While many marketing functions are now being assisted by algorithms, one area seems to be immune to automation – customer engagement centres.

According to recent reports, contact centres currently employ millions of people globally and are set to grow by 10% annually until 2019 at least.

This seems counterintuitive. First off, marketers are frequently asked to do ‘more with less’ and keep headcount growth to a minimum.  

Also, technology is emerging which could automate a large part of what these contact centres do. Online chat on websites and bot interfaces for messaging apps could be used to answer most customer queries using artificial intelligence.

So, why are customer engagement centres set to grow by so much? Why are they not feeling the same pressures to reduce size and headcount as other areas of marketing? And will ‘thinking’ marketing systems, or cognitive marketing, ever replace humans for ongoing customer engagement?

To find out, Econsultancy recently hosted a roundtable workshop with IBM Watson Marketing in Mumbai, India. Senior client-side marketers from a wide variety of industries were invited and discussed contact centres and cognitive marketing.

Here are three factors delegates said were critical to the future of customer engagement.

1. Privacy concerns

Today, marketers understand their customers by analysing their online behavior. They look at what their customers click on, what they do on the brand website, and what they buy. 

According to recent research by Pew Research Centre, consumers appear to be quite comfortable with marketers using this data. The study found that only 8% of consumers surveyed considered purchasing habits as ‘very sensitive’ data.

To gather the data necessary to automate contact centres, however, companies will have to start analysing much more customer engagement data. They will need to digitally capture emails, texts, and phone calls to train the customer service algorithms.

Looking back at the study, we see that these data sources are considered ‘very sensitive’ by far higher percentage of consumers.  

Because of this greater sensitivity, brands are holding off on using this data now. Participants felt that their companies do not want to have pushback from their customers when asking permission to use email and voice data to train an automated contact centre system. 

Participants noted that these sensitivities may change, but for now privacy concerns are one reason why brands have been slow to automate their contact centres and use more customer data for cognitive marketing.

2. Technology issues

Another barrier to adopting automation for customer engagement is the fact that many attendees felt that the technology is not yet ready.

There was concern among attendees that customer engagement data ‘has not proved itself to be statistically correct’. Using the combination of incorrect data and automated customer service could potentially cause a major mistake and harm the brand, they asserted.

Instead, participants felt that companies will move toward automation incrementally. That is, they will instead use new technology to change customer engagement programmes piece by piece and avoid trying to automate customer service all at one time.

For example, brands which have a lot of social media inquiries may start by delivering automated responses on social platforms. These could tell customers when, exactly, they will get a more comprehensive, manual response from a customer service agent. 

Watch this short clip where Sriman Kota from IBM Commerce explains more about how marketers will be using new technology such as the internet of things (IoT), image processing, and natural language to enhance their customer engagement.

3. Unanswered questions

Finally, attendees pointed out that there are big questions about customer engagement for which companies do yet have good answers.

These include: 

  • Is it possible for us to sort our customers’ questions and issues into well-defined categories?
  • Do we understand our customers’ problems well enough to respond to them automatically?
  • Can we predict that future prospects will have the same questions and same issues? 

When marketers can answer ‘yes’ to these questions, then customer engagement automation will happen very quickly, participants argued. While there is still some uncertainty about what customers need, however, then companies will struggle to automate customer engagement, even if technology issues and privacy concerns are overcome.

To move more towards automation, therefore, brands should focus on understanding their customers better by methodically tracking customer questions and other issues which they face.

Attendees were unsure about exactly when customer engagement automation will happen, but some felt that it is imminent.  A recent report by Gartner backs up this notion, indicating that by 2020 85% of customer interactions with brands will not involve a human being on the other side. 

Marketers, it seems, have a lot of work to do to make cognitive marketing a reality.

A word of thanks

Econsultancy would like to thank all the marketers who participated on the day, subject matter expert Antonia Edmunds (Business Leader of IBM Watson Marketing in Asia Pacific), and our keynote speaker Sriman Kota (Cognitive Engagement Executive in Asia Pacific for IBM Commerce).

We hope to see you all at future Mumbai Econsultancy events!