How to deliver value from AI-at-Scale
August 2024
AI is ‘business critical’ but why do so many enterprises struggle with implementation?
Artificial intelligence (AI) is steadily moving closer to the centre of enterprises as we recognise its potential to boost productivity and revenue. However, a recent study by Salesforce indicates that while 85% of IT leaders believe that AI is key to improving productivity, the majority of businesses (81 per cent) cite data silos as a hindrance to digital transformation efforts, with 62% reporting that their data systems are not configured to fully leverage or scale AI solutions.
Clearly, there is a disconnect between what we intend to achieve with our AI projects and what is actually delivered. As this report has shown, AI is a promising technology, but that doesn’t mean it is easy to implement as part of a company-wide digital change program. In fact, as a tool, it is entirely dependent on the data it is provided. Similarly, if there are bad processes in place, adding AI and other automation solutions will only accelerate these issues.
The problem lies in how we approach digital transformation in the first place. Too many of us get caught up in the hype surrounding these new technologies and adopt them almost overnight, without necessarily identifying how they will improve our business performance or customer experience.
In fact, most of us, when committing to an AI program, never consider reviewing our data architecture or current operational model prior to implementation. It is almost as if we are adopting AI as if it is going to solve all our problems, rather than establishing why we have chosen this particular tool and what we are trying to achieve with it.
For example, in the same report, Salesforce found that 95% of IT leaders identified integration issues throughout implementation. Interestingly, only 26 per cent of these organisations felt that they had established a completely connected user experience across all channels. This is the result of how we manage, store, and share data throughout the enterprise.
Most digital change programs are extremely complex because there are so many teams, processes, and systems involved, many of which are siloed or disconnected. Therefore, when implementing AI, it’s important that our IT teams establish a unified data model for the entire business. That is, one streamlined data cycle, where all data is easily shared between teams. Most importantly, all data needs to be accountable, accurate, and actionable across each stage of the business cycle.
If our data architecture is plagued by inaccuracies, poor processes, or if data is unstructured and siloed between departments, how can we implement and use AI effectively? Especially when the data that we do have is not a true reflection of how the business is currently operating.
Any business that wants to make the most of AI needs to take a step back and review its current operational model. It is important that we establish our ‘digital maturity,’ that is, where we currently stand regarding our digital transformation journey. For example, are there any bottlenecks or unnecessary processes that are hindering business performance? Perhaps there is an issue with how contracts are updated and shared between the finance and legal departments. It is important that we are self-reflective and really take our time to review the way we currently operate so we can get a clearer picture of what issues need to be addressed—especially before we start adopting new technology or scaling these solutions across the whole organisation. Failure to do so will prove costly and time consuming, as we will likely encounter further challenges in the long run.
By doing this, we will be able to establish where a particular technology or solution would be better suited and where change needs to be made. Over time, we will have a much clearer understanding of how data flows between our systems, teams, and throughout various departments, and we will be able to better align our go-to-market (GTM) activities.
Digital transformation is all about reconsidering the relationship between people, processes, and data. Technology does not need to be groundbreaking; we should start by integrating systems and streamlining processes. All data and workflows need to be properly structured and fully optimised. Most importantly, digital transformation should be treated as a continuous journey; it shouldn’t be rushed, especially if we are hoping to use this new tool to make major strategic decisions for, or changes to, our organisation.
We should ideally focus on identifying existing operational issues and reviewing legacy systems before making any serious investments. Once we have identified an issue that needs to be addressed, we can then start adopting one digital tool at a time, across each department in a phased manner.
There can be no doubt that AI is here to stay. Indeed, it could hold untold potential for us if we are looking to improve our end-to-end performance and gain an advantage over our competitors. However, it only works if we approach it in the right manner and know how to adopt AI effectively.
Read More Productivity & Innovation
