The data and analytics field is vast. When people ask me what I do professionally, I tell them I work in digital analytics, and unless they are in the area, they have no idea what that means. Some people do analytics work for retail stores, logistics, the stock market, etc. It seems like everyone is doing some analytics these days. Even those in the website/mobile app field can sometimes struggle to explain the difference between marketing analytics and product analytics.
But one of the most significant areas of confusion over the years has been understanding the difference between digital analytics and business intelligence. I have been in many conversations in which organizations tell me that they don’t need a digital analytics product because they have a business intelligence product or vice versa. So in this post, I will explain how I describe the difference between these two disciplines in conversations.
Most of those who follow my blog posts should be familiar with digital analytics. I define digital analytics as the collection and analysis of digital user behavioral data to use that data to improve or optimize digital properties and experiences. Digital analytics products track digital actions (events), campaigns, content, user path flows, and other behaviors customers take when using websites or mobile applications. Typical vendors in the space include Google Analytics, Adobe Analytics, Amplitude, etc. In the past, I have written about how I believe many of the different types of digital analytics products will converge over the next few years.
Business intelligence products have become very popular within organizations, and you would be hard-pressed to find an organization that doesn’t have a business intelligence product. Business intelligence products provide a high-level summary of KPIs that are critical to the organization. Often, business intelligence products take the form of high-level dashboards shared with executives. Business intelligence dashboards often combine data from digital analytics, CRM, physical stores, internal data warehouses, etc. Popular vendors in the business intelligence area include Tableau, Power BI, Looker, and Domo.
Digital Analytics vs. Business Intelligence
So with a few basic definitions, let’s dive into how digital analytics and business intelligence products are different.
Data Sources & Cross-Platform Metrics
Business intelligence products often incorporate data from many different sources. I like to think of this as the “greatest hits” of data from multiple data systems. While streaming any type of data into digital analytics products is undoubtedly possible, most organizations limit data to websites and mobile applications. But as the world becomes more digital, we are seeing more and more customers send digital analytics products like Amplitude data from stores, call centers, and even physical products.
One of the key selling points of business intelligence products is that they can combine metrics from different platforms in ways that would be challenging in one standalone platform. For example, let’s imagine that the digital analytics platform reported an organization had 1,000,000 unique visitors on May 3. The CRM system showed that 20,000 marketing qualified leads (MQLs) were created on the same day. The organization could use a business intelligence product to divide these two metrics to create a brand new KPI called MQL/Unique Visitor. While there may not be an easy way to connect those unique visitors to sales MQLs, at a high level, it may be possible to view trends and see if there is a relationship between the two. While this organization could import MQL data into its digital analytics product, most would choose to do it in a business intelligence product.
In the old days, this type of work would have been done in Microsoft Excel (the OG BI Tool!), but Excel had limitations on data import and database capabilities. I think of business intelligence products as Excel on steroids. The power of business intelligence products is that they can easily combine multiple data sources and empower organizations to mix and match all sorts of metrics from different systems. Often, the joining factor will be the date, but in some cases, other primary keys can be used to join data from different sources.
While some of this could be done in digital analytics products, it would be complicated and time-consuming. Dashboards in digital analytics products tend to focus on summaries of data related to websites and digital applications.
The most significant difference between digital analytics and business intelligence products is in the area of data exploration. While data exploration can occur in both types of products, they are done in very different ways. In business intelligence products, there are typically limits on the types of reports available. For example, if there is a KPI for sales, business intelligence products can break it down by sales rep or region. But in digital analytics products, data exploration includes metric breakdowns and many other report types that don’t exist in business intelligence products. Here are a few examples:
In digital analytics products, there are times when you would want to view how customers navigated pages or events. This can be useful to understand page flow or event flow drop-off and fix any flow leaks. But reporting on path flows requires time-stamped, sequenced data associated with unique visitors versus aggregated data. Creating an accurate path flow report in a business intelligence product would be challenging.
Digital analytics products are often used to build conversion funnels. These funnels plot key checkpoints in conversion flows to see how many customers make it to each step. While they sound similar to path flows, they are different in that they are less focused on all of the paths customers take and more interested in a specific set of steps taken. Conversion funnels are also built such that customers have to perform the actions in a set order to be included. This order sequence requirement means that the digital analytics product must understand which customers have completed each step and in what order. While a business intelligence product could likely report on how many times event1 and event2 took place, it would be difficult to understand if it was the same user who performed both events and in the correct order.
Cohorts & Segments
One of the most powerful aspects of digital analytics products is the ability to build ad-hoc cohorts (or segments) of users. These cohorts can be based upon event behavior, attributes, or navigation behavior. Once created, cohorts can be used to compare different groups of customers, and cohorts can be sent to other systems for personalization or marketing efforts.
Most business intelligence platforms are not user-centric. They focus on numbers more than users. Therefore, it is not common to use business intelligence products to create cohorts of users for analysis or marketing purposes.
A core component of digital analytics is the concept of identity. In digital analytics it is important to know if the current user is the same as a user who used the digital property last week. To address this, digital analytics products have built mechanisms to identify users and determine if they are known or unknown. Some do this via third-party cookies, and others do this via first-party authentication.
Business intelligence products have not traditionally attempted to perform identity resolution. While they can view and join metrics by a customer ID, they are not built to review anonymous user data and determine if the user is a previously known entity.
Understanding which and how many of your customers return to your digital experiences over time is an integral part of digital analytics. Digital teams use digital analytics data to see what features or marketing campaigns drive retention so they can form habits and generate revenue. Reporting upon retention requires identity resolution to know if the customer who is currently engaging with the digital product has been there before and how often.
Business intelligence products can report on usage, but many are not built to understand if the same users are returning again and again. There may be some ways to do this by leveraging customer identifiers, but this has to be coupled with time-series data for each customer and reports that use statistics to show retention buckets and time windows. These capabilities are rarely present in business intelligence products.
Another difference between digital analytics and business intelligence products is how often each type of user engages with the product. Business intelligence products are usually built for and used by upper management and executives. While lower-level staff may use the tools to develop reports and dashboards, the primary recipient of the reports and dashboards is often executives. Business intelligence products often tout how easy it is for executives to learn about their business via business intelligence products.
Digital analytics products are also built for executives, but they are also heavily used by digital analysts, marketing analysts, or product teams. Since digital analytics products provide both high-level and granular information, digital analytics products are accessible to almost anyone in the organization. Executives can view high-level dashboards in digital analytics products, but only the data-savvy ones will dig deeper into the data. I believe that the complexity of digital analytics products was one of the contributing factors to the rise of the business intelligence industry. One of the popular business intelligence products was founded by the former CEO of a digital analytics product. He was frustrated that he couldn’t see the high-level metrics he needed to run his business from his digital analytics product!
Digital analytics products primarily collect data from websites and mobile applications. However, in recent years this has expanded to include many other data types (e.g., store data, call center, etc.). However, the collected data is often at a very granular level. Common data points might include clicks or swipes on buttons and links, viewing specific pages, and phases entered into website search boxes, etc. Most organizations collect event data in the billions each month, and this data is aggregated in reports within the digital analytics product.
While not always the case, business intelligence products often collect data at a less granular level. For example, if you use a business intelligence product to show CRM data, you might feed in leads from Salesforce. This data will often not be as granular as hit-level data on a website. While there are exceptions, many organizations send summary information to their business intelligence product instead of duplicating the source data and all of its granularity. Another example might be piping in orders and revenue from a digital analytics product.
For most organizations, it is required to have a digital analytics product and a business intelligence product, not one. As described here, these products are different but can be complementary. Perhaps one day, there will be industry consolidation, and one vendor will own digital analytics and business intelligence products, but that hasn’t happened so far. Even Google, which owns the largest digital analytics product, purchased a business intelligence product (Looker).
I think digital analytics products might one day be able to address many of the business intelligence use cases, but I think it will be difficult for business intelligence products to tackle digital analytics use cases. While I think the two products will be separate for the foreseeable future, if I were to bet on one overtaking the other, I would put my money on digital analytics overtaking business intelligence versus the other way around.
For now, if your organization attempts to argue that it only needs one of these products, I encourage you to have them review this content and better understand the differences between the technologies. If your colleagues are insistent that only one product is needed, I suggest asking them to demonstrate how they would perform digital analytics use cases in a business intelligence product and vice versa. Typically those who argue for using one product have not had experience with both types of products or are simply looking to cut budgets. It is easy to make the case that digital analytics and business intelligence products are very different, have different objectives, different audiences, and solve different problems.