Analyse talent-flight to stop the brain drain
FUNDAMENTAL questions came to mind when reflecting on last Tuesday’s column “crying out for recent and relevant data”. And reports that the Fiji Bureau of Statistics data available to adequately analyse tourism and other sectors were out-of-date by up to four years, 2019 being the latest available. The analyses to be done were to support the formulation of “policies, address labour shortages, and steer the growth of the economy”. The data, it said, was inadequate and out-of-date.
DPM Prasad calls out Opposition Party
Why is it then, that Deputy PM and Minister of Finance, Professor Biman Prasad is able to make comments that exactly 29,719 left the country in 2023 in pursuit of education, migration, and employment opportunities? Calling out Opposition party members on their position on the so-called exodus of our people, he said they were “lies and misinformation.”
A stickler, it seems, for evidencebased pronouncements, the Hon Prasad apparently has access to data from which statistics was derived to confidently make that statement.
On reflection as mentioned above, that gives rise to fundamental questions. Is that same data available and accessible to other sectors? If so, why are other sectors, and the Opposition party, not able or willing to back their statements with databased evidence? If not, then why, in the name of transparency and freedom of information, do they not have access to the data to do their own analyses.
A starting point when considering data-based evidence for policy and decision making is to understand the value of detail data, not summarised, or aggregated, nor disaggregated by arbitrary segmentation. Whichever industry sector, or government department you’re in, the starting point for data analysis that provides for statistics, business intelligence, reports, and artificial intelligence (AI) - is data.
What is your data potential?
Three fundamental questions you need to be able to answer: What is your data potential? Are you in data deficit? Do you want to understand the potential of your data, and then analyse and monetise data?
From Hon Prasad’s statement on the number of people leaving the country, it may be reasonable to conclude that data is collected continuously and can be made available. We do have the data.
Ok. Let’s consider a few industries and look at scenarios and examples of data analyses possible with access to raw, detail data. We will look at Government, banking, and telecommunications and due to space restrictions leave manufacturing, retail, airlines, healthcare, and other industries and government agencies to another time, or direct questions to the feedback email signature below. First let’s understand that reporting from your current operational systems, directly from your operations data, whether you’re using business intelligence (BI) tools, operational system reports, CRM systems, sales systems, core banking, ERP and accounting packages is not what we’re referring to. Not to understand your true data potential. And the scenarios and examples below will demonstrate why.
Realising data potential through integration
Data potential is realised and maximised with the integration of data from your “Systems of Engagement” such as call/contact centre systems where you interact with the customer and “Systems of Record” such as accounting systems where you maintain transactional records. This is where you get maximum data leverage and have the opportunity to realise your data potential. Integration of systems data with website data and social/digital data yields maximum, compounded returns. But for the scenarios below we will keep it simple by sticking to the more traditional “Systems” for now.
Government
The Hon Professor Prasad: “When you look at the statistics, people are leaving all the time for education, migration, and employment.” He said on one hand, remittances were at an all-time high because of the movement of people, however it was a fact that we still needed labour here. Due to this we will continue investing in technical training through the three universities and other institutions.
These comments immediately point to the sources of data. Education data, immigration data, employment data. These data sources would be excellent candidates for integration. Remittances data would be another source of data you’d want to integrate into the first three. And perhaps data on available places at the universities in the coming years. Why integrate?
Here’s why. Whatever report or statistical system the Minister of Finance is relying on for making his comments is likely a report consolidated from multiple sources via spreadsheets or probably extracted from arrivals and departures data in which case the data would reflect what’s on a person’s passport and their declaration on the departures card.
Better decisions and policy making
And yes, with a certain degree of confidence you can report and make statements based on that. But to develop strategies and policies, integration of the three data sources will provide more insight when integrated with better information to education, and employment to make better policy decisions.
For effective decision making the questions that need to be answered will be complex to answer.
Questions such as: of all the people who departed our shores, how many doctors, nurses, accountants, how many teachers, hotel workers, how many labourers left; what was their education and certification levels, where did they receive their qualifications, what did they specialise in. What was their country of destination? Which, if any of our universities individually or collectively are offering places in those areas of specialization, are there sufficient places available and will there be sufficient numbers graduating to provide for Fiji’s needs given the rate of departure of that particular specialisation.
You can only answer that question, that complex question, with integrated data. Over a longer period of time, using historical data, we can add remittance data and ask the same complex question — attribute the response to remittances and assess which areas of specialisation we should prioritise for incentives to remain in the country, and which to allocate less budget to, based on evidence-based trade-off analysis.
Banking and finance
Systems of Engagement and Systems of Record (we’ll refer to them collectively as data marts) can provide answers to subject-specific questions. And with each new data mart, IT has to repeat its development efforts including sourcing data that already exists in another data mart. And that approach can become expensive and difficult to manipulate to answer complex business questions.
A Customer Management data mart can answer questions such as what products and services do customers with similar, profitable profiles have? Are particular regions or branches more successful at cross selling? Which mortgage customers do not have online banking?
The Risk Management data mart can answer what are the predictors of default? What is our exposure at default? What are the external risks and worst-case scenarios?
(An interesting challenge would be to ask those questions of your current systems and data marts. Or at least ask IT by when they could return you the answer sets)
And when you combine or integrated data from the two data marts, Customer and Risk Management, you can ask and get quick responses to high impact, high value complex questions like: which of our customers are vulnerable to market risk volatility? Which customers used up their credit line, got their limit raised, and paid it off successfully. Is there potential fraud, gaming the system?
Telecommunications
Just as with the Government and financial services and banking scenarios, the complexity and sophistication of questions change and increase in complexity as you add more data sources. The types of questions that can be answered with more data integration are more valuable. Combining or integrating data from data marts helps eliminate duplicate data and moves you toward more effective data re-use hence lowering costs and improving effectiveness.
Landline data marts focused on Network Route and call Path Management can help answer questions such as: what network segments are being most affected by busy-hour traffic patterns? Which specific inter-connection trunks are generating higher than normal originating traffic volumes? What is the costoptimal traffic routing scenario for busy-hour traffic volumes?
Access Cost Management data marts will help answer questions around what low-cost termination contracts are not being fulfilled each month? What is the average cost per minute of use, by carrier, by end office? Are particular end offices generating higher than normal access costs?
Combining or integrating the two data marts — route/path and access cost management enables much higher value questions such as: Which of our traffic termination agreements are vulnerable to price increases? How often do we exceed volume pricing on routes, and what does the extra capacity cost? Which customers are driving network usage, and is that usage sufficiently driving revenue to off-set termination costs? What usage plans should customers have, based on their network usage, to optimise margins?
Starting point and key to success
Whether you currently have data marts or not is not an issue because most government departments and businesses would have Systems of Engagement or Systems of Record, associated data marts, or at least spreadsheets and other data sources. You can benefit greatly by bringing data sources together for analysis that provides insights you’ve never had before.
Once you’ve considered what it means to integrate data, how your ability to answer more sophisticated, complex, and relevant business questions for high-impact decisions and immediate action you will have started, you will be able to confidently answer the question: “What is your data potential?”
Then comes the understanding of what level of data deficit you’re in and how to recover from that deficit while getting benefit from your data. And of course, the ROI, the analyses, the monetisation of data through decisions that enable increased revenues, lower costs, and provide great customer service.
One of the key ingredients of successful integration of data for analyses that helps realise your data potential is an integrated data model designed specifically for data analytics and data integration. This is where the key difference lies between reports and other systems and the ability to handle complexity in data analytics.
NALEEN NAGESHWAR is a data and digital strategy consultant. A Fijian citizen based in Sydney, he runs his own consulting practice Data4Digital and is managing partner Australia, NZ, and Pacific for AlphaZetta Data Science and Analytics Consulting. For feedback, email: naleen@ data4digital.com. The views are his and not of this newspaper.