5 tips for an effective business intelligence strategy

As Business Intelligence (BI) Consultants, we are commonly tasked with inheriting a murky business intelligence strategy that involves cumbersome BI tools, data quality issues, and report modifications. We look at a given report and we find that it’s been built in a vacuum and – after studying the underlying tactics – we wish someone took a little more time up front to design and implement a more foundational business intelligence strategy. Uncovering disjointed data sources, irrelevant KPIs or reports that don’t roll up right can make a bad experience worse.

Here are 5 powerful tips that you can use now for your business intelligence strategy to make sure you don’t pass the buck to the next guy on similar projects.

1) Identify the right KPIs

First thing’s first – this process must start with the business users, not with IT. After all, IT doesn’t make the data. Business users make the data and, therefore, they should own it. The IT team is there to work alongside and guide the business team in order to answer their most burning questions.

Now with that said, on to Key Performance Indicators (KPIs). I believe that KPIs should be broken into perhaps two primary buckets:

  • Representative – These are your dollars and sense reports. These are strategic like year over year sales, sales actual vs. target, Profit Margin, etc.
  • Operational – These come in many forms that are usually industry specific like percentage of claims paid, the number of rejected parts, anything having to do with the operational efficiency of your business.

2) Add in protocols to maintain data integrity

The currency of the information age (data) can best be used when it follows some basic principles. Chances are your business already has some form of processing and collecting data. However, that data isn’t always cleaned, scrubbed and sometimes maintained to the degree that we need for it to be effective.

Integrity – When your data has integrity, you will avoid situations where, for example, you have multiple names in multiple systems that effectively mean the same thing. This can have huge impacts on the quality of your analysis. For example:

  1. In your Data Warehouse, you have a value called ACME Company
  2. In your OLTP System, it’s strictly ACMECompany
  3. But to your users, you know them as ACME-Company

Governance – The art (yes, it’s an art) of maintaining and sourcing data – with business users owning the data – creates a well-oiled machine in terms of Business Intelligence Strategy. When IT owns the infrastructure, it frees up resources that would normally be dedicated to writing business requirements to create better models, and make better design decisions. As business analysts, start asking yourself questions as to why data needs to be there, and who owns it. This aligns the development process with key stakeholders and performance indicators.

As IT folks, we know what “ask the business” usually means and we generally like to avoid the political discourse coming from the top with most projects. So, keep in mind there will be a level of integration amongst your teams that will demand collaboration on multiple levels. It’s only fair that if we want to be successful with any strategy that we do not let obstacles hinder our ability to execute.

3) Data Warehousing

This is your IT Department’s time to shine. The infrastructure and framework drives data insights and should be built on the standards that were agreed on in your data governance policy.

The goal is to store, retrieve, and cleanse data for your analysis. Begin by determining where your data sources reside. Ask yourself and your team some of the following questions:

  • What technical stack should we use to extract the data?
  • Where are our technical proficiencies?
  • On-premise or Cloud solutions?
  • What ETL Tools should we use?
  • From where do we need to access this information?
  • How will we handle security?

Cloud-based warehousing has become quite trendy, and for good reason. It boils down to a couple facts decision makers should not ignore.

  1. Reduces latency
  2. Provides layered security options sometimes offsetting liability onto the cloud manufacturer allows us to sleep better at night and may be nice if you have to deal with compliance policies.
  3. Creates an available, accountable system with access from anywhere in the world
  4. Grants us the ability to process huge workloads by leveraging data center compute power.
  5. Reduces cost

4) Insights and Reporting

Great, so now we have KPIs, we have quality data, we have infrastructure, now how do we leverage all together and tie them into insightful and meaningful processes or workflows?

First, let’s go back to the governance policy and find our audiences. They can range from senior management to informational users, to operational users.

Take that information, and let’s think in terms of dashboards, reports, and tools– these tools allow users of all types to get a detailed or high-level view to answer the question: Is everything okay with my business? But they can also allow some sophisticated drill down capabilities to dig into the information you truly seek.

There’s a couple of ways we can render insights, allow people to get the insights they need specifically, and collaborate with your team. Let’s start with the basics:

Implementing insights into your workflows, or key metrics out to your team, is a great way to leverage all the hard work you’ve put into place. Let’s inform our teams so they know that they’re performing better, let’s inform our customer base that they’re important by sending them coupons. Let’s include hard metrics as well as soft metrics like satisfaction, and expectations met as key metrics to your business so employees don’t get one tracked on a single KPI. It’s important to approach your employees with care because they ultimately decide whether your businesses are successful.

5) Tools

Selecting the right tools is important, but not as important as who’s going to be using them. Are your report audience end-users? Are they looking for self-service or ad-hoc reporting? Or… are they more laid back and respond well to the traditional canned monthly reports? Whatever their flavor, we should include that in the functional aspects of whatever tool we decide to use.

There’s a ton of good tools to look out for, but here’s some questions I ask myself when looking at the versatility of these reporting tools:

  • Does it enable my primary users?
  • Is there ample support or a community I can reference for help?
  • How easy is it to connect to my data sources?
  • Can it handle ad-hoc reporting well?
  • Can I share and collaborate with my colleagues?
  • What is the impact this tool brings to my day to day?
  • Is there a large gallery of visuals?
  • How do I iterate on KPIs I no longer find useful?
  • What about complex aggregations?

What’s next?

Once you have your BI strategy in place, make sure you begin reviewing your approach at adding intelligence components. Your next phase of intelligence can phase in machine learning, data science, or IoT (Internet of Things). Keep refining and keep evolving with your business. After these valuable tools are in place you can begin asking powerful questions and deriving meaningful insights that drive your business to new horizons. If you think this is something you want to approach and you might need some help, please feel free to contact me at aborgetti@flexmanage.com.

By | 2018-06-12T15:53:59+00:00 June 12th, 2018|Business Intelligence, Power BI, Uncategorized|0 Comments

About the Author:

Anthony is a Senior BI Consultant working out of Chicago. He has six years of experience in Business Intelligence (BI) with a focus on Power BI, Internet of Things (IoT), and Big Data. He possesses strong data warehousing and SQL skills, including SSIS and SSRS. He also has programming experience with power shell, C# and asp.net.