Last week I presented at the first Swiss Business Intelligence Day. It was an inspiring conference to attend, with world-class keynote speakers opening the day. They included Professor Stephane Garelli from IMD, Philippe Nieuwbourg from Decideo and Hans Hultgren from Genesee Academy.
After such an illustrious start, you can imagine that I was more than a little nervous to present my very non-IT perspective of business intelligence. However, the presentation did seem to go down well, so I want to share with you some of the ideas I talked about. Not surprisingly, with my passion for customer centricity and always with the end-user in mind, I took quite a different perspective from that of the majority of IT experts who were present.
BI should Collaborate More
With the explosion of data sources and the continuous flow of information into a company, managing data will become a priority for everyone.
The Big Data market, which more than doubled last two years, is forecast to triple in the next four, according to Statista. BI will have to exp and its perspective, work with more varied sources of information and exp and its client base.
In the past BI was inward looking. It ran data-mining exercises, reviewed corporate performance, developed reports and occasionally dashboards. It was, and still is in many organisations, mostly concerned with operational efficiencies, cost-cutting and benchmarking.
The above plot is my own, simplified view of how BI fits into data management within most organisations today. The other three quadrants are:
- Competitive intelligence (CI) uses external competitor knowledge to support internal decision-making. Although BI is sometimes considered to be synonymous with CI because they both support decision-making, there are differences. BI uses technologies, processes, and applications to analyze mostly internal, structured data. CI gathers, analyzes and disseminates information with a topical focus on company competitors.
- Investor Relations (IR) uses internal data to get external people, such as shareholders, the media or the government, to support and protect the company and its views.
- Market Research (MR) on the other h and is mostly outward looking. It studies customers’ behaviours & attitudes, measures images & satisfaction, and tries to underst and feelings & opinions. That information is then used, primarily by marketing, to develop actions and communications for these same customers.
The four quadrants, even today, usually work in isolation, but that will have to change with this new data-rich environment in which we are working.
BI is Ripe for Change
According to a recent (Jan 2014) Forbes article, BI is at a tipping point. It will need to work in new ways because:
- it will be using both structured and unstructured data
- there will be a consolidation of suppliers
- the internet of things will send more and more information between both products and companies.
- thanks to technology, data scientists will spend more time on information management & less time on data preparation. At present it is estimated that they spend 80% of their time on data cleaning, integration and transformation, and only 20% on its analysis!
In February GigaOM echoed these thoughts, claiming that we are not in BI 2.0 but rather 4.0. They said the volume of data and the number of people now exposed to it, makes data availability to everyone essential. No longer does BI involve only the CEO and IT specialists, it concerns everybody.
Google glass, as tested by Virgin, is a good example of this. It delivers real-time, on time and relevant information to Virgin’s hosts and hostesses, to meet, greet and advise its passengers. Their customer support team can accompany their VIP guests and warn them of delays and gate changes as they happen. Google Glass enables them to get out from behind their desks and interact more with the guests they are trying to please.
BI must Deliver More Synthesised Knowledge
According to a recent Business Intelligence report on management’s opinion of their data, they are currently frustrated. They say that it comes from many disparate sources and is rarely if ever available in real-time. They can’t easily access it without the help of IT and it takes too long to customise it to what they need. What is particularly interesting in the findings, is that management were not saying that they don’t need information; in fact it actually looks as if they want to have access to more data. BUT more of it in a way that makes it easy to find what they want, when they want it.
Another finding from the survey shows executives’ thoughts about data delivery. Currently they are getting their information primarily through emails and spreadsheets. I find this shocking that today we still expect management to take the time to wade through all the data in order to draw their own conclusions. Less than one in eight of the C-suite is getting dashboards, which is their preferred medium (>><<). They also want mobile delivery so that they can access information on the go.
This study provides us with a simple plan to satisfy their needs and to help us meet our own challenges of data abundance. This is what we should prioritize, since we can no longer continue to do what we’ve always done in the same way we’ve always done it. The BI priorities are as simple as ABC; accessibility, business impact and consistency (>><<).
BI needs to Provide Simplified Access
Information should be provided where and when it is needed and in such a way as to have most impact on the business. This means making it easy to review, and quick and simple to draw conclusions. This is why the number one dem and from business is dashboards.
Dashboards have the advantage of imposing consistency (>><<) so no time is lost in underst anding what the information is showing. With the availability of more information, comes the challenge to make it available to more people. And more people will also mean more and different needs.
To underst and the accessibility challenge I find the tree is a great metaphor for what we struggle to achieve. The roots can be compared to all the different sources of information we have at our disposal. The trunk is like all the integrated information that is reported in dashboards and the branches, twigs and leaves are the different data warehouses we create.
Whilst a one-page overview is sufficient for management, others will need greater granularity. Therefore we need to make information available at different levels of detail. My experience suggests three types of information sharing.
- The leaves are like data warehouses where the raw or nearly raw data sits
- The twigs are the information repositories where analysed data and information resides
- And the branches are the knowledge libraries where the integrated actionable insights sit
What I have learned from setting up numerous data warehouses, information repositories and knowledge libraries, is that it is not easy. Not because of any technical complexity, but because of winning the needed internal support for the project and getting the essential acceptance for global access to the information. It takes more than technology, it takes a culture change in many cases too, and this is the real challenge. Stopping the “information is power” mentality means finding ways to counter the opposition who claim confidentiality of their own data whilst also requesting access to everyone else’s. In addition, even if people need information, they will generally not make the effort to go looking for it, if there is an easier way, such as by asking someone else! All these issues need to be resolved for an integrated database project to succeed.
One way to encourage the culture change mentioned earlier, is to demonstrate the business impact of what you are providing. The desired impact won’t come by delivering spreadsheets, it will come from dashboards (>><<).
So how do you summarise a company in a one-page dashboard, especially those which are present in multiple categories, globally? Well, often the simplest way is not to try to cover the total business, but rather the top categories and markets that would cover 70% – 80% of total sales. In most cases this would be sufficient to underst and the main priorities for management.
Of course at category level each business unit should be able to get access to more detailed information, as should the regional presidents, if you are working in such a complex business environment.
The real power of dashboard information will come from data integration, where both internal and external information are synthesised, for a holistic view of the business. I have worked on several projects that combined internal information with consumer data for a complete business report. The consumer information came from promotions, call centres and CRM activities, and was combined with market research on product and communications performance, to provide a solid base of consumer underst anding. This can then be presented alongside the more usual financial information that executives are already receiving. Having a complete overview of the business has far more impact than individual, silo’d summaries and enables management to make decisions more quickly and easily.
Another challenge when setting up and integrating databases, is in the harmonisation of their master data. When you are working with consumer data, this challenge can be multiplied by ten if not one hundred. For example, consumers will talk about a pizza, without specifying the br and, sub-br and, variant, flavour, packaging and size that would be used by the business to define it. So you have to find a way to translate what the consumer is saying, into the products as recorded internally.
The consistency of the master data will even increase in importance and complexity, with the expansion in available data sources. In addition, the fact that more people will get involved, will confound things even more, since their needs will differ.
Asking Better Questions of the Data
Accessibility, business impact and consistency are vital to the success of the new BI’s data management and usage, but I feel the urge to add one more thing. That of asking the right questions of the data. Although BI is used to asking questions, I think Market Research (MR) are the real experts in questioning. Therefore they should be involved in ensuring integrated databases are combined in such a way as to permit easy extraction of whatever level of information is required, or whatever perspective might be taken.
For example, BI is used to running forecasts. Those usually start from a review of past data and current reality to develop forecasts based on complex algorithms. They will do this within their teams with perhaps input from finance. MR on the other h and, is more likely to work from societal trends and develop plausible future scenarios, brainstorming across the organisation to gather a wide array of perspectives. Both perspectives are complementary and combined, they make a powerfully readied organisation.
Making more data more accessible to more people will certainly help this question development, as I think getting the right answers depends upon asking the right question, don’t you?
These were just a few of the ideas I shared at the Swiss BI Day in Geneva. How do you see business intelligence adapting and changing as a result of the increased information availability happening today?
C³Centricity used images and graphs from Statista, Microsoft and Virgin in this post.