Business Intelligence became mainstream since 2010s. The boom of the start-ups, the apparition of the smartphones and the tech culture pushed societies accross the globe toward digitalization. In parallel, software companies developping tools allowing this data to be visualized and create reports did a great job to democratize access to numbers. So you might wonder: what is exactly "Business Intelligence"?
Definition of business intelligence (BI)
Business Intelligence gathers all the data analysis and visualization techniques, allowing business people to read and understand the activity and to make decisions based on the numbers provided. BI covers topics such as reporting dashboards and reports being sent around accross the organization, but also all the tools making it possible to carry out deep-dives and perform self-service analysis such as Qlikview, Tableau Software or PowerBI or ESSBASE (Excel add-on).
3 main components to business intelligence
It's a small piece of all the data-related activities and jobs - but a critical one for business people. Business Intelligence is about:
- Owning the KPIs and business metrics definition to ensure everyone is getting one and only one definition and not to confuse leaders. This means owning data dictionnaries, documentation and so on. It's a tedious piece of business intelligence, but most probably the most critical one.
- Structuring the right information to make business decision by building special data tables named datamarts which are going to be helpful for decision-making. Not all the data is relevant and the level of granularity required is to be agreed with the business: if too high level, then numbers won't deliver any insight and too granular, they will confuse people.
- Presenting the information to business people by building reports and dashboards and ensuring that every exec in the company is having the right report to steer the business and achieve their business goal.
Owning the KPIs definition
Define the truth
Revenue, turnover... all these metrics seem to be quite common and understood, right?
Actually, when it comes to business, each company will look at these components slightly differently. Some metrics are pretty straightforward to define. For an hotel, that would be the number of nights occupied. However, other metrics trigger more questions: for a shoe company, should we look at the gross sells or net sells, post customer returns? Both metrics reflect the same reality but have different stories to tell. You might want both: but it's up to BI to clarify the differences and make this explicit and understood.
From business to code
These are tipically question the business need to answer with words. Now, for Business Intelligence, it comes to define them also in terms of technical definition: where is being stored this data? What should we factor in or remove as filter to get it correct? Which SQL queries should be considered as source of truth ? Business intelligence put into code what the business tells what's worth analyzing.
Datainsightout helps small and medium companies crunching their data. What about yours?
Speak the same language
Business Intelligence plays a critical role to align the teams and to read correctly the metrics and numbers which are the bones and flesh of the organization. Finance department is generally a strong stakeholder. There should be no disconnect or multiple definitions of the same metric. Finance department is a key sponsor for the BI department. Typically, financial performance reporting, including actuals, budget and latest forecast should also be managed by business intelligence.
Business Intelligence structure the information
Once all metrics are defined, it comes to store them properly: which are the dedicated tables which are mandatory that business intelligence need to build? Usually, all numbers are stored into production tables: these tables used to do run the business and support the day-to-day activity. They are not meant to be used for analysis. Bugs can happen, information is very granular, IDs are everywhere, information is spread accross multiple servers... not easy to use at all.
Business intelligence needs to pull out the relevant information from these production tables, transform them, ensure they are as clean as possible, align them with the business definition and store them at the right level of aggregation in new tables named datamarts which are going to be shared with business decision-makers.
Dimension aggregation should be also defined carefully. The business might not want to get the full details. Aggregated tables might be enough to answer 99% of the questions. Datamarts must answer straightforward business questions. It's a balance between keeping the size low and getting the most relevant information.
Datamarts for marketing, HR, Sales...
You could have dedicated datamarts for sales people, for marketing people, HR people or project owner people.
Marketing people will want to get a marketing datamart at a customer ID level. They will analyze the journey of each customer. Marketing will understand how a customer signs-up, how he activates and get demographic information to build a segmentation.
Sales will want a breakdown of all the sales, details around each portfolio, sales rep, pipeline forecast or any customer care information. Finance will want another view. Each departement will want to have a dedicated database with dedicated datamarts, allowing them to be a data-driven company.
Business Intelligence builds dashboards
Once this very tedious piece of work is achieved, it's up to business intelligence to build the reports and other data visualizations which will be consumed by the business.
Many options to build reports
The panel of tools to build reports on the market is very broad. Nowdays, most of the solutions are the same. Large companies will frequently use multiple reporting software at the same time. It all comes down to pricing and number of users. Some dashboard solutions are 100% in the cloud, others require more technical skills, others are easy to use via heavy use of drag & drops. It all depends.
Dashboards for all departments
All projects and business sides should be covered. Business intelligence could and should collaborate with all the departements. C-level executives should have daily high level reports. Purchasing should know about the evolution of the stocks. Marketing should get performance related information regarding their latest campaigns. Sales should get reportings on quota and pipeline.
Business intelligence is the department to create if you want to buid a data-driven company. Reports could be created to cover all the activities of a company. The right level of reporting should be created: not to detailled but not to high-level. Experts in BI are highly valuated and seeked on the market. BI can transform your company.
Having good business intelligence reports is nothing without a great data storytelling.
Alertings and real-time reports are also handled by business intelligence - but it might require a particular data infrastructure set-up. A production line breaks? A customer drops-off? Volume goes down? BI has the tool to monitor this and push alerts. Company executives can have real-time control on their perimeter. Money could be saved thanks to business intelligence.