In a recent survey, 76% of C Suite leaders said they expect to be implementing AI technology in the near future*. With this in mind it is inevitable that the next piece of ERP, HR or CRM software that your organisation buys will have inbuilt AI capability; be that for reporting and forecasting, using Agents to create bespoke training packages, or to support your Sales team with real-time insights into customer interactions. And over the next five years these capabilities are set to grow as software suppliers develop Large Language Models (LLMs <- get used to this one, it’s the game changer) that mine the ‘information that matters’, rather than rely on LLMs which have consumed the entire internet.
However, all of the benefits we are about to see through this technology transformation have one characteristic at their heart: data. Whether it’s structured in your General Ledger, unstructured in your HR teams inboxes, a doodle in your creative's notebook, or a mix of everything, data is what will actually drive benefits for your business, regardless of organisation size or industry. With this in mind it’s easy to see why people are calling for organisations to be ‘Data-Led’, although how one might achieve this can be a little unclear.
Veran has been working across multiple industries for over 13 years; we specialise in back office and technology transformations. Our focus is bringing visible and tangible benefits to business and end users and in that time we have supported hundreds of data transformations and gained some tangible insights into what works. Our top three critical success factors are Culture, Capability and …Culture. So important it’s in there twice!
In order for data to be accessible – in the right place, providing the right insight, backing up (or debunking) your intuition - your organisation has to believe that the cost of change is going to be worth it. Removing data siloes, stripping away data as ‘power’, introducing data standards, individual responsibility and ‘fix at source’ are all concepts which we intuitively agree with, but we all know cannot be implemented unless your organisation has a culture that supports individual accountability and mutual benefit.
So as you embark on this next stage of the AI revolution, all areas of your business must understand they are building a different looking organisation:
Culture: bring HR in to the journey and make sure there is an agreed definition of ‘what good looks like’ for a data-led organisation, every staff member knows why you are pivoting towards it and what role they play.
Capability: your technical and data staff can implement standards and controls, exception reporting, data lakes, role-based access controls, automation. Expect these teams to give you good guardrails for success and the tools to support a data-led future, and make sure you engage them with clear definitions of what you need the outcomes to be.
Culture: Data-led organisations value Data Ownership, have clear roles and responsibilities for managing it and hold themselves to account for quality and accuracy. Training (maybe with an AI Agent?), learning and celebrating success remain stronger drivers for compliance and adoption than the possibility of misconduct!
Obviously there are risks you should consider and plan for, including some generic ones for any organisation thinking to put data in the centre. These include:
Data does not replace insight, intuition or talking to people: don’t rely purely on the data for any qualitative analysis, rather use it to inform decision making and strategic direction. If your gut is right, the data should back it up!
Avoid vanity metrics: with loads of standardised data available to generate Key Performance Indicators it can be easy to start measuring anything and everything. Focus on measures that drive action and can be shown to have real substance – does the metric reflect the truth?
Bad data is worse than no data: engage your suppliers in rigorous conversations about any LLM they are proposing; where does it come from? How do they tend it? What are they thinking the next phase looks like?
As an emerging technology AI has some specific data risks: recently I attended a demonstration from a supplier who has been working with partners to create an LLM that is specific to their industry, and the phrase used was ‘we don’t need our systems to tell the difference between a llama and a labrador’. This stuck with me as it points to a trend around refinement and maturity in the sector, steering towards a future which has minimised AI hallucination through having too much data to play with to one which has the right information (be that HR, Finance, Customer, etc) doing the right tasks. It’s hugely exciting for business leaders and staff, and as a technologist who does not want to see a world dominated by massive energy-consuming data centres, a change for the better.
Final note: as part of this exercise I asked ChatGPT the same question as this article, after I had written it. Unsurprisingly there was a focus on the positive, collaborative, empowered decision making possibilities of a data enabled AI future. Not as much on the culture and people aspects, and only a nod to the risks. This human is happy to still be at the helm.
Comments