology solutions like AI give the customer the feeling of having an advisor saying, for example, these are the shoes and bag, which will go with that dress. We call that a bundle. It gives them advice on how to style up, with information on what other people have bought. In essence, it gives the customer confidence. They feel like they are getting a style guide or advice from the brand that is supposed to be an expert in that field.
“It’s early days with AI at the moment. To start, we had a few fears. The first was that personalisation might be scary for the customer; we don’t want to cross the line. We needed to know how much is good enough. Just as in a store you don’t walk behind the customer; you give them space. Online, you also need to be there for them to call out to you. The second fear we have as a retailer is whether the data is actually available to make the correct AI integration. The machine can learn so much more if it has good data. This is the reason we work with a third party such as Increasingly; to get data quicker and cleaner.
Reducing guesswork and satisfying customers
“AI will definitely impact marketing in the world of fashion retail. Firstly, it will help the retailer to give the customer what they want. Secondly, it will enable the retailer to identify the correct customer at the right time. This will enable marketers to make decisions and take calculated risks. AI will reduce guesswork and give the CMO confidence that their strategy is correct. Similarly, in merchandising, AI allows the retailer to buy what the customer is looking for and to stock at optimum levels. There will be less analysis of the past, and more focus on future predictions.”
As well as the major tech vendors referenced earlier in this article, it’s important for readers to appreciate that there are many start-ups now on the market, who are disrupting the industry status quo with their AI solutions. One such company is Increasingly, and we hear now from the CEO and Co-Founder Sri Sharma, who works with brands such as Reiss.
“AI will be increasingly used to both scale marketing, for example, AI email copy optimisation. On-site we will see more sophistication in how retailers engage with their customers on a 121 basis. In general, AI will solve all tasks that cost too much to scale manually and secondly AI will help us predict the future better. Automation Insights is helping companies automate content creation. Increasingly AI is helping companies cross sell with massive impact. AI in marketing has already happened, for example Google Adwords is using AI in its core algorithms. Many innovations are on the cusp and proven. We are going to see mass market adoption.
“Companies need to recognise that if they don’t innovate, their competitors will. To avoid getting left behind they have to test innovative solutions. This tends to be easier done by leveraging third party solutions as this feeds into the approach of succeed big or fail fast.”
AI is a general-purpose technology, which touches all businesses and every part of all businesses. But the reality is that it is a long and laborious task to see something happen in the real world. Speed of deployment is key. You need to deliver real life instances as soon as possible, rather than talk strategy continuously. Companies like Kortical in the UK are offering a safe and fast route to embed AI in business.
Businesses that are much more proactive in adopting AI are seeing much greater returns, even at this early stage. McKinsey’s 2016 report confirms this, concluding that those already investing strongly in AI are already making on average 10 per cent more than companies partially investing or not investing in AI. Marketing Director Martin Harrison explains that Kortical’s mission is to make AI available to all types of businesses. They believe it will be as ubiquitous as Excel and that it will have the same impact in a much shorter period of time. You would never have dreamed 40 years ago that Excel would be used across almost every business.
Their platform, The Kore, makes it easy for businesses to become AI enabled, and to therefore grow faster and become more efficient. It enables Kortical to create bespoke solutions for businesses by automating a lot of the data science process. At moment, there simply isn’t the level of expertise available, so companies like theirs are having to do a fair amount of hand holding. Medium to long-term, their ambition is to be software as a service provider.
Predict, automate and innovate
Harrison explains the three main things AI currently does:
- Prediction – every business decision is a view of what the future looks like and what the future holds. The more data driven it can be, the more effective they are. AI and machine learning provide another boost; another layer to that ability to predict. Kortical recently worked on a project with the UK National Health Service (NHS), focused on predicting how long a recipient might have to wait for a compatible organ. The NHS had spent a few years refining this process, but using Kortical’s model, they were able to improve it by 10 per cent in just over 4 intense weeks.
- Automation – Every business achieves efficiency by having repetitive tasks streamlined and standardised as much as possible. AI vastly increases the amount and depth of tasks that can be automated. It brings a layer of cognitive ability to decisions you are making; a layer than hasn’t been there before. As a result, huge efficiency gains can be driven. Kortical recently worked with a UK high street bank, automating an email process. Historically a small business would write to their bank and wait two days for a response. Emails coming into their business centre, which is a £1.4M cost centre, were assigned to the right customer service team. Before that someone had to manually open and read each one to figure out the appropriate ticket to raise. Using robotic process automation and AI, Kortical were able to automate this process in just over one month.
- Innovation – is where things start to get really exciting. As Co-Founder Andy Gray explains: “No business can stand still, particularly in today’s world. The really interesting and exciting aspects of AI are where it’s used to unlock new business models, new market strategies, new propositions.”
An immature market
Kortical have also developed a chatbot called Obi, which does not require vast amounts of training data to get up and running. It can sit alongside a call centre operative and watch how they interact with customers. It learns and in a short period of time, it starts to replicate. A bank would therefore select 2-3 of its best call centre staff to train it. Within less than 2 weeks, it can start to add major value. Incrementally it takes over and grows its own functionality. For most call centres, a relatively modest 10 per cent rate of automation can result in over a million pounds in savings.
The market for AI at the moment is very immature. Businesses are still trying to figure out what their vision is for AI actually is. They are also still trying to understand where their data is and how it can be structured for AI. At the same time, they are grappling with a severe lack of talent. The need is not only for data scientists but also the engineers who can manage the software development to deploy it. There is then a further requirement for project managers who can run as AI project; there is a real shortage in these areas too.
What’s driving the current AI summer
Cost saving is definitely what is selling AI to businesses right now; it’s what unlocks the budget. But we are definitely seeing improved customer outcomes and service, especially in chat bot and call centre space. As it’s not, as some predict, a way to fire staff. Many are using it as a means of scaling more effectively, because their teams can achieve more.
One of the current challenges facing CMOs is that they are inundated with requests from different suppliers for multiple different point AI solutions. If a CTO or a CMO is buying 5-20 solutions, all driven by the same set of algorithms, they are building up a huge amount of technical debt, which isn’t sustainable in the long term.
It’s still early days and some may see these benefits creepy. But the AI is improving all the time, and as we adjust to it, the more it will just feel good and not jump out at you. For years we’ve been learning to work with and be more like machines. AI allows humans to be more human.
Top 10 tips
- Personalisation is a misnomer. Used properly, AI allows us to personalise without being intrusive.
- It is now much harder to make an impact if you retain a traditional approach and ignore AI.
- It is crucial to understand the problems to be solved in order to ascertain which AI technique is best.
- To succeed, you cannot rest on your laurels. Ongoing research and innovation will be key to maintaining a competitive edge.
- We need to teach young people, and our staff, skills such as creative thinking, collaboration, mind mapping and goal setting.
- AI is like a new drug we have yet to become addicted to.
- AI will reduce guesswork and provide businesses with the confidence that they have the right strategy.
- If you don’t innovate, your competitors will.
- AI is in truth still immature and many companies are figuring out what their vision for it is.
- Customers want personalised products not personalised marketing