Creating an Effective Data and AI Strategy: views from the frontline

Regardless of industry sector, most companies are eager to seize on the transformative promise of AI and push ahead with projects. Once again, boardrooms are turning to their tech teams to get them in race and win. So, how do information, technology and product leaders establish themselves as the ‘go to’ resource for insightful leadership on AI? More importantly, what does it take to craft a successful – and outcomes led – strategy to operationalise this tech across the business?

We asked tech leaders from some of the UK’s biggest and most innovative brands to join us for breakfast and share their AI experiences, hopes, dreams, and pragmatic adoption plans. Here’s what they said.

Bridge the knowledge gap – focus minds on strategic use cases

There was universal agreement that tech leaders have a key role to play when it comes to educating board members and business leaders on the potential benefits and limitations of AI.

Chiefly, these conversations need to be focused on ensuring that everyone understands the resource requirements involved and the importance of committing to use cases that will offer the most potential value for the organisation. 

This can prove a tricky endeavour and some of our tech experts highlighted how ‘Mexican standoff’ style conversations are not unusual – with business leaders saying, ‘we’ve heard there are benefits to be gained from AI, so make it happen’, to which tech leaders respond ‘OK, where would you like us to start?’

Doing AI for AI’s sake isn’t ideal, so prioritising impactful AI use cases that deliver the most direct value for the business should be everyone’s priority. To this end, tech leaders will need to work alongside business leaders to identify the best candidate projects for investment. Ultimately, the consensus in the room was that ‘doing’ AI in a meaningful way is all about starting small and building an AI strategy iteratively as deployments go live and real-world outcomes can be evaluated.

Finally, our contributors highlighted the importance of ensuring that boards and business leaders don’t overlook governance and understand the ethical, compliance and legal guard rails that will be needed when deploying AI.

What did we learn from lockdown?

Lockdown highlighted how long commutes had eaten into time with the family and had reduced the quality of our lives. It also revealed our authentic selves to colleagues. Anikh calls out that, with family members, pets, and personal living spaces on camera during video calls, the people we work with saw a side of us they hadn’t known. It was fun, human and deepened connections.

Lockdown also opened new opportunities for many. Remote working was a great geographical equaliser, says Ray, widening the talent pool now that distances between homes and offices were no longer a factor.

Of course, lockdown revealed some hidden truths as well. One of them was trust. How much trust a leader had in their team soon became evident. Micromanagement crept in in some places. The irony was that many people showed they could thrive in a remote working setting. Anikh adds, the best leaders realise greater team autonomy could lead to greater results.

AI is an organisation-wide endeavour

In addition to identifying business opportunities, our tech leaders highlighted the importance of taking the lead when it came to ensuring that a strong framework is in place for organisation-wide data management.

Similarly, AI projects should feature cross-functional teams that include members from the organisation’s compliance, risk, legal, security and data teams. A top consideration here will be to ensure that AI tools and solutions being considered will not put the business at risk and that long term and lasting value is the outcome of any implementation.

As part of this assessment, consideration needs to be given to how much time it will take to train AI models so these perform as expected – and how this could affect the business case for investment. Similarly, discussions need to be had around whether switching to AI driven logic will endanger existing value chains. For example, exploring if there is a risk of divestment in relation to access to data that is subsequently captured and managed by an external AI tool.

It’s all about the data – becoming a data driven organisation

Tech leaders know that creating a strong data foundation is essential for AI success. That means that data needs to be consolidated, centrally managed, cleansed and accessible – which is by no means a trivial task. Problem is that in many organisations, senior decision-makers aren’t always on board with what it takes to marshal internal data capacity and initiate the data pipelines needed to power AI.

The good news is that everyone in the room felt that the current focus on AI at board level is helping to re-frame organisational attitudes in relation to data. As a result, recognition of data as an active resource that needs to be appropriately managed is growing. And that’s unlocking access to funding tech leaders need to undertake the vital data enablement activities required to underpin new AI-related capabilities.

Indeed, some of our tech leaders drew parallels between the current AI hype to what happened when Cloud first appeared on the scene. For them, it’s providing a much needed conversation opener for wider discussions around initiating data enablement/hygiene programmes as part of wider AI investigation projects.

Who’s accountable?

One interesting discussion point that came up on the day is who ultimately holds the risk in relation to AI projects.

Aside from the importance of ensuring that LLMs are unbiased, transparent and accountable, our tech leaders highlighted the question of accountability in the widest possible sense. For example, who owns the risk when it comes to adopting more data and AI – is it the business or the tech team. And what happens when AI goes wrong?

A key concern discussed was whether the tech team is responsible for testing out the extent to which the board is right. And who determines if the value generated is commensurate with the initial investment made? Importantly, our experts felt it was important for tech leaders not to over promise and to engage in regular dialogue that keeps everyone updated on day-to-day deployments.

With this in mind, the feeling in the room was that it’s important to have open and transparent conversations with business leaders around what problems we want to use AI to solve and who takes ownership of risk as the organisation drives to become a more data driven.

The AI hype wave – driving progress and pragmatic thinking

Everyone in the room agreed that the current AI hype wave is encouraging organisations to look at the potential value contained within their data and ask important questions like ‘what is our data strategy – and how does it need to change?’

In terms of moving ahead with an AI strategy, our experts offered these top tips:

• The starting point for AI is data – and conversations around potential AI use cases will open the door to getting the budget to drive and catalyse the long data journey that is critical to success.
• Build a business case – your starting point should always be ‘why are we doing this?’ and ‘what problems are we trying to solve?’ IT leaders need to work with stakeholders to identify use cases where AI initiatives will deliver against strategic objectives that are important to the business.
• Start small and adopt an experimental mindset – while keeping the investment cost vs. potential outcomes breakpoint in mind. With so much to play for, focus on pragmatic pilots and test first approaches vigorously so learnings can be built upon. Each project will mark a step forward in terms of getting data sets into a much better state overall.
• Focus on governance and compliance – ensuring that you implement a structured approach that features best practices for AI project development.
• Set realistic expectations – help business leaders make the journey from AI hype to AI reality and the importance of viewing everything through the lens of ‘how can we get new/better value from our data’. The aim of the game is to think about tomorrow, while investing in today.

Has this article piqued your interest? Speak to us.