The “do-or-die” mandate for Digital Transformation, has become the most important strategic decision for all enterprises, in every market sectors, even the traditional ones.
The avalanche of Digital-Native market disruptors in the entire supply chain of every economic ecosystems has forced the seemingly well-established incumbents - some with a bicentennial history - to plan and push for their Digital Transformation.
Statistics show that about 90% of all organizational at global level, has plans for, or have already started their journey towards their future state Digital Enterprise. This is while the numbers on success rate are quite worrying: Over 70% of tech-savvy organizations (like Telecom) and over 93% of traditional organizations (e.g., Oil & Gas or Automotive), have failed in achieving complete transformation due to a number of pitfalls and inefficiencies.
One key prominent factor of their failures is lack of properly designed Key Performance Indicators (KPIs) that is needed for a Digital Transformation and fit for the Digital Economy. That means through their Digital Transformation journey, they cannot get a realistic, forward looking picture of their progress and fail to create the needed transparency to allow for strategic navigation during the change and suffer a fatal disconnect among their ranks that would not allow leadership to cascade their visions down to the teams in the tranches.
Traditionally, organization have under-utilized KPIs as limited measures against compliance and regulatory requirements, with mostly look-back (retrospective) effectiveness.
In order for KPIs to match the dynamics of the Digital Economy, organizations should expand their usage into creating visibility into what is ahead in the market as predictive measures, capable of providing prescriptive triggers with action items for corrective and re-calibrating tasks.
This ability, that would have been the subject of sci-fi movies in the late 20th century, is now quite attainable, and is being adopted at an accelerating rate, thanks to the advances in Artificial Intelligence (AI) and its learning child, Machine Learning (ML).
This expansion has upgraded KPIs from a basic score keeping system used by management of silos, into an Augmented Corporate Portfolio Level Navigation and Business Strategy Decision Support service.
ML-enhanced KPIs (aka Smart KPIs), are learning models by design, capable of predicting the trends movement of future, by learning from the measured values of the past. They get “Smarter” as they receive more training on the data that comes in, and become more accurate, relevant, and efficient, over time.
Smart KPIs: the competitive edge of Digital Unicorns
Smart KPIs are great competitive advantages for organizations during their Digital Transformation journey, and their Digital Enterprise future state. The Unicorns of the Digital Era, like Amazon, Facebook, Google, and Alibaba, have pioneered Smart KPIs and contributed significantly to its development and wide acceptance. They in-turn have maximized their benefit form this advantageous power to push the incumbents of their markets – or any market they have decided to target – to the side and in a number of occasion, off the cliff.
The secret sauce of their benefit maximization from the Smart KPIs is how they have incorporated them into their strategic navigation structure, using the predictive insights and the prescriptive guidance to lead their organizations. This is a whole new level above and beyond traditional use of KPIs as look-back measures, serving a post-mortem, “lessons learned” perspective.
Successful enterprises use multi-level Smart KPIs which would provide corporate guidance at portfolio, solution, program and even team levels.
For example, at portfolio level, an enterprise may choose: Customer Revenue Growth Acceleration (on the Net Promoter Score Per Customer) and Value Streams Improvement to serve as a strategic guidance for the rest of the levels, to provide the baseline on what is the most valuable measure for the enterprise, for their own Smart KPI setting. It also tells each function in the organization what they should focus on in their joint or cascading efforts, and what is being measured and monitored by the leadership.
Smart KPIs can expand an organization’s reach in the ever-changing competitive market environment, through greater personalization of offers for customers. ML can be used to identify customer segmentation schemes that are most valuable to focus on for targeting, and the Smart KPIs can be established based on those segments to measure and visualize the performance of the efforts.
There is no right or wrong answer regarding the number of Smart KPIs to serve the organization, but in general terms, as we move higher on the corporate leadership level, the smaller number of Smart KPIs provides a more effective in creating a holistic view of the channels and decision support required for leading them.
Smart KPIs need Smart Data Pipelines:
Smart KPIs use Machine Learning Models, which have an unquenchable thirst for good quality data. That is what they train on and get “Smarter” as a result.
Providing the Data Pipeline is no easy task even in our current era of Big Data. To identify the most relevant Data and the required features, the organization needs to develop an in-depth understanding of the KPIs that are to be measured and how they will augment the business decision making process.
The data coming from our measurement points and tools is in need of clean up, completing, massaging and preparation. A big part of this process also uses Machine Learning to improve the data sampling process and optimize the collection and enrichment functions.
Smart KPI ML models should be considered as “living and breathing organisms”. Once they are trained, validated, and put into service, they will enter the Continuous Monitoring stage, where their accuracy and efficiency goes under an ongoing check to ensure they are still relevant and return proper measures.
As the data that is fed to Smart KPIs may change over time (Data Drift), and what needs to be measured many need revision (Model Drift), we would need to retrain the existing Smart KPI ML Models using the new Data or revise the models and train them for the revised conceptual model. This ensures that Smart KPIs that is put into service, continue to lend their competitive advantage to the strategic decision making level.
If your organization have an existing MLOps, then this is where your Smart KPIs ML Models need to go through for their data validation, training and model validation and serving. If you do not already have one, you would need to plan for establishing your own MLOps, as DevOps pipelines are not the best fit for supporting them due to their key architectural, testing and serving differences.
Smart KPIs benefits for the enterprises:
A successful implementation of Smart KPIs, if well monitored and maintained, can result in many improvements in connectivity and visibility of leadership’s “smart” and market-relevant decisions across and through the ranks in the organizations. As defined at multiple levels of the organizations, they enable leadership to stay in a productive connection with the middle management, allowing for an effective collaboration between the levels.
Smart KPIs serve as the organizational compass, to guide and lead the entire collective of organization activities in alignment with the corporate strategic decisions. Their chain of transparency and decision support provides the line of sight in both ways and allows for all ranks to see their North Star and compare it with their current direction and re-align as needed.
They also enable the management and leadership to have a live view on the performance of all Value Streams and delivery pipelines and take not only corrective actions to fix an existing issue, but to plan ahead and execute incremental re-alignments in response to the prediction of upcoming trend movements in the market.
Smart KPIs also democratize the predictive and prescriptive information across all Value Streams, Silos, and function groups in the organization. This provides a holistic view on the quality of services provided to the customers, their satisfaction and impact on the revenue across all channels. The cross-org and top-down transparency allows for Smart Collaboration between all teams involved in the Value Streams, all based on live and relevant data enriched with prescriptive augmentation in optimizing their synchronized efforts towards hitting higher marks.
Digital economy is growing 2.5x faster than the Global Economy with 6.7x higher ROI compared to non-Digital segments and is expected to grow to $23 trillion by 2025.
Digital disruptors are making it impossible for non-digital-native organizations to stay relevant and competitive in the market without going through Digital Transformation.
Smart KPIs are establishing themselves as a vital competitive advantage, leveraging the power of Artificial Intelligence (and Machine Learning) to shift KPIs beyond their traditional post-mortem measuring tools into predictive and prescriptive systems that would serve the organizational decision making hierarchy, from strategic leadership to tactical implementation, and democratize that power, insight and transparency across all silos and their serving teams.
It is now the time for all enterprises to embrace the future, by using the foresight and predictive power of Machine Learning through Smart KPIs, and to use its prescriptive power to make “Smarter” decisions for their market survival and expansion.
Arman Kamran is an internationally recognized executive leader and enterprise transition coach in Scaled Agile Delivery of Customer-Centric Digital Products with over 20 years of experience in leading teams in private (Fortune 500) and public sectors in delivery of over $1 billion worth of solutions, through cultivating, coaching and training their in-house expertise on Lean/Agile/DevOps practices, leading them through their enterprise transformation, and raising the quality and predictability of their Product Delivery Pipelines.
Arman also serves as the Chief Technology Officer of Prima Recon Machine Intelligence, a global AI solutions software powerhouse with operations in US (Palo Alto, Silicon Valley), Canada (Toronto) and UK (Glasgow).