Overcoming the challenge of AI

Overcoming the challenge of AI
19th January 2021

'AI will streamline complex tasks, identify insights that can truly make a difference in the world and make incredible leaps in our levels of productivity.' - Ursula Hardy on the benefits of adopting AI in your business strategy.

Ursula Hardy, Platform Solutions Manager

In the US, 68% of businesses increased investment in AI technologies during the Covid pandemic, whilst one in four CTOs thinks that AI and machine learning will have the greatest impact on global recovery. Even so, it is easy to be daunted by the complexity, the mystery and the possibilities of AI. The reality is that it will streamline complex tasks, identify insights that can truly make a difference in the world and make incredible leaps in our levels of productivity.

But how can you start on the journey of adopting AI in your business strategy when it is such a complex subject?

The foundations of AI

First and foremost, AI depends on a large, healthy set of data. The only way for accurate patterns and insights to be uncovered is for AI algorithms to analyse carefully collected data that reflects reality, otherwise the output will not be reliable. It’s from that data that valid interpretations can be produced. For example, Football coaches are now adapting AI to analyse hours of video footage from past games to provide insights on how to train their teams to compete better. This is only possible with the data collection practice in place.

To incorporate a similar AI strategy, you should first invest in a strong data strategy – data that is captured with intent, consistency and transparency is key. Football clubs have a policy of always recording each match. The process has been for analysts and coaches to spend time watching and taking note of observed patterns and manoeuvres. Now, instead of only relying on human eyes diligently reviewing the footage, football clubs have the additional power of AI to help identify patterns that may previously have been missed.

Data strategy fundamentals

Your company, like many, may have the luxury of developing a fresh data strategy. Often they find themselves with a vast amount of data, and only now are trying to understand what exactly it can do for them. It’s a common pain point for existing data collection to be done without a purpose, or for the data itself to be inconsistent in format, or even that it is hidden within the business across various teams and channels.

So, to advantage of AI, your data strategy needs to address some key principles.

  1. Be objective: what are you trying to achieve with your data collection? Are you tracking metrics across KPIs? Are you gathering data to understand user behaviour? Be aware of the intent behind data capture and categorise accordingly.
  2. Review the data you are collecting: is it in the correct format?
  3. Avoid silos: valuable insights are often best found across areas within the business. Transparency is important.

Fundamental to your AI strategy is the surfacing of data that it is collected with intent, in a consistent format for analysis and made transparent across the organisation. Utilities companies are incorporating AI driven chat bots, but in order for these chatbots to work they must be able to access the data stored in different areas of the business, such as account information, payment history, existing charges, etc. Imagine if a customer were trying to access their account balance, but the chatbot didn’t know they changed addresses? The resulting confusion creates more work and poor results. Basic data strategies need to be in place to solve these day-to-day problems.

Don’t start from scratch

There are solutions to help guide the process when you are tackling how to implement a healthy practice in data collection, or retroactively taking advantage of the vast amounts of collected legacy data for value. Finding the right Digital Experience Platform (DXP) means that businesses must think critically about how their overall business strategy fits with their digital strategy, before finally settling and determining their data strategy. In this practice, the complex questions of determining the best solution will come up naturally if done correctly. Furthermore, the additional burden of keeping up with sporadic digital policy regulation changes, can be handled through the use of the DXP – saving organisations from collective headaches of developers to the C-suite.

A well established DXP will establish a strong connection point with customers. It will also enforce a deliberate data collection policy – as a result, producing a healthy data set that accurately represents reality. With this foundational step established, organisations are poised to integrate AI and can begin to evolve their processes.

You can read more from Ursula in AI Magazine.