We live in an era of unprecedented skills shortages. At no point in SA's history has there been such a chronic dearth of quality skills, and with emigration on fast forward, don't expect the situation to improve any time soon.
Nowhere is this situation more starkly apparent than in the business intelligence (BI) market. There are a number of reasons for this:
* BI is a growth market - it is surging far ahead of the rest of the market, and in such a case, demand for skills will always exceed supply.
* BI is inherently complex and fragmented. Specialist skills are required at every stage of a full-blown BI project: data warehouse designer, data mart or OLAP cube designer, ETL (extraction, transform and load) specialist, front-end tool designer, report developer, and database administrator.
It's a tall order - almost an impossible one to fulfil given the skills shortage - and this list of required skills helps explain why few users are taking advantage of the potential of BI: corporate adoption is stuck at around 7%.
It also helps explain why there is such a backlog of jobs in all corporate BI environments: the available bodies are simply stretched to breaking point.
That's the skills required at the back-end: what of the skills required at the front-end? Despite all the promise of BI, users really do still need a relatively high level of skills if they want to ask the right questions of the system.
However, few business users have the time or inclination to grasp and embrace the complexity of SQL instructions, or to construct complex questions.
What they do know is how to ask questions in natural language English, and this is emerging as the best way in which to deliver BI to users.
It's similar to the way in which people interrogate Google: no one needs any training at all when asking Google, or other search engines, to find information - and the results are returned almost instantaneously.
The strength of data search is that it provides a familiar interface and metaphors, while applying natural language processing techniques.
Corey Springett is strategic business manager at Progress Software South Africa.
The answer is a concept known as data search and discovery. The strength of data search is that it provides a familiar interface and metaphors, while applying natural language processing techniques. The interface is the search box: users type in the words that describe their interest. They are presented with data and navigation choices, which allow them to iteratively filter out the data they don't want. The filtering logic will probably be the Boolean logic we use with Google and Yahoo.
Natural language processing can be applied to users' requests. There are five essential drivers for companies to adopt data search and discovery:
1. Immediacy: Users need answers as soon as possible.
2. Investigation: Users don't necessarily know what questions to ask.
3. Investment: Budget is tight, and it costs a great deal of money to develop formal reports. Some queries are not of a sufficiently high priority to justify developer time, so they are ideal for data search and discovery.
4. Inaccessibility: Users need to be able to find information in an existing report, and traditional BI applications will not necessarily be able to find them.
5. IT efficiency: IT users can boost productivity by producing reports, finding answers or helping end-users with their information requests.
Data search and discovery is ideal for organisations which have a reporting backlog which could be addressed by data search, either via IT or users; which find decision-makers and managers lacking the information they need; and which operate in fluid, dynamic and frequently changing circumstances, and accordingly need flexible reporting tools.
For many people and companies, this approach will solve the bulk of their operational BI requirements.
* Corey Springett is strategic business manager at Progress Software South Africa.
Share