A practical guide to building AI & Analytics maturity

Advancing AI maturity is no longer a question of ‘if’ but ‘when.’ It is an opportunity facing every organisation, sector and leader today. 

According to a recent study, 12 per cent of firms have advanced their AI maturity enough to achieve superior growth and business transformation. These companies attributed 30 per cent of their total revenue to AI. Interestingly, they even enjoyed 50 per cent greater revenue growth than their peers during the pre-pandemic times. Additionally, they also outperformed their competitors in customer experience and sustainability. The report further revealed that the share of AI maturity in companies will increase rapidly and significantly, doubling from 12 per cent to 27 per cent by 2024. 

With this in the backdrop, it is really important for companies to have a holistic perspective on AI and analytics maturity that would propel the business and have competitive advantages in the long run. 

AI and analytics maturity warrants a centralised knowledge group or team that oversees conceptualization and implementation of  organisation-wide analytics and AI projects. 

Let’s look at a few industry examples here. 

Walmart operates in over 10,500 stores in 24 counties. The organisation works with data on an enormous scale across organisations, including operational efficiency to user experience. With retail competition at an all-time high, the company needs to evolve routinely and at scale. Certain key parts of its evolution involve using ML and algorithmic data analytics to build immersive and relevant experiences for customers and partners. The retail giant can execute its operations seamlessly owing to its strong AI & Analytics platform (ELEMENT), that enables high-performance innovation. 

Likewise, Amazon has ‘Flywheel’, where this approach enables seamless running of its analytics and AI initiatives. Simply put, ‘Flywheel’ is an engineering concept detailing how businesses can conserve energy and maintain momentum. It encourages a constant flow of energy, proliferating it to other areas of the machine. 

Here’s a flywheel sketch; original sketches were drawn by Jeff Bezos, 2001 (left) and David Sacks, 2014 (right). 

Using analytics and AI within an organisation requires a fair amount of focus and energy. However, once the wheels start turning, there is no stopping; it becomes much easier to maintain pace by ensuring continual boosts to the wheel. Interestingly, with this model, AI innovation in one department or team can be transferred to other areas for business growth.

For instance, if you visit the Amazon Go store to purchase a few groceries for dinner and ask ‘Alexa’ to search for a recipe, its AI model can predict the specific type of saucepan or frying pan you are most likely to purchase. 

Check out AWS’s data flywheel model here

So where does this lead us to; an attempt to outline a…..

Practical perspective on building AI & Analytics maturity in an organisation

According to Forrester, nearly 60 per cent of leaders recognised AI-related initiatives as the most critical factor in successful digital transformation. Moreover, AI is the second-most important initiative for enterprise leaders today, second only to using data-driven insights to improve products and services. 

While big tech and early adopters of AI have managed to develop in-house capabilities and platforms to run their analytics and AI initiatives at scale, most modern enterprises still succumb to siloed data infrastructure, delays in deploying AI/ML models to production and struggle to instil best practices and data-driven culture internally to achieve desired goals. 

So what drives the AI & analytics maturity journey within an enterprise? 

Most traditional studies would tell you that to build AI & Analytics maturity you would need to ensure four factors— 

Leadership 

The analytics maturity of an organisation depends on the outlined strategy as well as the level of executive leadership and support. In his book Competing on Analytics, Thomas Davenport characterises an analytics competitor as an organisation that champions analytics from the top. The in depth research reveals the following critical success factors: 

  1. Top-down support – the top-down leadership and executive support needed to drive an analytics agenda within an organisation is critical to the Analytics capability level of the firm.
  2. The type of industry and size of the firm – the nature of industry is a strong determinant in the analytics capability of a firm. Digital natives such as Google and Amazon have a higher disposition towards analytics initiatives and maturity compared to traditional firms. Studies also indicate that within an industry, the larger a firm, higher is the analytics maturity level. Therefore, industry and size are important variables to be considered from a competitive perspective.
  3. Analytics strategy – firms that have a clear and well-defined analytics strategy and road map have higher business analytics capability levels compared to organisations that have a tactical approach to the competency.

IT (Data, Tools and Systems)

  1. A data ecosystem is one of the key non-negotiable prerequisites for an effective analytics strategy. Successful analytics depends on the type of data made available in an organisation through internal or external sources. The digital world throws up various types of unstructured data (beyond the well-known structured data component) such as image data, speech data and text data that needs to be managed well prior to analysis. In analytics terms, the key imperatives on the data front need to be: 1) Is the data relevant? 2) Do we have enough volume of data? 3) How accessible is the data? 4) Does the data reflect all types of sources? and 5) Can the data be trusted? It is also necessary to study the real value of data to address problems and the likely potential of data to be utilised for development.There are multiple interesting views or categorisation on this dynamic state of data; a common one is referred to as the seven ‘V’ and it classifies data under one of seven ‘V’ dimensions; Volume, Velocity, Variety, Variability, Veracity, Visualisation, and Value.
  1. Enabling technology is defined as an ecosystem that allows for the storage and processing of data inputs using multiple hardware and software processes, such as ETL and governance capabilities, either internally or through partner systems 
  1. Tools are the functionality provided by the technology ecosystem that allow the organisation to access, manage and analyse data effectively and efficiently. Organisations today combine multiple tools with niche functionality to build a comprehensive tool landscape that complements the organisation requirements in data analysis and is related to business analytics endeavours.

Human Capital (Talent, Skills and Competency)

As data becomes available, the ability to analyse and drive insights from data is becoming more critical. This requires human skills and expertise that are extremely dynamic and ever-increasing. Human skills in the business analytics domain include technological skills, business skills and management skills. Most importantly, an appropriate people strategy that helps companies hire the right AI talent and foster innovation, collaboration, and drive a data-driven culture should be in place. 

Organisation (Structure, Process, Culture and Governance)

One of the most important components of analytics maturity is the organisation factor to support analytics initiatives. Various examples assign significance to this factor (of organisation) that is composed of cross-functional data-driven decision-making, embedding analytics within the processes, tightly articulated roles and responsibilities and key constructs around building internally vs. outsource/partner. Studies also emphasise the need for robust processes, the need for a formal organisational structure to enable business analytics and the concept of cost-benefit analysis, thereby pointing towards the overall significance of the organisation factor in business analytics capabilities.

But this may not be enough. The dynamic, competitive environment pushes organisations to look beyond…

Beyond the foundational factors, organisations need accelerators to speed up their maturity journey

Unless an organisation does not have the right use case for its analytics and AI initiatives, it will ultimately become an esoteric proposition, incapable of adding value to one’s organisation. ‘Use cases’ are probably one of the most significant factors worthy of consideration. Use cases are vital resources for better pathways for applications and methods. Companies will need to evaluate best practices internally and also across partners from a wide cross section of industries to arrive at the right use cases. They will need to build their capabilities to extract out of the box ideas across functions and diverse communities. It will be imperative to use originality for determining emerging technologies and build archetypes. This will help enterprises to speedily raise the level of solutions and customise the same for their internal requirements

With innovation moving at a complex speed, enterprises will have to work with multiple partners. These partners could range from established technology vendors and consultants, to start-ups to other vendors. A good strategy should include a firm base for forming partnerships that reinforce execution stability while alleviating risks. Given the dynamic and rapid growth of innovation, it is imperative that without the right partnerships, organisations would find it difficult to scale and maximise the analytics throughput.

  • Data and analytics translators 

In recent times companies have woken up to the fact that to achieve success in AI and analytics needs more than merely hiring data scientists. A curious mix of data architects, Visualisation pros, and data engineers is being well understood by most organisations. However, when it comes to putting the data & analytics to use, communication-or rather, lack of it-between the data scientists and the executive decision makers can cause problems. The two sides often don’t speak the same language and may differ in their approach to and respect for data-based information. Hence the need for Data and Analytics translators. Translators are that ones that propel organisations to achieve real growth by bridging the two worlds and giving business decision makers clearer understanding of the goal.

Understanding AI & Analytics is often challenging, it is rightly called an art and a science. Success is possible by practising evangelism, and consistently following a positive acceptance of change. There is immense benefit in accepting the magnitude of propagating the value of analytics. By focusing on the value of analytics and communicating the same, it is possible to create a compelling motive for change. In order to develop trust, it is imperative to have a strong game plan and measure success. The differentiating factor for organisations lies in the manner of transformation that AI & Analytics teams achieve within the enterprise. These enterprises exude achievement and are feted with recognition and value. 

  • Operationalisation, Governance and standards (Coding, business value) 

Many organisations suffer from poor prioritisation, execution and alignment practices and end up with a backlog of analytics projects. Therefore, an organisation needs to ensure that the resources are utilised in a way that aligns with organisational priorities. Analytics execution within an firm can be segmented into six key areas: building analytic models, deploying analytics models, managing and operating analytics infrastructure, protecting analytics assets through policies and procedures, operating a cross functional analytics governance structure and driving analytics strategy to deliver business value within the organisation. Hence, there is a need to have an effective Business Analytics project governance with established standards that aligns analytics programmes with business requirements.

Conclusion 

As AI and analytics continue to mature significantly, organisations need to start working towards developing a more robust framework to realise the value of AI & Analytics initiatives. A good, well thought out framework acts as a catalyst for business growth and helps companies to differentiate themselves in the ever-evolving competitive landscape and stay ahead of the curve. While leadership, people, process, technology, and infrastructure can get your AI and analytics initiatives to a certain level, the additional accelerators suggested in this article would help companies unleash innovations and scale to a new level. 

This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.

Leave a Reply

Your email address will not be published.