{"id":107009,"date":"2025-12-17T09:00:00","date_gmt":"2025-12-17T17:00:00","guid":{"rendered":"https:\/\/www.crazyegg.com\/blog\/?p=107009"},"modified":"2025-12-18T10:57:00","modified_gmt":"2025-12-18T18:57:00","slug":"embedded-analytics","status":"publish","type":"post","link":"https:\/\/www.crazyegg.com\/blog\/embedded-analytics\/","title":{"rendered":"Adding Embedded Analytics to Your B2B App? Here\u2019s What Users Want (Research & Analysis)"},"content":{"rendered":"\n
It\u2019s now the norm for B2B apps to have customer-facing analytics. <\/p>\n\n\n\n
Grammarly Pro, for instance, comes with a dashboard. Canva\u2019s reporting is also pretty nifty. And Coursera for Teams lets users track an assortment of metrics, including new employee competencies. <\/p>\n\n\n\n
If you\u2019re planning on adding embedded analytics to your app, you\u2019ve made a good decision. Done right, they are a strong differentiator. But it\u2019s vital to first understand how to create user experiences that are both satisfying and add genuine <\/em>value. <\/p>\n\n\n\n
What Is Embedded Analytics?<\/h2>\n\n\n\n
Embedded analytics is the integration of a reporting dashboard into a customer-facing app. It\u2019s most common in B2B SaaS (as in the Canva example below), but many B2C companies also offer it. Banking and gaming apps are two well-known examples.<\/p>\n\n\n\n<\/figure>\n\n\n\n
There are three parts to an embedded analytics solution:<\/p>\n\n\n\n
\n
The database:<\/strong> Contains the data the user dashboard will show<\/li>\n\n\n\n
The semantic layer (or data modelling layer):<\/strong> Retrieves and structures data from the database in response to dashboard requests<\/li>\n\n\n\n
The user interface (UI): <\/strong>Presents the data through charts and visualizations<\/li>\n<\/ol>\n\n\n\n
Up until fairly recently, most developers had to rely on APIs and design their frontends from the ground up. The only other option was to embed clunky iframes. However, things have changed. Most vendors now offer SDKs that let users define models and build dashboards quickly to integrate them into their apps. Software engineering skills are still required but the dev burden has significantly reduced. <\/p>\n\n\n\n
We conducted a user test to identify which core features customers wanted from analytics tools in general<\/em>. We also asked participants about their biggest frustrations with popular platforms. <\/p>\n\n\n\n
To tailor these principles to embedded analytics, the results were combined with the in-house expertise of our team. The focus was on creating a practical UX foundation that would be useful for B2B apps for which reporting was a supplementary feature and not <\/em>the primary category. <\/p>\n\n\n\n
The resulting prioritization matrix tells you which features to prioritize in your own embedded analytics. You should include Mission Critical <\/em>and Must-Have<\/em> features before rollout, adding others\u2014in the Secondary Focus <\/em>and Optional Enhancements <\/em>quadrants\u2014as resources allow. <\/p>\n\n\n\n<\/figure>\n\n\n\n
Let\u2019s look at each of the reporting features and UX elements in the matrix in a little more depth and define exactly what they mean and when and why they matter, starting with those in the highest priority quadrant. <\/p>\n\n\n\n
\n
Clean, intuitive user interface: <\/strong>A clean UI was a must-have feature for participants. It was also apparent from responses that they struggled with cluttered dashboards in other platforms, with GA4 often mentioned. One user said that \u201cUsing Google Analytics requires a PhD.\u201d <\/li>\n\n\n\n
Strong out-of-the-box features: <\/strong>Users wanted to be able to log in and track key areas of activity with an absolute minimum of setup time. Lars Lofgren, who I interviewed for this article, said long setup time drove high churn at Kissmetrics: \u201cThe platform required a lot of work to get up and running, and this was a major issue for users.\u201d<\/li>\n\n\n\n
A wide range of drill-downs: <\/strong>One interesting finding was that users varied widely in their drill-down requirements. There was little uniformity in demands\u2014some wanted to understand micro-trends, and others wanted to compare specific data sets. As such, wide-ranging and <\/em>highly granular drilldowns should be a priority.<\/li>\n\n\n\n
Easy-to-understand visualizations: <\/strong>Most users understood data primarily through visualizations (rather than table breakdowns). One respondent went so far as to say that \u201cnumbers are pointless unless compared to other numbers in visualizations.\u201d Visual representations in our test were praised because they were perceived as being well-suited to the activity being tracked\u2014pie charts for percentages, graphs for trends, and bar charts for categories. <\/li>\n\n\n\n
A large integrations library: <\/strong>One of the biggest frustrations among users was the inability of analytics platforms to integrate with their existing tech stack. This shortcoming was often a deal-breaker\u2014users said things like, \u201cI don\u2019t want to go between dashboards,\u201d <\/em>and, \u201cWe have a process that works. We don\u2019t want to change it.\u201d<\/em><\/li>\n\n\n\n
Custom event tracking: <\/strong>In addition to plug \u2018n\u2019 play functionality, users repeatedly spoke about user-defined behavioral logging. This should fall high on your priority list if <\/em>it\u2019s essential for core decision-making in the org. To give you an idea of the demand for custom tracking, one user said, \u201cWe currently have 50 custom events we\u2019re tracking.\u201d<\/em><\/li>\n\n\n\n
Compliance with data laws: <\/strong>This is a must-have, and prospective buyers will often avoid non-compliant platforms simply because they are bound by company regulations.<\/li>\n\n\n\n
Clear, shareable documentation: <\/strong>One user summed up the need for documentation perfectly: \u201cI would like to share written documentation with my colleagues to make sure that it doesn\u2019t present any technical or information security issues for us.\u201d <\/em>Demand for detailed technical documents was largely in the context of securing team approval, rather than onboarding. Buyers needed to know that there wouldn\u2019t be any security or technical issues. <\/li>\n\n\n\n
Unified reporting that connects multiple sources: <\/strong>Unified reporting provides packaged insights for data from multiple sources. This way users can act on it without the need for heavy analysis. AI features now make this possible in a meaningful way without the need for any <\/em>direct input from users. <\/li>\n\n\n\n
Customizable dashboards: <\/strong>Customizable dashboards are a secondary focus because they were not universally sought after by the study participants. Most users are happy to work within the existing layout of a dashboard, especially if it\u2019s well designed, and will only make minor modifications if they are given the choice. <\/li>\n\n\n\n
Data provenance (capture and storage transparency): <\/strong>Data provenance is a highly detailed account of how <\/em>data is captured and where <\/em>it is stored. This fell relatively low on the demand spectrum\u2014simply knowing that compliance was accounted for was usually enough\u2014but it was raised. <\/li>\n\n\n\n
Fast (ideally real-time) data: <\/strong>Real-time data was a lesser concern among users. Contrary to the way it is touted as a must-have feature, real-time analytics will often not be a primary requirement among users of embedded analytics.<\/li>\n<\/ul>\n\n\n\n
Should You Offer Analytics?<\/h2>\n\n\n\n
If you answer yes to one of the following questions, you have a strong reason for including them in your app. <\/p>\n\n\n\n
Do analytics drive core decisions? <\/h3>\n\n\n\n
Tracking is often a core operational necessity. In these cases, embedded analytics are a baseline feature, and the quality of your dashboards compared to other providers can make or break a deal. <\/p>\n\n\n\n
You must <\/em>include analytics if your customers rely on your platform for any of the following: <\/p>\n\n\n\n
\n
Compliance obligations<\/li>\n\n\n\n
Staff activity monitoring that feeds into HR decisions<\/li>\n\n\n\n
Sales and revenue tracking<\/li>\n\n\n\n
Marketing campaign performance <\/li>\n\n\n\n
Inventory management <\/li>\n\n\n\n
Financial reporting for tax preparation<\/li>\n\n\n\n
Embedded analytics are particularly crucial if your users have well-defined operational data needs and<\/em> limited tech experience. Pulling data from Shopify\u2019s databases into an in-house BI platform, for example, is beyond the scope of many independent merchants.<\/p>\n\n\n\n<\/figure>\n\n\n\n
Are analytics necessary for market parity? <\/h3>\n\n\n\n
You shouldn\u2019t follow the market for its own sake. Fads come and go, and chasing them mindlessly is a sure route to wasted resources. <\/p>\n\n\n\n
Instead, aim for market parity if you meet the following two criteria:<\/p>\n\n\n\n
\n
You see a trend towards embedded analytics among your competitors. <\/li>\n\n\n\n
Your customers demonstrate a clear desire for them in focus groups and sales calls. <\/li>\n<\/ol>\n\n\n\n
We\u2019ve seen a category shift towards embedded analytics in project management apps like Monday and Asana (pictured below). Dashboards tend to be relatively straightforward with limited functionality. They are add-ons rather than essential functionality, but users regularly cite them as valuable in online reviews. <\/p>\n\n\n\n<\/figure>\n\n\n\n
Will embedded analytics drive meaningful <\/em>differentiation? <\/h3>\n\n\n\n
Embedded analytics can <\/em>be a competitive differentiator if they report metrics that drive change and users can act on them<\/em>. <\/p>\n\n\n\n
This is a key point: you can only differentiate by improving an outcome for users, not by providing glitzy charts. If you can\u2019t identify specific decisions in your customer\u2019s workflows, realistic scope for influence, and definite outcomes tied to the metrics that matter to your customers, meaningful differentiation likely isn\u2019t a possibility. <\/p>\n\n\n\n
Grammarly is an excellent example. As one of the market leaders in writing assistants, its analytics tools are both unique and useful. Organizations often want to test different writing tones with a view to improving customer engagement.<\/p>\n\n\n\n<\/figure>\n\n\n\n
How to Design Your Customer-Facing Analytics: 5-Step Playbook<\/h2>\n\n\n\n
The features in the embedded analytics matrix describe what a finished platform looks like. However, they are the result of a multifaceted process whirring away under the hood. <\/p>\n\n\n\n
Providing \u201cstrong out-of-the-box functionality,\u201d for instance, requires you to actually know what functionality users want. In fact, more than this\u2014not just what they want <\/em>but what they\u2019ll actually use<\/em>. <\/p>\n\n\n\n
\u201cClean design\u201d and \u201cdata visualizations\u201d sound simple enough. But pick a vendor that\u2019s misaligned with your company\u2019s software engineering capabilities\u2014one whose \u201cno-code toolkit\u201d is a set of rigid iframes\u2014and you\u2019ll find them impossible to provide. <\/p>\n\n\n\n
With this in mind, I spoke to two experts who know analytics inside out\u2014Crazy Egg CEO Hiten Shah and former Director of Growth at KISSmetrics Lars Lofgren<\/a>.<\/p>\n\n\n\n
![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/16131910\/Canva-Screenshot-1024x576.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/17143742\/Embedded-Analytics-Feature-Priorities-1-1024x614.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/16131934\/Shopify-Analytics-Dashboard-1024x576.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/16132011\/Asana-Dashboard-1024x576.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/16132022\/Grammarly-Analytics-Dashboard-1024x576.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/15153852\/Embedded-Analytics-Solutions-1024x614.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/16132210\/CrazyEgg-AI-Overview-1024x576.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/16132229\/Crazy-Egg-Home-View-1024x576.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/16132245\/Crazy-Egg-Tooltips-1024x576.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/16132256\/Crazy-Egg-Plain-English-Example-1024x576.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/ceblog.s3.amazonaws.com\/wp-content\/uploads\/2025\/12\/16132311\/Advanced-Filters-Example-1024x576.png\")
<\/figure>\n\n\n\n