“The adoption rate of ChatGPT was over 100m in the two months to May 2023.”
“Data, analytics and AI (D&A) is revolutionising our strategic decisions, bringing cost savings and a deeper understanding of commercial forces. What could possibly go wrong?” Jan Henderyckx, Regional Leader for D&A at BearingPoint highlights the perks and the pitfalls of getting business value out of data.
Thanks to D&A, companies are boosting productivity on a broad scale.
To benefit from the data revolution, companies must transform in stages and:
- Put business action at the heart of the data strategy through empowerment.
- Organise themselves and the data ecosystem.
- Adopt risk-mitigation planning.
The challenge is to transform masses of data into business value without getting lost in it. Unfortunately, many – if not most – Big Data projects fail, with firms stumbling on their quest to become data-driven. In 2017, 60% of Big Data projects were destined to disappointi, while in 2018, 80% of analytic insights did not transform into relevant business objectivesii.
Data ‘undermining’?
Once the organisation and its data literate people are aligned to a data-centric strategy, the next step is to study available data and determine if it underpins strategy or undermines it. This is key in mitigating the risks involved in the transformation.
The data product should lead to tangible outcomes and should therefore be CART:
- Compliant (with ethics and regulations, throughout the entire lifecycle)
- Actionable (can it be applied?)
- Reliable (can it be scaled?)
- Trustworthy (is it reflecting the real world?)
Take the case of a leading financial news provider who needs to deliver data products that draw from a wide range of data sources and formats in real time. The different sources are ingested, curated and processed in its own Artificial intelligence (AI) Generative Predictive Transformer (GPT) based model. The model allows answers to be found for users with questions that range from the data focused “Give me the figures on Taiwanese consumption of sugar”, to the deductive “How can I evaluate this bond?”.
When D&A is applied correctly, data products should seamlessly provide the relevant insights depending on where the data was derived from:
- The company internal data
- Data from the company’s eco-system
- Generally available data
But remember to ask if the model has been fed with relevant and qualitative data. “Computers are useless”, as Picasso said, “all they can give you are answers”. The answers you get may be statistically relevant, but they will not tell you how relevant. After ‘cleansing’ (removing duplicates or misnomers) and analysis, the insights are presented to business leaders.
Data dictum
While the methods are inherently mathematical and, let’s face it, rather complex, the reporting is designed for easy comprehension using visuals and interactive dashboards. The point is to get the data to reveal what needs to be done.
With original roots in finance and pharma, D&A is now used globally, across all organisation types, in every sector. It helps make more informed decisions, more quickly. Yes, experiential methods have their value, but in a world where the only certainty is change, let the data speak.
Analytics help organisations understand the world as it is, with no room for wishful thinking. Now, the question is, can your company afford not to embrace D&A? As long as it’s done properly of course—with the right pre-thinking, planning and risk checks.
i Gartner 2017
ii Harvard Business Review, October 2018
