What are Business Intelligence tools? Why use them?

Business Intelligence tools facilitate better reporting and analysis – tools like Microsoft Power BI or Tableau automate and simplify the process of generating reports that are otherwise manual, saving time and reducing the potential for human error. They also provide a platform for comprehensive data analysis, enabling users to gain a deeper understanding of what’s happening within the company.

They enhance decision-making across the company – by aggregating, manipulating, and presenting data in a clear, digestible format, these tools allow managers and executives to make better business decisions based on accurate, up-to-date information.

They create more comprehensible visualizations of your data – they present data in a visually intuitive and easily understood format. Compared to traditional spreadsheets, they make it easier for stakeholders to understand and interpret the data, and make more effective data-driven choices.

When should you implement and begin using a BI Tool? What do you need to have in place ahead of time? 

Implement a BI tool from the time you have five employees – even small companies can benefit from a BI tool. Tools like Power BI and Tableau offer free versions that can help you get started. As your company grows, you can consider paid licensing and bigger implementations. 

You need to have your data in working order – before implementing a BI tool, ensure your data is in a good state. It doesn’t need to be perfect, but data should be checked thoroughly before releasing the BI tool to the rest of the team.

Centralized data storage improves your BI functionality – BI tools can work with flat files like Excel files and CSVs, but they’re more powerful when connected to databases, data warehouses, or data lakes.

Who should own BI? Who should be involved in the selection and implementation process? What third parties might be involved?

IT if you have it, whoever will champion it not – in larger organizations, the IT department often takes ownership of the BI tool. In smaller companies, it could be anyone in the C-Suite from the Head of Analytics, a CMO, a COO, or even the CEO—although the CEO owning it is not ideal. Marketing departments often find BI tools extremely useful and if they’re introducing the tool, the Chief Marketing Officer can lead the charge.

Those involved in selection and implementation include: 

• Developers –  they are the ones who will be using the tool most frequently, so their input is crucial.
• Analytics team – they can provide insights into what skill sets they already possess, which can guide the selection process.
• C-Suite – they have a holistic view of the organization’s needs and can provide strategic direction.
• New Chief Data Officer or Chief Marketing Officer – they might have experience with a particular BI tool from a previous company and offer input based on that.
• IT and Analytics Groups – they can provide technical expertise and insights into how the tool can best serve the organization’s needs.

Selection

What are the major available Business Intelligence tools on the market? What are the pros and cons of each?

Qlik Sense
Pros	• Great at data modeling – they just recently purchased an ETL tool and they’re the first to have an ETL or API tool to bring in data.  • It’s easy to model within the tool before building • Data-centric and they make datasets available • It comes with good pre-built functionality – e.g. the ability to click and filter a dashboard easily. This is great if you have a beginner developer.
Cons	• You almost need to have an expert from the beginning – some of the other tools you can play with. Qlik is more complicated in getting the model set up and ready to visualize.  • You want demo dashboards – to help you understand what you have and can do with a tool.  • You hit the limits with available customization pretty quickly – table formatting and fonts are common complaints, and there’s no way around it. Qlik is like Apple–some people love the design, but if you don’t like what you’re given then you’re in a bad place

Tableau
Pros	• Very easy to get started • Getting data connected is very quick – whether it’s a database or flat file, it takes seconds to minutes. • Drag and drop capabilities – you can build simple quickly. • A lot of customization and the most flexibility  • Plays with a lot of different data sources really easily – if you have data coming from all over the place, Tableau is a great choice.
Cons	• With the customization comes the need for an expert – I’ve seen companies used as the front end of a product that a company was selling. There’s so much you can do with it. The more you want to customize and develop, you’ll need an expert. • With Tableau Public, you can’t save locally – it has to be saved to the cloud and you can’t connect to any databases.  But if you’re working with CSVs you can use it forever. Their free version can be saved by those who have Tableau Public or Reader downloaded.

PowerBI
Pros	• More advanced chart types come packaged • Better Data modeling than Tableau (but not Qlik) • If you’re a Microsoft shop – if you already have Azure and have someone who knows Power Query and Power Automate, those are going to lead you to choose PowerBI—because the integrations are so clean.
Cons	• Less customizable than Tableau – but more than Qlik. • The free version is free until you have to start sharing it with other people

Implementation

What kind of tools and data should you have integrated and connected with your BI tool? 

Integrate your Single Source of Truth – a single source of truth (SSOT) is a concept that an organization should have one source of data that everyone agrees is the real, trusted number. It could be a data warehouse or data lake. This SSOT ensures the integrity of data across different departments and teams, preventing discrepancies and confusion. 

Data integration tools can expand the pool of data available to your BI tool – tools such as Stitch, Evo, and Fivetran can be used to pull data from various sources and store it in a single location. These tools use APIs of common SaaS tools to extract data and store it in a data warehouse. 

Data you may want to integrate includes:
• CRM • Marketing • Finance  • Social media • Operations  • Event Tracking		• HR  • Talent reporting  • Logistics  • Sensor data (e.g. tracking shipments or physical store data)

You may have to transform your data – sometimes, data from different sources may need to be transformed to make it compatible with your BI tool. This could involve changing the format of the data, combining data from different sources, or cleaning up the data to remove errors and inconsistencies. Some tools like Power BI or Qlik have data modeling capabilities and can work with data that isn’t already organized in a data warehouse. But having your data organized in a warehouse before it goes into the BI tool will make your life a lot easier. 

Include whatever data is relevant to your business – the data you collect can be used in a variety of ways, from tracking employee training to monitoring customer behavior. The key is to identify what data is most relevant to your business and how it can be used to drive decision-making and improve performance.

When should you use a BI tool vs. reporting within your CRM, GL or ERP?

Reporting within point solutions works for teams doing siloed work – if the only group interested in the data is the one using the tool, then reporting within that tool can be sufficient. For instance, if Finance is the only one interested in the financial data in your data, reporting through your ledger may work. Reporting within these tools can limit cross-functional insights. 

Integrated reporting is crucial for cross-functional insights – as soon as other stakeholders, like the CEO, want to see how different functions are performing and how their interplay is driving results, integrated reporting becomes necessary. This is where a Business Intelligence (BI) tool comes into play, connecting data from different functions to provide a holistic view of the business.

Point solutions often have reporting limitations – tools like your CRM and general ledger are not designed for reporting. They are made to track projects, clients, opportunities, etc. Their reporting capabilities are often limited, making it difficult to gain useful insights from the data. For complex reporting needs, it can be beneficial to pull data into a BI tool for better visualization and understanding.

How does your data storage affect BI implementation? How might a data warehouse affect your implementation? 

Consolidating your data to a single source of truth will make your BI tool more effective – storing all of your data in one place is crucial for effective BI implementation. This ensures that all your data is consistent and reliable, providing a single source of truth for your business.

A data warehouse or a data lake can benefit your data strategy  –  you can start with a free version on a small scale, but if you’re planning to be a data-driven company, it’s best to build your strategy and get started with the tool you want from the beginning. Switching between different databases can be challenging and may require a lot of rework.

Keep as many calculations as possible in the Data Warehouse – keeping calculations, SQL modeling, and transformations in the data lake or warehouse makes it easier to switch BI tools in the future if needed, as all the front-end work won’t have to be scrapped and rebuilt in the new tool.

What work might you have to do to ensure that your data is usable and clean? 

Key data cleaning practices include: 

• Standardize as much punctuation and capitalization as possible –  this can be particularly important when dealing with company names. 
• Ensure accurate data types – for example, numeric fields should not be treated as text fields and vice versa. You can’t perform calculations on text fields, and treating numeric fields as text can lead to errors.
• Review for missing values and outliers – always check for any missing data and whether there is any data you don’t want to include in your set because it’s an outlier or a misinput. 
• Check for common misspellings – especially when dealing with dimensions where you have a small number of values, like company names.
• Look at averages, minimums, and maximums in your numeric fields – if you have a unit price field and the maximum value is significantly higher than the most expensive product you sell, something might be wrong. This can help uncover issues quickly.

What common dashboards might you create? What are common reports or tables you should run?

	Dashboard Type	Dashboard Description
By User	Departmental	Every department in a company should have its own dashboard to monitor their progress and performance. For instance, the marketing department needs to know how their efforts are translating into results, while the finance department needs to keep track of the company’s financial health. These dashboards are typically restricted to the respective departments or individuals who need access to this data.
	Executive	A consolidated view of all the departmental dashboards. This provides a high-level understanding of the company’s overall performance, enabling better decision-making.
By Purpose	Explanatory	Most dashboards are explanatory in nature. The goal of these dashboards is to tell a story about what’s happening in the company or a particular department. For example, instead of simply presenting “sales by month,” an explanatory dashboard might highlight that “sales have increased by 30% over last year.” This directs the user to the key insights they should be focusing on.
	Exploratory	These dashboards are designed for managers and analysts who need to dig deeper into the data. Unlike explanatory dashboards, exploratory dashboards are not designed to tell a story. Instead, they provide access to data, allowing users to extract insights and draw their own conclusions.

The effectiveness of a dashboard depends on being easy to understand and actionable – whether it’s a departmental dashboard, an executive dashboard, or an exploratory dashboard, the goal is to enable better decision-making and drive business success.

Sample Marketing Dashboard

Customer Feedback Dashboard

Usage

What roles can help with making the most of your BI tool? 

BI Experts – have a background in business intelligence or analytics. They are more likely to have strong backend data engineering skills, making them comfortable working with databases and data warehouses. They can think through user experience, build dashboards, and understand business requirements. 

Analysts – business-focused individuals who are adept at interpreting requests and results. They frequently work with data, but may lack a strong background in data warehousing or data engineering. As a result, data might need to be prepared for them. They are skilled at interpreting results and generally have a higher level of data visualization capabilities than a BI expert.

Data visualization specialists – usually have a background in data engineering or data warehousing, but their skills in these areas might not be as strong as a BI expert. Their strongest skill set lies in understanding the business and translating that into visualizations. They excel in user experience, building engaging and easy-to-navigate dashboards, and high-quality design.

How can you increase uptake and fluency of employees across your organization with your BI tool? 

Offer data literacy training around BI implementation – before you even select a tool, starting with a data literacy training can significantly boost the engagement and adoption of data insights. You can offer employees a basic understanding of what to expect from the tools, how to interpret data, and why certain data representations are used over others. 

You can offer more sophisticated trainings as your use of data matures – as you progress in your data journey, training requirements evolve. Initially, end users might just need to learn how to interact with the dashboards. As the business starts developing its own dashboards, end users may need specific training on how to work within the BI tool.

Have a data strategy – to get the most out of your BI tool, build a strong strategy about how you want to use the dashboards, who will see them, who will develop them, and how you will control who in the company gets to see what.

Some super users can participate in training from developers – not all end users need to be involved in developer training, but including some of the most important users help them understand what they can ask for and prevent implementation problems. By sitting in on beginner-level development training, end users can better understand what to expect from the tool.

What are common ways for businesses to increase the value they get from their BI tool? 

Self-serve analytics is a potential end goal for some companies – at some point, many companies don’t want to create a bottleneck with only a few developers available to answer every request. There’s often even initial resistance from employees who are used to working with Excel and creating their own answers. As you add dashboards and developers and the demand for data increases, you can create controls on data access and allow employees to create their own analytics and dashboards. Your pace will depend on your company and employees—though setting it up for sales early on is often beneficial. 

Aim to surpass Excel reporting, not replace it – the transition from Excel to a modern BI tool should not be about mirroring your old reporting in a new tool. Instead, focus on finding answers to questions and create detailed reports to support decisions. Understand why your reporting existed and look to improve on it.

Hire a BI Expert – don’t try to use a team member who isn’t a BI expert to implement a BI tool. Sending a non-expert to a training won’t equip them to build everything. This often leads to poor user experience, as the dashboards may not be engaging, the data may be incorrect, and the insights lack value. Hire an expert, even if they’re only a part-time consultant. They can provide guidance, build what you need quickly, and train your team to maintain and develop the tool in the future.

Get the whole company on board with the BI tool – implementations should really be organization-wide. It’s often one department that leads the implementation because they’ve heard the tool is great for their specific needs. If they limit the tool to one department, that can cause tremendous problems down the road when others want access. Every department can benefit from dashboards and the insights provided by the BI tool. Therefore, it’s crucial to get the whole company on board to maximize the value of the BI tool.

What are data visualization best practices that you can implement to increase the sophistication of usage? 

Avoid pie charts with more than three slices – pie charts can be useful, but they become confusing quickly. Humans are not good at interpreting angles, so it’s difficult to compare the sizes of different slices. If you have more than two or three variables, consider using a different type of chart.

Embrace bar charts – some try to vary their chart types because they want variety in their images. However, bar charts are one of the easiest types of charts to understand. Our brains are good at interpreting the relative length of bars. Don’t shy away from using bar charts just for the sake of variety. 

Use color sparingly and purposefully – as soon as you introduce colors, users want to try the whole rainbow and needlessly use lots of colors. You don’t need to assign each bar in a chart its own color. This can overwhelm and fry the eyes of the viewer. Instead, use color to highlight and accent key data points. 

Replace red/green with orange/blue – keep in mind that a significant portion of the population (up to 8% of males) has red-green colorblindness. Avoid using red/green and consider colors that are more easily distinguishable, like orange/blue for negative-to-positive charts.

Don’t reuse colors with representational significance – if you do use a color to represent a particular category in one chart, don’t use that same color to represent something totally different in another chart. This can lead to confusion and incorrect assumptions. 

How do you measure and track successful implementation and usage of your BI tool? 

Look at user engagement and adoption – a successful BI tool implementation is judged by high user engagement and adoption. Track usage stats like how many clicks happen in dashboards, dashboards opening, and user count. Monitor which users never log into the tool, and which ones use it frequently.

A BI tool is only as valuable as the insights it extracts – if the BI tool is helping you make better decisions by providing valuable insights, then it’s doing its primary job. Even if you have well-designed and frequently used dashboards, they might be failing if they’re not relevant to your decisions. You should be able to point to specific instances where a dashboard or chart provided an insight that led to a beneficial change in your business operations.

What are the most important things to get right?

The strategy for how you want to use data – most people when they get to the point of BI selection are thinking about tools capabilities or skills capabilities. The first thing you should think about is how you want to use the tool. Do you need a high level of customization? Do you need a high level of design? Do you need to embed them? What will your reporting look like? Is it important to do heavy data modeling in the BI tool? Where is the data coming from? Do you need AI or

Design a strategy for data usage before selecting a BI tool – before you even start considering the capabilities of different BI tools, it’s crucial to have a clear strategy for how you want to use data in your organization. This strategy should guide your selection of a BI tool, not the other way around. 

What are common pitfalls? 

Hiring one data company to do too many things – if you hire a Salesforce consulting firm to set up your CRM and set up your dashboard. A person who develops and builds out your CRM won’t be good at dashboards. The two skill sets won’t have great overlap. 

When designing your data usage strategy, consider the following:

• Customization needs – Do you require a high level of customization in your BI tool? 
• Design quality – Is high-quality design a priority for your organization?
• White labeling and third-party reporting – Do you need to white label the tool or send reports to third parties? If so, you’ll need a robust licensing system.
• Embedding reports – Will you need to embed these reports somewhere?
• Data modeling – Is it important to do heavy data modeling in the BI tool?
• Data sources – Are you pulling in multiple data sources? 

Don’t select a tool just because it’s popular or advanced – It’s easy to get swayed by popular tools or those with advanced features like AI capabilities. However, it’s essential to consider whether these features align with your organization’s needs and readiness. For instance, while AI is undoubtedly the future, it may not be necessary for your current operations.