In the last 10 years, organisations have been taking active steps towards becoming largely insight driven. Yet many businesses still find themselves falling short of harnessing the true value of the powerful duo - AI and Data. Why? Could ‘Cultural Barriers’ be hindering the most well laid of plans?
Helping leaders make more informed, effective, and intelligent business decisions is the core idea behind analytics. Yet, entrenching data, evaluation, and evidence-based reasoning into the organization’s decision-making processes has fallen short of mainstream application in a wide variety of cases. In a report published by Deloitte on Becoming a Data-Driven Organisation, surveyors found that “fewer than four in 10 (37 percent) place their companies in the top two categories of the IDO (Insight Driven Organisation) Maturity Scale, and of those, only 10 percent fall into the highest category. The remaining 63 percent were aware of analytics but lacked infrastructure, found themselves still working in silos, or were expanding ad hoc analytics capabilities beyond silos”.
Reasons for this vast discrepancy ranged from lack of leadership direction to even a lack of trust in a machine’s interpretation of happenings. This would result in falling back onto traditional working mechanics and failure to expand on AI’s value proposition. In each of these cases, the organization’s culture has played a significant role in the acceptance and successful implementation of deep data analytics.
As most executives eventually realize, embedding AI-based decision-making into an organization involves re-wiring the business at a fundamental level. While the initial work might seem substantial, the rewards are to be leveraged well into the future.
Here are some relatively simple ways to help employees become active participants in charting a path towards truly being an Insight Driven Organisation (IDO).
Analytics is already mainstream. However, most businesses still rely on structured data that fits into rows and columns, missing out on critical insights that are to be gained from product images, customer interaction recordings, and social media messaging. Herein lies the argument for adopting technologically advanced AI-driven software that can sift, consolidate and show appropriate recommendations from troves of data. To make the shift, it’s essential organizations give due consideration to these four avenues-
a. Data Science is a Team Effort – Traditional working mindsets are often the most significant resistance to the adoption of AI, making it imperative that welcoming data-driven decision-making starts right from the top.
b. Cultivate the Talent- It’s essential to nurture a diverse team that is curious, numerally proficient, and capable of walking the median line between analytics and business requirements. This progressively creates a shift in the right direction.
c. Go experiential to Data-Driven–Augmenting human decision-making with data-driven insights means trusting the algorithm’s suggestions. It also entails building the autonomy to act on those insights without higher-level authority. If every proposal generated by the systems has to be approved, then the purpose gets defeated.
d. Start with a Minimum Viable Product– While a business system with all bells and whistles may seem ideal, striving for one from the get-go is a fundamental mistake. Instead, look towards speeding up development and achieving gains within weeks with minimum viable products that can iteratively be scaled up.
As the first line of defenses towards building a data-powered organization is taken down, the attention now needs to swing towards setting up the required mindsets for success.
a. Explain the why’s - People often have qualms about adopting newer technology in the workplace, mainly because the need is never understood. Why break something that is working? But by demystifying data analytics and explaining the competitive edge the company is seeking to gain, executives can rapidly help in creating a unanimous acceptance of AI-driven insights
b. Bolster morale and highlight common ground – From the fear of obsolescence to strict reliance on the human rationale to build customer-driven relationships, an organization usually has a lot to contend with when taking basic calls suggested by algorithms. In such cases, making a good story, investing in building the correct information around new systems, and even encouraging adoption through scorecards and perks, can significantly help break down mental barriers.
c. Bridging the gap between technical and business – Much like translating software for languages, a new class of experts are now coming to the fore. Known as ‘Analytics Translators’, they play a crucial role in identifying roadblocks and streamlining business requirements while ensuring adoption goes smoothly.
d. Equal Budgeting for integration and adoption – It’s easy to sidestep this critical element. Companies are often found to allocate most of the budget towards developing the systems, forgetting that workflow redesign, communications, and training are equally vital. Successful implementors of AI spend almost half of the analytics budget on activities that drove adoption.
Once the basic machinations are in place, it’s time to develop acceptance. Facilitating AI implementation with Organisation Learning has shown to increase adoption exponentially. A top-down approach to educate teams works best and helps align priorities across the board.
Aim to target leadership with broader, higher-level educative content. OL should make executives understand the impact on workers’ roles, barriers to adoption, and talent development. Furthermore, offering guidance on instilling the underlying cultural changes required goes a long way towards a smooth implementation.
For those at the core of the business, the ‘Insight-Driven Decision-Makers’, OL should help bring them up to speed with sessions that incorporate real business scenarios. This focus enables staff to derive maximum advantage from AI’s insight capabilities and take effective business calls on, say, product features and launches.
Once a model is securely up and running, an oft-forgotten faction is the analytics and translators (those who act as the intermediaries between IT and Business). Here, the organization needs to up their game by continually focusing on sharpening the hard and soft skills of those directly involved in creating reports and summaries. This will ensure that algorithms being used are established on the most cutting-edge innovations in data science and that analytical reasoning is aptly applied to solve business problems.
Most AI transformations take up to 18-36 months to complete, with larger projects taking a few years. Herein lies the biggest challenges for most organizations – loss of momentum in the light of growing cost and negligible results. Address these challenges by targeting smaller gains in development and enforce adoption by
a. Incorporating learnings and pushing aside misgivings – A key focus area for leadership, who need to walk the talk, steadfast commitment to AI needs to get highlighted even considering failed experiments. By highlighting what was learned in the pilots and reiterating the need for more significant data, organizations can rapidly scale up implementation and adoption efforts
b. Making business accountable and not the analytics team – Analytic teams are simply owners of AI products. It is the business factions that need to own products and see them to successful completion. A scorecard that effectively captures performance metrics is usually an excellent way to align analytics and business goals.
c. Retaining an eye on With vs Without - Comparing the results of decisions made with and without AI can encourage employees to embrace the generated reports. For instance, if on further investigation, executives learn that non-AI-supported forecasts are typically accurate, say about 50% of the time, that immediately bolsters the case for relying on something far more tangible than just intuition.
As is evident, successfully establishing a data-driven culture is the key driver that helps companies scale from carrying out analytics projects in silos to a fully insight-driven organization. Given the research, the willingness to act on analytically derived insights, to make decisions, change processes, and adapt behaviors is best achieved when the organization in its entirety decides to take the leap of faith and encourage a unanimous move to the future of data-driven business.
Such actions create a favorable cycle of events. Interdisciplinary teams bring together the diverse skills and perspectives needed to build new and practical tools. With time, this leads to workers absorbing more collaborative and synergistic practices. As they work more closely with colleagues in other functions and geographies, their vision expands. Workforces move from trying to solve discrete problems in their departments to completely reimagining business and operating models. This culture shift enables the speed of innovation as the rest of the organization adopts the test-and-learn approach that successfully drove the pilot AI initiatives.