OPEN FORUM: Demystifying big data, analytics

OPEN FORUM: Demystifying big data, analytics
Data-Science-vs.-Big-Data-vs

Data must be analysed and decisions made in real time — volatility. If you do not act on an alert, you lose the chance forever.

By Zorodzai Chenhidza

“YOUR assumptions are your windows on the world. Scrub them off every once in a while, or the light won’t come in.” — Isaac Asimov
Isaac Asimov, an American author who wrote or edited over 500 books, 90 000 letters and hundreds of short stories, summed up in one word the biggest issue that clouds decision makers’ judgment: Assumptions.
Assumptions are misconceptions that paralyse the decision maker and stops decisive strategic action for achievements of results. Some organisations that have decided not to implement analytics have assumed that analytics (the discovery, interpretation, and communication of meaningful patterns in data) is a passing phase in information communication technology.
They are waiting for the storm to pass instead of riding the wave to get the thrill and the momentum to mount the bigger waves. What will change is how data is captured, stored and analysed. But data will always influence decisions, and cognitive analytic systems only grow smarter and more useful the longer they are in use.
Others say they have no budget for analytics.
When you do not analyse your data, you miss opportunities to cut operating costs or increase revenue.
As long as you operate on intuition and gut feeling, chances are you might never have a budget for analytics. Analytics projects can be implemented in incremental steps that shorten time to value, cater for areas of high priority and build resources for further focus areas. You can start with one project and move on until the whole organisation is data driven, proving the assumption a fallacy.
The term “big” intimidates some potential adopters. True, big data got its name from its volume, but that is not the only characterisation of data which makes it big. Big data is about persistent fast streaming data — velocity. It is also about diversity of type and structure (audio, logs, transactional data, text, video, social media) — variety.
Data must be analysed and decisions made in real time — volatility. If you do not act on an alert, you lose the chance forever.
Assuming big data is all about volume is wrong. Different solutions require different quantities, types and quality of data, and you can investigate whether your data is enough and suitable for your target project.
Big data analytics is not about storing large volumes of data. No. It is about getting value from data. If it creates a competitive advantage, enhances the corporate culture, destroys process silos, drives action to strategic objectives, and grows revenue and market share, then that data has value. Store what creates value. Do not hoard data, because storage does not come cheap.
That leads us to data ownership. You need no town all the data you use for their analytics.
You can buy or use free data collected by someone else. Non-governmental organisations, government agencies, industry associations, public enterprises, research institutes and many others come to mind.
And finally, the assumption haunting most companies in Zimbabwe is the lack of human resources. Analytics platforms have a steep learning curve irrespective of the vendor.
To groom the right resources for analytics might take ages, but that should not stop you. Find the right partner who can build your solutions to match your skills level and fully complement you.
I hope the windows, those analytics gateways, have been cleaned clear, and you can press the start button for your analytics expedition. Bon voyage!

Zorodzai Chenhidza is with OLSPS Zimbabwe (Pvt) Ltd, an advanced data analytics company with regional headquarters in Cape Town, South Africa, and business solution implementations in over 32 countries, including African countries. She is an engineer and a predictive data analytics consultant contactable on email: zchenhidza@olsps.com

Connect With Us

Fingaz Polls

CEO term limits...good or bad idea?