A question of language: Getting the AI message across


It is always interesting to attend trade shows and find out what’s new. It is also, however, slightly depressing to hear what’s not new, especially when it is an issue that we have been discussing for some time. At the National Retail Federation’s event in New York City recently, I asked a number of retailers for their views on the trade show. Over and over again, I heard two things.

First, on every booth, everyone was talking about the same buzzwords (omnichannel, machine learning, artificial intelligence and customer experience). My contacts, however, had found it difficult to get any real insight into what those might mean for their business. The second thing was that almost every vendor was talking about “products” and not solutions to problems. Taken together, this meant that most people had left the event without really understanding how their organisation could benefit from AI or machine learning. This seemed like a wasted opportunity.

Will AI deliver benefits?

Many retailers see that AI is finally starting to deliver real-life benefits to early-adopting companies. Most of them see its importance for customer experience and customer service. They may not realise, however, that AI-powered robots could do much more, such as run their warehouses – and even automatically order stock when inventory runs low.

ai message

AI-powered robots could do much more than increasing customer experience and customer service. They can f.ex. be used to running warehouses – and even automatically order stock when inventory runs low.

During the event, I had a long conversation with one particular customer from LATM. We talked about self-learning algorithms and use cases for machine learning. I asked him where he saw potential benefits from machine learning and AI. His response was that there are no concrete use cases or strategy on this topic. Using chatbots could be an interesting use case because everyone can see the potential of using smart agents on digital channels to increase and improve customer service, or simply to reduce costs in customer support. We also, however, talked about different concepts and IT requirements and how to avoid ad hoc approaches that might not scale, but where one department wanted to jump on the hype!

Sign up to this webinar series and learn about the practicalities of artificial intelligence and making AI business-smart

It is certainly true that AI, in particular, IS the subject of a huge amount of hype at present. Everyone is interested. In my opinion, however, companies need to have a certain analytical maturity before they will really be able to make AI and machine learning part of their core strategy. They have to be able to run before they can walk, and many companies are nowhere near mature enough yet.

As a company, you need to know where to use AI and where it will bring value, and that means understanding more about analytics. In other words, the early adopters are getting value because they have this maturity. Simply jumping on the bandwagon will not deliver results.

Investing well, generating value

Part of development analytics maturity is to make sure that you have the right talent and capabilities in place. Thinking about AI as part of a core strategy does not mean that companies have to start with a big-bang approach. The use case-driven approach can provide quick wins and measurable value for the organisation, and it is definitely a good way to start.

What are the prerequisites? AI starts with machine learning, and that requires being able to analyse your data. That, in turn, means that you need the right data. When defining your AI strategy and your first use cases, you need to be confident that you can manage the core steps of every machine learning process. These are:

  • Data exploration.
  • Data cleansing.
  • Model building.
  • Results presentation.

It takes time to develop analytics maturity. A company wishing to invest in – and generate value from – AI and machine learning must follow the maturity path into and on from (big) data analytics. There is plenty of potential, but companies should be sure not to try to run before they can walk, or they are unlikely to see a good return on their AI investments.


About Author

Ioannis Stavridis

Sr Manager Global Practice Customer Intelligence at SAS

Global Practice Customer Intelligence, Ioannis holds a Master of Software Engineering and Economics degree and has a longer than 15-year track record in the area of successful analytical CRM, Customer Intelligence and BI implementation projects. Before he joined SAS he worked for several consulting companies in a managerial position and acted as trusted advisor in several companies. Additionally his experience includes project management in Customer Intelligence, Business Intelligence and CRM projects in many of Europe’s largest international retail, telecommunication and broadcasts, banks, utilities, automotive and companies. His main focus areas are Omni-Channel and Contextual Marketing Management solutions, business case and business value development and Customer Intelligence process improvements. Throughout his career he worked in both national and international companies and gained international experience through working with multi-cultural project teams.  His experience of having been a program manager, a business advisor, and a implementation consultant helps Ioannis to act as a trusted advisor for clients as well as helping them to operationalize their Digital strategy.

Leave A Reply

Back to Top