The ‘data-driven promise’ is…
Add ‘actionable insights’ and you will be able to grow, innovate, compete and manage risk.
You can definitely file this in the ‘easier said, than done’ basket.
Given our experience of over 6+ years and across 100+ clients in Australia, we’ve seen, heard and experienced the real pains of delivering on this promise.
Here are the top 5:
#1 Data Platforms & Data Pipeline
Heavy upfront investments in data platforms & data pipelines.
Data lakes, enterprise data hubs, big data platforms and the like don’t just work out of the box. They take significant effort to design, build and manage. Once built, you must get data into these platforms in a consistent, secure and scalable way.
#2 Data Scientists
Dependency on ‘hard to find’ & ‘hard to retain’ experience and expertise in data science.
In every industry, there is a shortage of quality data scientists, with the operative word being ‘quality’. If you’re lucky enough to find and attract the good ones, it takes time for them to become familiar with your environment, industry and business priorities. Not to mention holding off the recruiters who are promising them great opportunities elsewhere.
#3 Models in Production
Long lead-times for models to be performant in production.
Building and improving models to find insights can take a long time. Even once completed, these models still have to move out of test and dev environments. Productionising models requires additional skills and lead times before they produce any results.
#4 Siloed Solutions
Siloed dashboards & analytics solutions make it hard to find correlations and root-cause.
Those who have pursued their analytics journey with dashboards and visualisation tools can attest that once these dashboards are built, “no one looks at them anymore”. One reason is the sensory overload – how many dashboards can you look at and what are you looking for? Another reason is that business requirements often change so quickly that an entirely new set of dashboards is soon required.
#5 More Talk, Less Action
Descriptive, not actionable analytics. Hard to get ROI.
Assuming you get through the first 4 hurdles (problems) and assuming there has not been an organisational restructure, change in technology stack and you’ve held on to your data science teams – you are likely to now have the start of some analytics and insights but not quite in a scalable or actionable way.
So after about 18-24 months, your leadership will be asking, “where is the ROI” on this ‘data-driven’ and ‘analytics’ promise?
Autonomous Analytics With Machine Learning & AI
Check out our webinar where we demonstrate how advancements in cloud, automation, machine learning and AI technologies make low cost, low friction, fast ROI insights available to all companies and departments.