- What Is Business Intelligence
- What Does A Business Intelligence Analyst Do?
- What Skills Should The BIA Bring To The Organisation?
- Where Should A BIA Sit In The Organisation?
- What Are The Risks Of Under-Employing A BIA?
- Where To Next For The BIA?
The business world has never been more reliant on business intelligence capability. The rise in data analyst employment in the US and the recent call in Australia to ‘teach coding to kids in schools’, suggests an understanding that a future workforce will need to tap into the magic of data and make it meaningful. One challenge for business leaders, who will continue to need to harness internal information for performance analysis, is the increasing need to also test; ‘what next?’, ‘what if?’ and ‘what’s out there?’ scenarios. As global online business and consumer-led markets in goods and services grow, relevant external information and trend analysis are key to ongoing business competitiveness. The challenge for all business is that disruption, including machine learning, artificial intelligence and Big Data analytics capabilities, is the new norm. Businesses can expect the need to access and apply intelligence in the next five years to continue to expand. Fortunately, new online tools and specialist BI start-ups are filling the current capacity and capability gap in data analytics. This article aims to support the business manager’s understanding of the business intelligence analyst role to enable their effective input within the organisation. If you are looking to establish a role for a business intelligence analyst, this article also describes some of the in-house reasons for appointing a BIA and the expectations about the breadth of skills that may be required to best address the business need. For those looking to contribute to a business in a business intelligence analyst role, this article will provide some insights into where the role intersects in the business environment and what capabilities you can bring to the role. As a key enabler, the purpose of the role will be to support business competitiveness while identifying new opportunities for the business to move forward.
What Is A Business Intelligence Analyst?
Business Intelligence (BI) can be described as the sets of information provided through data analysis and knowledge management, which can inform decision makers on areas for response. This can be in relation to emerging external trends or changing internal performance demands. A simple way to consider BI is to look at whether the data that informs business development goals are drawn from internal systems or external sources. Internal information in large organisations will generally come from business wide enterprise systems, such as SAP or ORACLE database systems, or can be drawn from a range of more distributed, and often informal, data capture systems from across the organisation. This data can then be analysed for BAU reporting and tested against other information to identify how the organisation is performing against expectations. External sources may be directly business driven, such as market surveys which provides information on customer expectations or a more formal analysis to identify what trends are ‘out there’ to which the business may need to respond. The external information would also provide information on the competitive profile of like businesses. In responding to external change, businesses can increasingly draw on external ‘self-service’ online analytical resources. In marketing, the capture of social media trends via Google Analytics, Kissmetrics or similar analytical tools is now a familiar activity, but making sense of the array of information may not always be straightforward. In describing the BI framework in an organisation, Chris describes his view with the onion analogy.
“If you take Business Development as the primary purpose of BI, then peel back that layer you will have BI roles which are either internally or externally focused. The next layer is processes, data and people. The external branch has a focus on competitive intelligence while the internal is focused on business intelligence and performance"
As with any component of a business, the wider Business Intelligence System needs to deliver value and the cost of adopting and adapting large enterprise data systems is being challenged by lower cost and more agile cloud-based and web services options. In many large organisations, the two forms of external and internal information intelligence demand are merging through the Data Lake concept. This is taking over from formal Data Warehouse designs, where inputs are typically highly structured and outputs are packaged for a level of standard reporting. The Data Lake means data can come from multiple sources and be ‘pooled’ in a less structured form which can be ‘tapped’ as needed with new analytical tools or programmed queries. Open-source packages, such as Apache’s Hadoop, which is used by large online services such as Amazon, Twitter and eBay, is an example of a programming tool which can be deployed to pull data from a suite of data storage clusters. This information is used to process data as well as address new business questions. Martin Fowler simply describes the Data Lake as a store for raw data, in whatever form the data source provides. There are no assumptions about the schema of the data, each data source can use whatever schema it likes. It's up to the consumers of that data to make sense of that data for their own purposes.
Diagram based on work by Martin Fowler (2015)
As shown in the table developed by Tamara Dull Director of Emerging Technologies, SAS Best Practices below, the growth of the Data Lake concept is also driving the demand for data scientist skills to enable business intelligence analysts over other business professionals. While Martin suggests in his diagram that ‘we’ select data for each need, the reality is that selecting, extracting and analysing the ‘right’ data from the Data Lake is still the realm of a specialist data scientist or analyst. Table sourced from Dull (2015) While the ‘one-size-fits-all data system’ prevalent in the enterprise suite of large organisations is changing rapidly to more agile approaches, challenges remain around data security and statistical validity for the Data Lake adopters. The structured data form is still likely to hold as essential for some time fobusiness-criticalal areas.
What Does A Business Intelligence Analyst Do?
Just as information management and information systems are becoming less rigid, the Business Intelligence Analyst role needs a level of agility to inform as well as respond to changing business needs. The principle role of the Business Intelligence Analyst within the overall Business Intelligence System of an organisation is to provide a vertical bridge through the business to communicate information of high value to support decision making demands. If the BIA’s role takes a ‘business insights’ approach and the analyst is enabled to work with decision makers, information managers and external service providers to continually engage in business development, they are more likely to deliver value. Often the first step for an incoming BIA is orientation and understanding the business current state. As Chris explains, “The BI has to be able to consider the organisation’s existing systems as well as the culture of the business. They will need to understand ‘What are the business development objectives? What systems and mechanisms for data capture, storage and processing are being applied? What methods are being used for analysis, ie how automated are the systems? How creative or mechanical are the requirements?’” The Business Intelligence Analyst may span different parts of the organisation, often working from either a more internal or external point of view. In large organisations, the BIA would be expected to bring a level of knowledge of Big Data handling and could be expected to inform future intelligence gathering priorities. As a first step the BIA may be expected to work across the organisation to provide a stock-take position around the “4 ‘V’s of Big Data”, a term coined by IBM to provide a framework which can inform business management: Volume – How much data is currently managed / out there? Variety – What types of data are managed / can be accessed? Veracity – How reliable / useful is the data? Velocity – How accessible / usable is the data within the information system To these I would add a fifth: Visualisation, and others have also added Value to this list, which we’ll touch on next.
What Skills Should The BIA Bring To The Organisation?
While the types of skills required will be highly dependent on the size and scale of the organisation, the BIA can be expected to add value from the outset through their technical, analytic and problem solving skills. They will often come with backgrounds and qualifications in Information Technology, Data Science and Computer Engineering. Despite these backgrounds, the role needs to be clearly distinct from the IT or Engineering specialist who may have an expectation of undertaking analysis to construct a business technology solution for the organisation. Rather than delivering specific technical solutions, BIAs bring an understanding of data gathering and analytics to intelligence and knowledge forming. In any organisation this can be a broad activity which can cover:
Running complex reports from existing databases
Organic activity where special projects are established to analyse information on emerging markets
Analysis of market testing to support new product and services development.
For business marketing oriented BIA roles, an organisation would expect knowledge and experience across a broad range of social media platforms and media analytical tools. For these roles a double degree in information technology as well as communication and media qualifications would be an advantage. However, if the organisation anticipates expanding the role in the future, or needs a BIA ready to step up to a senior or management level role, the job description will usually need to describe expectations for broader business knowledge, communication expertise and potentially relevant sector experience. Successful candidates would typically be expected to have postgraduate qualifications, such as an MBA, or broader relevant business experience. Chris has a clear view on this; “BIA roles can often be constrained if the organisation sees the main skill as simply using data to produce statistical information. This skill set is drawn from people with computer science, data science and statistics expertise and often weighted toward technical or industry settings. While these skills are highly valuable, people selected with a focus on these skills alone will often have a very weak role in influencing and informing decision makers.” In the US, according to online data available from Indeed.com, the demand for data scientists and analysts with key skills, fluctuate during the year and track closely with the number of job vacancies for programming skills in R (for example). This may may indicate reactive or short-term recruiting demands for these skills within the short-term annual business cycle. In my experience, while much emphasis is often placed on the value of data analytics, an effective analyst at management level will be able to look at data in the context of business development. This will consider external competitive or cooperative opportunities, in-house and key stakeholder inputs as well as considering existing corporate knowledge. Data from Indeed.com for the US jobs market for 2014 - 2016 shows that while the term 'Data Analyst' appears in around 2% of all job postings, ‘Business Intelligence’ and ‘Data Scientist’ appear in fewer postings. In comparison postings which refer to 'Business Development' made up between 14% to 16% of the total jobs posted in the same period. This suggests to both employers and job seekers that the value add by Business Intelligence Analysts into the high level of Business Development demanded should be managed effectively to enable the alignment of these capabilities.
Where Should A BIA Sit In The Organisation?
In larger organisations, the BIA may be a member of a business development team working closely with other specialists who can collectively add value to informing strategic planning and business development. The type of agency given to the BIA to be a ‘transboundary’ agent – that is to be able to liaise through and across the organisation and across business functions - can be critical to the effectiveness of the role. Ideally, this equips the BIA with the ability to have an understanding of and be able to access a range of structured as well as unstructured information in the organisation. As a data translator ultimately feeding information into the ‘Decision Support System’ (DSS) of an organisation, the BIA role needs to go beyond reporting the status quo. The BIA role sources appropriate data, applies a suitable analytical method or methods, interprets the resulting information and then constructs 'a rich picture' in the analysis. The resulting decision support information may include; a description of context, identification of trends, highlights of emerging associations or changes within or between organisations (competitors), which could demand repositioning or business review.. Chris’s view concurs “Although BIAs have technical skills, they don’t belong in the IT Systems arm of the business. And although a BIA wouldn’t necessarily know a great deal about the layer above it, ie generally won’t be across benchmarking or profitability compared with using data to produce statistical information, they must be within the Strategic and Business Development Team of the organisation.”
What Are The Risks Of Under-Employing A BIA?
A major challenge for any organisation is the amount of work and skill required just to keep the business as usual (BAU) processes and reporting demands running. With a skilled and capable BIA on hand, there is a risk that the role is absorbed with BAU analytics and reporting and dealing with data wrangling from an existing system. This type of work constantly deals with the ‘known knowns’ of the organisation rather than venturing into the unexplored area of data discovery and over time will place a limit on business development. The size, maturity and risk appetite (or aversion) of an organisation also plays a part here. As Chris also reminded me “Finding the gaps in information as well as asking ‘what isn’t in the current data or information set?’ is an important value add offered by a BIA.” The Donald Rumsfeld conundrum is a useful test for both a manager looking to appoint a BIA as well as an initial approach for a newly appointed BIA to explore information knowledge within an organisation. Rumsfeld appeared to ramble at a press conference at one point during the Iraq War saying; “We know what we know and we know what we don’t know but we don’t know what we don’t know.” Although Rumsfeld possibly didn’t communicate this point as well as his advisors had intended at the time, a close colleague of mine, Dr Andrew Patterson, adapted this approach in a risk management toolkit to look at disease import risks in the UK.
If an organisation only deals with ‘what we know we know’ – or ‘what we don’t know we know’ – that is, ‘we know the answer is there in the data and we have a report on it already or we just need to adapt a report to fish out the data we have’ - then the business will not really be using the BIA as effectively as possible to push business development. Vincent Granville, writing in Data Science Central, describes this boundary pusher as a ‘Horizontal Analyst’. Granville proposes that while the vertical analyst may typically take on the BAU role, the horizontal analyst will pursue approaches which have greater potential to contribute to a return on investment. To effectively utilise the BIA, as Granville suggests, the business has to carve enough of the role out for the BIA to ask: ‘what don’t we know, that we don’t know?’ This type of intelligence analytics often goes beyond the organisation and ideally sits within a broader strategic outlook activity. For me another barrier to effectively utilizing the BIA is my fifth V of Big Data analytics – Visualisation. New online data-analytics tools, such as Tableau, are enabling visualisation of large datasets and analyses of multiple data sets as never before. However when comparing notes with Chris we have both experienced the challenge of translating data into a visual format a decision maker can reasonably interpret. As a GIS specialist working to develop disease risk information in the very early days of digital mapping in government, I developed a risk assessment report which included a relevant map clearly showing an encroaching disease risk within Europe which posed a national threat – or so I thought. While my manager liked the report (‘the rich story’), he discounted the map saying he didn’t think it was helpful. After some further discussion, my manager had a light-bulb moment and could see what the map was saying. From that point on, the risk assessment brief always included a map, but I was reminded never to simply assume graphical literacy. Chris describes a similar challenge: “Even if the decision maker has the capacity to digest and understand how things got there (in the analytical report) – what are the implications? Do they have the business nous or the resources to do something about it? In my experience, it is rare for leaders to be able to take a helicopter view and understand the data capture through to interpretation and action. This means the BIA must always be looking at how to best tailor the information to the viewer and a single type of data representation may not be adequate.” From our discussion a simple outline of the iterative nature of the BIA role in supporting decision-making process could look like this:
Where To Next For The BIA?
For organisations demanding expertise from a BIA, a set of challenges to consider in a new appointment or role review are:
is the job specification forward thinking enough?
can the BIA also inform on emerging tools and online services to streamline data analytics demands relevant to business development?
what are the current/near future demands on decision making?
what is the business growth in data capture plus storage capability and potential to utilise analytics from a range of platforms?
In his book on Disruptive Analytics, Thomas Dinsmore describes how new online services, such as Tableau which is being adopted at an extraordinary rate to enable desktop analysis, marks a substantial change in organisational analytics. Dinsmore remarks, “we are now in the era where conventional data warehousing theory no longer applies”. Another trend, (which perhaps explains the flat statistics in the jobs analysis for BIAs in the US), is that BIA specialists are now moving rapidly into the startup community. Angel.co currently lists 1073 Business Intelligence Startups in the US with an average value of $US4.3m. The writers in VentureBeat magazine recently stated that “the new funding (for BI Startups) comes during a mad storm of business-intelligence funding deals, acquisitions, and product releases. Most recently, consider Hitachi Data Systems’ purchase of Pentaho, Salesforce’s new Wave cloud service for analytics, and GoodData’s $25.7 million round.” One start up profiled in the article, Looker, raised $30mill in 2015 for developing business-intelligence software which people can use to pull up visualisations of data sitting in corporate databases. This shows a race for new analytics among the big data players which could signal a vibrant future for people equipped with BIA expertise. For organisations moving from BAU processes to embrace predictive modelling, optimization techniques and implementation of machine learning, the BIA is moving into a role that needs to be at the leading edge of organisational response.
About Chris Ong, a key contributor to this article:
Chris started his career as a chemical engineer with Brown & Root (now KBR) and BHP Billiton over various roles. Now at The University of Newcastle (Australia) as Business Intelligence and Development Manager, Chris works to glean what makes consumers tick and aims for his insights to influence successful outcomes for the University's strategic direction.
Other Resources and Further Reading:
Khan, Rafi Ahmad; Quadri, S M K (2014) Business Intelligence: An Integrated Approach, International Journal of Management and Innovation; Taipei
Dinsmore, T. (2016), Disruptive Analytics: Charting Your Strategy for Next-Generation Business Analytics Chapter Self-Service Analytics 199-230 Springer