By Scott Logie, Customer Engagement Director, REaD Group and Chair of the Customer Engagement Committee of the DMA (Data & Marketing Association).
When it comes to building customer engagement, using machines to assist in the process is already commonplace. Machine-learning models, chatbots and AR- and VR-based content are not only being used but also well received by the end user.
In fact, there is plenty of scope for some areas, such as gamification for example, where usage hasn’t quite caught up with consumer demand just yet. As an obsessive daily step counter, if someone tied my daily step count with rewards in exchange for my data then I would sign up immediately.
One of the many questions the use of machines raises is whether it is necessary to tell the end user that they’re engaging with one, rather than with a person: is it human and if not, does it really matter?
Automation needs data to succeed
Automation brings efficiency, speed, accuracy and enhanced information security. Thanks to the proliferation of APIs, the requirement to transfer data is reduced. They can manage the ‘heavy lifting’ required on the huge amounts of data that brands hold. But automation needs quality data to succeed. Research has shown that while 76% of C-Suite executives have AI and machine learning (ML) initiatives in their company roadmap, 75% aren’t confident in the quality of their data.
The same survey found that poor quality data caused AI/ML projects to take longer, cost more, and fail to reach the anticipated results. Implications of data inaccuracy included miscalculating demand (59%) and targeting the wrong prospects (26%).
From a business perspective, then, the automation of data cleansing not only provides cleaner data, but in turn also provides increased insight, better marketing decision-making, triggered campaigns, live personalisation and improved business planning. Automation also improves regulatory compliance, increases customer loyalty, improves the customer experience and provides brand protection.
Automation vs. interaction
Automating the data cleansing process improves your datasets and it is this data that underpins the automated communications with customers and prospects. Does that mean that automated comms can replace every consumer interaction? It seems not: research has shown that that people really want to talk to people when the request is complex or where there is a need for a complaint to be made: 74% of consumers would sooner complain about a product or service to a human rather than a chatbot. But making sure that the conversation can be identified as moving in a particular direction and the right intervention is in place should be more of a priority than whether the end individual knows if the operative is real or not.
And in reality, it won’t be too long before we allow our own machines, home assistants, phones or even our fridge, to engage with a brand’s own machines and make decisions for us – making the identification moot.
AI, data and ethics
One question often raised around the ethics of AI and automation is the continued threat of whether people will be redundant or passed over because of the machines. The important thing to remember is that automation is only as good as the data that drives it, and machines are good at the repetitive tasks that people find boring. This frees up employee resource for empathy, creativity, ethical decision-making and more. Research has shown that UK organisations would prefer to upskill employees from both a business and technical standpoint rather than using automation to fill the skills gap.
Data fuels success
At the end of the day, there are some constants that always need to be in place. The first is identifying who the customer or prospect is and being confident that you have the right person, and the second is having enough data of interest to make the interaction relevant. In amongst all the chat about AI, VR, AR and machine learning, it is vital to remember that it’s the data that fuels the success – or otherwise – of these technologies.
Our main aim, that of giving brands the right to be personal, is never more applicable than when there is a combination of machine and person doing the engagement. Not having the data infrastructure, or indeed the base data, in place means that the discussion about machine vs. person is irrelevant. Ultimately, having a clean, up-to-date, enriched dataset is vital to the success of any AI, chatbot or other technology-based pilot. And when you’ve found the right combination of person and machine, the customer relationship is not only maintained, but is also able to flourish.