We’ve been talking a lot recently about companies needing to use their data in order to stay in business in the future. We’ve even gone as far as saying that every company is a data company, whether they know it or not. And every business – regardless of the industry, product, or service – should have a data analytics tool driving their business. With that being said, it’s not enough to just have a tool. Every business needs a business intelligence strategy to take it forward.
As the Global Team Lead of BI Consultants at Sisense, I can say that the projects I’ve worked on where a BI strategy was involved, were more successful than projects without a strategy. The BI strategy played a major role in the setup, execution, and ongoing implementation of the BI platform. And it can do the same for you.
But what is a BI strategy in today’s world?
A business intelligence strategy is a blueprint that enables businesses to measure their performance, find competitive advantages, and use data mining and statistics to steer the business towards success.
Every company has been generating data for a while now. The question is, what are you doing with it? Here are the five steps you should follow when building a data strategy.
The modern platform for enterprise business intelligence:Get the free report
#1 Find out where you are (so you’ll know where to go)
Our go-to approach for analytics that feeds well into a BI strategy is the Evolution of Analytics chart (below). Originating with Gartner, this chart includes the analytic features needed for a full analytics strategy, and what our AI team believe to be the absolute future of analytics – Cognitive Analytics.
In order to know where to go, you must first find yourself on this chart. Most companies find themselves in the bottom left corner, in the Descriptive Analytics and Diagnostic Analytics sections. You likely already have some form of scheduled reports, are drilling down into your data, discovering what is in your data, and may even be visualizing to some extent. But as you can see from this chart, there is so much more you can be doing … and it’s easy to get there with the right business intelligence strategy.
#2 Plan your objectives (and map the supporting data)
At this stage, you will need to plan your business goal. It may be an oxymoron, but with so much data out there, the more focused you are in planning your business goals, the better off you’re going to be. Essentially, your data needs to address a business question or a business need. Do you want to be more efficient? Find a bottleneck in R&D? Share knowledge with customers? Add value to your solution?
With a goal of getting to the end of the chart with predictive and prescriptive analytics, you can ask questions like: Are we going to hit our targets by the end of the year? What is the market segment we should focus on? Is there a bundle of products we should suggest based on our historical performance? What are the main contributors to close a deal? And is there a set of different elements we should combine and/or follow to make a bigger impact in the market?
When working with customers we’ve found that a good place to start is with finance and sales data. Uncovering insights in this key area can make a major impact on the growth of the company. Let’s take revenue growth, for example. In order to analyze revenue growth, you will first need all of the sales information related to revenue. This information may come from Salesforce, or from your ERP system like Oracle, as well as from any other marketing technology that may hold customer experience information.
Bigger companies have complicated tech stacks with multiple touchpoints that are collecting piles and piles of customer data that needs to be evaluated to see which technologies are relevant to this specific objective. Again, the granular focus on your objective here will help define which datasets you need for this objective.
Also, keep in mind which types of data are missing as that may be critical in putting together the bigger picture and may prevent you from reaching the predictive analytics stage and the future of your BI strategy.
The end result of this step should be a set of KPIs that support your objective, and a clear map of the data sources you will need to analyze and visualize these KPIs.
#3 Define how the data will be shared (and how it will be distributed)
Before you start executing on the plan, there is one very important question that needs to be answered.
How are you going to release the solution?
I know, this seems like an obvious question, but when so much data is concerned, this is one of the most critical items to be defined in your strategy. There are two basic strategies that your company can take:
A Decentralized Approach
Some organizations empower its end users with interactive dashboards. This is where the term citizen data scientist comes into play. Citizen data scientists do not need to be data scientists, they just need to know their way around the data, and that begins with giving them access to more than just a dashboard. These are the employees you’d consider power users. They’re the ones that are defining what is needed in the department and need the insights to make decisions. In large enterprise organizations, this scenario can free up bottlenecks with IT and data teams, enabling departments to do a lot of the data analysis on their own.
A Centralized Approach
On the flip side, we have organizations that have decided to be in complete control of the data that is distributed, including who sees what, and how much of it they can touch. Take Nasdaq for instance. Their BI strategy took into consideration their sensitive data, huge distribution channels, and the need for better governance to reach one version of the truth. Building on this strategy, Nasdaq provides its customers with dashboards, but it does not provide them with the ability to work directly on the data models. There is a certain amount of drill-down that each customer can perform to see more details, but the data is strictly governed with system level, object level, data level, and row-level security.
If your solution is going to be embedded within your product, there are a few more questions you should be discussing, how and where to embed, how to scale when you’re going to hit the critical mass number of users, what will be the process of moving from development to product, and are you going to freeze the code during that time?
Also consider how the BI solution will be communicated to the company or customer, and how you will support the adoption of BI in the organization. Options include emails, meetings, PR, training, enablement, maybe even old-fashioned documentation.
Spend some time on this with your team members, stakeholders, and management and go over all the scenarios. As your BI strategy materializes, you will be thankful that you did.
#4 Deliver your solution (and focus your efforts)
You may think it’s too early to think about, but part of your BI strategy needs to be how you’ll deliver the solution to your end-users, or your entire company.
Some of this may come naturally with the decision on how you will share data. For instance, for a centralized approach (where the user has no room for adjustments) you will need to put more effort into training and documentation. For a decentralized solution, you may want to do short release cycles, asking for feedback on each release, and then incorporate that feedback into the next release.
By outlining the delivery method in your BI strategy, you can plan how you want to focus your efforts. Do you need to work on creating more data governance or put more effort on training and documentation?
#5 Find the roadblocks (and push through them)
Now that you have a clear idea of what kind of questions you need to ask, and what kind of data you need to support that, you will now need to understand who are the gatekeepers in this scenario. The gatekeeper in this situation is basically who or what is standing in the way between you and the data.
Your data can be stored in a database or may even be located with a third party vendor. If you are outsourcing HR services to a company that is managing your hiring pipeline, or if you have cloud-based service providers like Marketo for managing marketing campaigns or Quickbooks for financial services, then you will need to plan how to connect to their data and learn about its structure in order to use it properly.
For some third-party applications, there may be a simple API that can be used to import the data. For the IT department, it may be just the manager’s permission and a couple of signatures that you will need to find in order to access the data. This all takes time and resources that need to be allocated inside your BI strategy.
Another roadblock may be the personnel you need to implement your strategy. Don’t forget to plan your workforce ahead of time, not only for the first release but for future changes and enhancements down the road.
Go Big, go data
We are firm believers in a data and BI strategy. We’ve been with customers when the light bulb goes off when all the pieces fall into place when their first dashboard delivers that ‘Aha! moment’ that changes the shape of their business forever. It’s moments like these that make me proud to be a BI consultant helping customers build their data strategy and find insights in their data.
The digital transformation of your business must happen – regardless of the industry, product, or service – you must develop a digital transformation strategy that will drive your business forward.
The modern platform for enterprise business intelligence:Get the free report