<p>If there’s one thing that’s increased for everyone building and deploying analytics over the past few years, it’s complexity—more assets, more teams, more moving parts. Analytics has become mission-critical and with that a huge proliferation of dashboards, reports, data models, and dependencies. Analytic assets have grown exponentially for enterprises and for software companies who are embedding analytics increasingly pervasively in their apps.</p>
<p>More than ever distributed teams are all working on different parts of the analytics stack, and they all sit at the intersection of growing and changing data flows and data dependencies. It’s an emerging release control headache. Get it wrong, and it’s a recipe for broken analytics.</p>
<p>In this post, you’ll learn why strong source control and integration with your DevOps processes is increasingly crucial for analytics—and how we’re delivering it here at Sisense.</p>
<h3>Versioning and release management analytics headaches set to grow.</h3>
<p>Dashboard, reports, metrics, and mapping have become first-tier assets in release processes, and are now as critical to version, track, and control, as code assets have always been.</p>
<p>For example, without strong controls, versioning, and integrating with CI/CD processes, with the proliferation of analytics assets and dependencies, it’s too easy for a report to break or return incorrect results when a change is pushed to a data type in the underlying data model. Or when an update to a data warehouse schema is released, an analytic data model no longer maps to the updates. And with multiple analytics developers, things can get very painful very quickly if there’s no way to combine everyone’s contributions into one unified repository or see who contributed a given change (it’s one of the reasons internal analytics teams often try and avoid too much collaboration, for fear of breaking changes.)</p>
<p>Net-net, with so many teams developing analytics assets, and so many other changes happening across the data landscape, it’s easy for teams to get out of alignment―and that can mean a broken user experience and diminished trust in the accuracy (and value!) of data.</p>
<p>Whether you’re at a software company tasked with embedding analytics in your applications or sit within a data analytics team, ensuring strong version management and control around the analytics assets you create and manage is crucial to ensure you can efficiently scale, ensure quality, and have the visibility around changes to resolve issues when they occur quickly. Ultimately, it’s critical for everyone to work together as a team and have a strong integrated and collaborative process from development to test to production.</p>
<p>Achieving this means applying the same degree of release management, tracking, and change controls that has been used by DevOps and software developers for years and fully integrating your analytics development with the Source Control Systems that they use, like Git. As the world’s leading version control system, Git is the standard for managing and tracking revisions. Often, it’s used for managing tracking code changes, but it can be used with any file type or assets—like your dashboards and data models.</p>
<h3>The ugly truth: Analytics technology was never designed for this.</h3>
<p>Ultimately, the solution is to connect  your analytics deployment to a robust Source Control Management platform like Git. It’s the best way to trace easily, review, undo changes, review, manage branches, and control changes to your dashboards, reports, and data models, and ultimately keep release processes aligned and in sync across your environment―and teams.</p>
<p><img decoding="async" loading="lazy" class="alignright wp-image-735986 " src="https://cdn.sisense.com/wp-content/uploads/Git-blog-icon1-300x199.png" alt="" width="277" height="192" /></p>
<p>But the reality is a lot of analytics technology hasn’t, much of it is still pre-DevOps and pre-Git, and simply doesn’t play well with what are now industry standard source control systems. In fact, many analytics tools started out as single-user, desktop software and were never designed for the large-scale, collaborative demands of the modern analytics landscape.</p>
<p>For example, some analytics tools include rudimentary and proprietary built-in version management, separate from Git. What happens is that teams end up not using them at all. And DevOps teams can’t integrate it with their SDLC and CI/CD processes and standard tooling. While other analytic tools provide basic Git integration just for the data model for example but don’t include it across all assets, such as dashboards and other content, risking some assets getting out of sync with the release cadence.  That’s a headache waiting to happen.</p>
<p>If you’re looking to connect all your analytics assets to Git for total integration with your CI/CD processes, it’s been virtually impossible. Until now.</p>
<h3>Introducing total Git integration for Sisense Fusion</h3>
<p>Our design goal for Git integration was simple: integrate version control natively, without imposing additional restrictions or simplifying capabilities for ‘novice’ users, so you can integrate the complete asset chain &#8211; Sisense data models, charts and dashboards &#8211; with your software development lifecycle, DevOps, and CI/CD processes. We want to make broken analytics a thing of the past.</p>
<p><img decoding="async" loading="lazy" class="wp-image-735987 alignleft" src="https://cdn.sisense.com/wp-content/uploads/Git-blog-icon2-300x293.png" alt="" width="240" height="235" srcset="https://cdn.sisense.com/wp-content/uploads/Git-blog-icon2-300x293.png 300w, https://cdn.sisense.com/wp-content/uploads/Git-blog-icon2.png 640w" sizes="(max-width: 240px) 100vw, 240px" /></p>
<p>With the new Sisense-Git integration, you can develop new capabilities, commit changes, submit PRs, and revert changes &#8211; exactly the same way you do with the rest of your software. Not to mention get complete observability over who changed what across all your Sisense assets. No matter where you sit in the analytics development, or overall release process, there are some key benefits:</p>
<p>For example, if your job is building dashboards you now get the ability to test individual changes and integrated releases more easily and fully in safe development and staging environments and can push to production when ready.</p>
<div id="attachment_735984" style="width: 810px" class="wp-caption aligncenter"><img aria-describedby="caption-attachment-735984" decoding="async" loading="lazy" class="wp-image-735984" src="https://cdn.sisense.com/wp-content/uploads/Git-blog-1.png" alt="Selecting project assets in Sisense Fusion" width="800" height="512" srcset="https://cdn.sisense.com/wp-content/uploads/Git-blog-1.png 1000w, https://cdn.sisense.com/wp-content/uploads/Git-blog-1-300x192.png 300w" sizes="(max-width: 800px) 100vw, 800px" /><p id="caption-attachment-735984" class="wp-caption-text">Illustration: Selecting project assets in Sisense Fusion</p></div>
<p>And if you’re a release manager, you can now ensure that Sisense plugs into your automation and release process, and you can standardize processes across your environment.</p>
<p>Finally, for software developers, you can develop and embed analytics that are built in Sisense (not just in your code) while checking in updates and managing changes in Git, rather than using proprietary source control management and versioning.</p>
<div id="attachment_735985" style="width: 810px" class="wp-caption aligncenter"><img aria-describedby="caption-attachment-735985" decoding="async" loading="lazy" class="wp-image-735985" src="https://cdn.sisense.com/wp-content/uploads/Git-blog-2.png" alt="Managing sources, diffs and commits in Sisense Fusion" width="800" height="513" srcset="https://cdn.sisense.com/wp-content/uploads/Git-blog-2.png 1000w, https://cdn.sisense.com/wp-content/uploads/Git-blog-2-300x192.png 300w" sizes="(max-width: 800px) 100vw, 800px" /><p id="caption-attachment-735985" class="wp-caption-text">Illustration: Managing sources, diffs and commits in Sisense Fusion</p></div>
<p>At Sisense, we think this new Git integration finally brings analytics into alignment with enterprise and software release processes, providing high velocity, manageable analytics development. I’m excited about how this will enable organizations to create more impactful data experiences for users.</p>


Why strong source control and version management is now critical for analytics success

curve-image-unique-image-unique