Summary: What is embedded analytics and why is it so important? What should you look for in an embedded analytics platform? While the requirements vary by company, I believe that certain elements are essential in a good embedded analytics tool. In this article, we explore 7 essential elements you should look for in any good embedded solution.
A growing trend, embedded analytics refers to the integration of self-service BI tools into non-BI applications. According to Gartner, “Embedded analytics is a digital workplace capability where data analysis occurs within a user’s natural workflow, without the need to toggle to another application.”
Why is it so important? How do businesses use embedded analytics platforms? Here are two important ways:
1. Software vendors embed existing BI and reporting tools into their platforms. For example, an ERP vendor might embed an analytics solution within their product. Why? This approach provides a few big advantages:
- It improves their product offering. Analytic capabilities add another selling point to their software. Some vendors even offer these features as paid add-ons, meaning they instantly create a new revenue stream.
- It improves the customer experience. It lets customers analyze their data, create reports, and enhances overall data visibility.
- It’s far easier than building their own BI solution. Creating a custom reporting solution is difficult and usually produces an inferior result. Embedding a pre-built solution is faster and results in a better customer experience.
2. Other businesses (non-software vendors) use embedded analytics in a couple of ways:
- They replace inferior reporting features found in other software packages.
It’s the biggest complaint we hear about enterprise software. The reporting features don’t meet their needs. This typically happens when a software vendor tries to create their own reporting features.
ERP reporting is a common example. ERP packages include a reporting module, but they’re often limited. Their customers can either replace the ERP system or replace the reporting. Most choose the latter, and adopt an embedded reporting tool.
- They bring business analytics to the users. Depending on which survey you read, the failure rate for traditional BI software is anywhere from 50% – 80%. Why does this happen? While I could list off many reasons, user adoption is one of the biggest. After all, if no one uses it, BI is worthless.
Embedded analytics helps solve this problem. It brings reporting and BI to the business users. It works it into their daily routines. It makes the BI features look like they belong, and keeps the user in their workflow. The result: Improved user adoption.
Now, this is just a brief explanation of embedded analytics. If you’d like to learn more, we put together a guide that explains it in more detail. In this article, let’s focus on features. What elements should you look for in an embedded BI tool? In this article, we’ll explore 7 features to examine in any embedded analytics software.
1. White labeling
Embedded BI is different than traditional BI in that it’s built into your existing software. The charts, dashboards, reports, and data visualizations must look and feel like they’re part of your existing software.
As such, white labeling is essential.
White labeling means that you can customize the BI software to look and feel like your current software. It means you can remove the vendor’s branding and replace it with your own. It also means you can customize the overall look and feel to fit with your software.
What areas can you customize? It should let you match the fonts, color scheme, branding, and any other aspect of the embedded analytics with that of your own software. When done correctly, white labeling provides a consistent and seamless user experience.
Be careful: Not every vendor offers the same white labeling options. Look for a solution that allows for total control over the look/feel of your generated reports and applications.
2. Friendly licensing model
One of the biggest “gotchas” with embedded analytics software: The licensing model. This is where embedded BI and analytics can get expensive in a hurry if you’re not careful. Here are a couple of licensing areas to inspect.
Per-user pricing is fairly standard. The vendor charges a seat fee for every person that uses their software.
But, what is a user? Is it an employee that’s creating the reports, dashboards, or analytics applications? Or, is it someone that’s accessing those generated applications? The answer can make the difference between a few users and hundreds. It’s an important distinction to make before a purchase.
If you plan to sell or distribute the analytics applications you create, watch out for distribution fees. This means you pay the software vendor a fee every time you distribute those applications to your customers.
For instance, if your customers have 100 users who need access to your generated report, you’re paying 100 different distribution fees. As you might imagine, this gets expensive quickly.
These are just a couple of different areas to look at before you buy, but you’ll find plenty more. The most important point: Don’t assume that all licensing models are the same. Some vendors have a flat fee, while others charge by the database. Some charge distribution fees, while others don’t.
Which is best? It all depends on your business and your needs. But, it’s important that you understand the potential pitfalls.
3. Key security features
“One of the most important aspects that you should look for when it comes to embedded analytics software is the security of your company’s data,” says Will Ellis, Founder of Privacy Australia and IT security consultant. “If your company already has security measures in place, will the software be able to adapt to meet your needs? If you are looking for new security measures, what kind of security does the software offer? Are there multiple layers of security that you can utilize? These are just a few of the questions that companies should be asking themselves before implementing any embedded analytics software. By doing this, you will be able to get a clearer view of exactly what it is that you need and choose your software accordingly. My ideal go-to when it comes to searching for features of embedded analytics software is the level of security. I like to look for software which not only restricts user access and protects data, but records actions, allowing me to find the source of any security issues that may arise.”
What types of security features should you look for? Here are 5 must-have security options:
- Row-level (or multi-tenant) security
Multi-tenant security controls data access at the row level. This means that many users can access the same application, but only view the data they’re authorized to access.
“Row-level security makes a big difference,” explains Sean Werick, Managing Director of Analytics at Sparkhound. “For example, if one person is accessing a web page with embedded analytics, he or she will see something different than a person accessing the same page who may have different security privileges. The data is in one central repository, but each person sees a different view of that data to which they are entitled to and this is generally done via Active Directory.”
- Single Sign-on (SSO)
A session/user authentication process, SSO eliminates login prompts when switching between applications in a single session. It lets users login in one place and authenticates the user for all authorized applications.
- User privilege parameters
User privilege parameters personalize features and security to individual users. Saved to a user’s profile, user privilege parameters control user-specific features throughout every report, dashboard, or BI application.
- Role-based and user-based security
This option lets you secure application access by individual users, or their roles. Each user can only see the applications they’re authorized to access. For instance, you might want your CEO to see different reports than your sales team.
- User-specific data sources
This feature lets developers control database access on a user level. They can create a single application that will access different data sources depending on the user.
It’s a common problem: Organizations license a BI solution based on their current needs, but don’t plan for scalability. Of course, I realize they can’t predict the future. But, there are some aspects of scalability that you can plan for.
Let me explain. I look at scalability in a couple of ways:
- Cost: If you’re successful with this software, will you be able to afford it? As touched on above, some solutions become prohibitively expensive as your needs change.
I’ve seen businesses get hit with rising user fees, distribution fees, run-time fees, and more. They licensed a tool at a low initial cost…and it ballooned out of control as they grew. Now, they’ve invested time and money into a solution they can no longer afford.
- Capabilities: Will this software support your long-term goals? I know…this might seem obvious, but it’s often ignored.
Look past your current project or needs with selecting embedded analytics tools. What capabilities does it offer? Is it tied to one technology stack, or will it work anywhere? Can you deploy it in-house and in the cloud? Is it web-based?
My advice: Plan for growth from the start. Ask the right questions. Is the licensing structure designed for growth? Is the software built on technologies that will still be current in a few years? If we decide to move to/from the cloud, will this still work?
“The number one thing to look for: make sure your choice of embedded analytics software scales,” says Werick. “Anything that is embedded, is typically an all-encapsulated product. To implement embedded analytics, such as into a web app, make sure the analytics platform itself scales.”
5. Easy integration
One of the most important aspects of embedded analytics software is also one that you can’t see: The architecture. Some vendors build their software on proprietary architecture. Others build it on open architecture and frameworks.
Why is this so important? It all boils down to integration.
Any embedded BI software you choose needs to integrate with your existing systems. The architecture could be the difference between a simple or difficult (if not impossible) integration.
I’ve seen businesses struggle for over a year trying to integrate a BI tool, before finally ditching it altogether. Make sure you understand how easily you can integrate a solution with your existing systems before you make the purchase.
“When choosing an analytics platform the primary consideration is speed,” says Ozzie Bock, CEO of AltiView Technology Group, LLC. “How quickly can I gain the insights I am looking for? To me the speed upon which I can make a business decision is paramount.”
I completely agree. In today’s data-driven world, speed is the name of the game. You can’t afford to wait around for insights. You need real-time information to make data-driven decisions.
When it comes to embedded analytics, look for speed in a few different areas:
- Speed to get up and running
How long will it take your team to become productive with the software? This ties in with the integration point above, but also with the overall learning curve. What’s the learning process and how long will it take?
- Speed to create reports, data visualizations, dashboards, etc…
What goes into a build process? Can you create applications in minutes? Hours? Days?
- Speed to get the required information
From an end user’s perspective, how easily can they get the data they need? Is it a multi-step process, or is the data at their fingertips?
7. Key capabilities
Finally, let’s explore some key capabilities to look for in any embedded analytics tool. I’m only covering the key points in this section, but the list could be much longer. Here are the most important features to look for:
Self-service analytics and reporting
Any good embedded BI tool should provide some sort of self-service options. For instance, some will let users create applications on their own. Others let you include filters in the reports that let users see their data in different ways. Make sure you understand the extent of the self-service options provided by the tool.
Ad-hoc reporting is essential because it lets end users create and distribute reports on the fly. The user selects the data elements he/she wishes to see in the report. Then, they can either export the report into a format of their choosing or email it from their web browser.
What if your end users need more information than what’s displayed in their dashboard or report? That’s where drill-downs come into play. An effective BI application starts with a high-level view, and provides drill-down capabilities.
For instance, suppose a dashboard displays product returns. It should also let the user drill down and see where products are being returned, reasons for the return, and any other pertinent information.
Any modern embedded analytics solution should provide application programming interface (API) integration. In short, this lets your application send and receive data from other software applications or services.
These days, users are connected to the web at all times. They expect a consistent experience across any device. Make sure you choose a tool that automatically creates responsive applications.
Interactive reports display as much or as little data as the user desires, and lets users perform data analysis in different ways. It starts out with a high-level view of business data and lets users filter, sort, and drill down to the most minute details. Interactive reports are so useful because they let users view their data in any way they wish.
A business dashboard displays high-level graphs and reports in one easy-to-use interface. It lets executives see critical data and can alert a company to problems before they get out of hand.
In the past, BI & reporting software were “read-only.” They pulled data from the database, but couldn’t write data to the database.
As businesses become more “data-driven”, we’re seeing this change. Modern BI reporting tools should include read and write capabilities. This gives users the ability to send alerts, perform actions based on data, and trigger workflows.
Automated alerts and notifications
With traditional BI reporting tools, you need to physically look at the reports or dashboards to gain any insights. Modern tools should be proactive, and let users set up alerts and notifications for critical data changes.
Modern reporting software should let you create and trigger workflows based on changes in data. This might include actions like approval workflows, sending emails, running reports, etc…
Hopefully, this article gave you a clearer view of embedded analytics. Do you have questions or comments? Feel free to comment below!