Back to List

IT-Managed and Self-Service Analytics: Best Practices

Scott Hietpas Scott Hietpas  |  
Aug 27, 2019
 
In this blog series, Scott Hietpas, a principal consultant with Skyline Technologies’ data team, explores the advantages and disadvantages of different analytics management approaches. For a full overview on this topic, check out the original IT-Managed vs Self-Service Analytics Webinar.
 
In the first two blogs in this series, we looked at the advantages and disadvantages of IT-Managed and Self-Service analytics. At a high level, IT-Managed analytics offers the advantage of a quality, single source of truth managed by a highly-skilled team that delivers experienced industry best-practices. However, Self-Service analytics has more scalability, offers direct stakeholder ownership, and allows more business area innovation. Ultimately, the best solution for many organizations is a balanced approach between these two options.
 

IT-Managed and Self-Service Analytics

Each of these approaches involve different architectures. An enterprise solution typically has much more layering of the architecture while Self-Service is more of a direct consumption of the data.
 
To get the most out of your data with a combined solution, you need to decide upfront what will fall in the realm of IT-Managed and what falls under Self-Service. I recommend that mission-critical items (like KPIs that your organization relies on, or that you measure yourself on) are often best still managed by IT. When accuracy matters and when it's critical, it makes sense to rely on a single source of truth provided by IT.
 
In cases where you want to focus more on innovation and there are many opportunities to explore, Self-Service can really fill a gap. Self-Service can be a great benefit to an organization if you don't need that same degree of accuracy, or the ability to get answers quickly is a higher priority.
 

Power BI Dataflows and the Common Data Model

Additionally, with innovations like Power BI Dataflows, we're starting to see where we can utilize the same architecture approach with Self-Service and IT-Managed. In this video, Microsoft shares how Power BI dataflows can be utilized from a Self-Service perspective within a full enterprise-scalable solution.
 
Power BI Dataflows allow us to store data in Azure Data Lake Storage Gen2, and we can use the Power Query capabilities within Power BI to ingest that data and build datasets and reports on it. Now, IT can define datasets and reports in that same space, Self-Service can find data models or datasets that IT created, and business users can also find datasets that other business users have created. This makes it a much more collaborative space for not just building the reports but also for building those datasets.
 
Another key component of this architecture is the use of the common data model. This allows us to put some metadata definition around the data lake or Azure Data Storage so it's more consumable by Power BI and other applications that implement the common data service.
 

Combined Solutions for Big Results

What started out as a big difference between how an IT-Managed solution might look and how Self-Service solutions might look doesn't necessarily have to be an “or” architecture. It can be an “and”. The same architecture approach could be leveraged by Self-Service and the enterprise. More solutions going forward are going to be a combination of both IT-Enabled (or IT-Managed) and Self-Service. Instead of trying to decide which one applies, the key is understanding how we can be successful in both of those areas.
 

How Can a User Tell the Difference Between Self-Service and IT-Managed Analytics Content?

Differentiating between Self-Service and IT-Managed can be a challenge. With Power BI, it's on the roadmap to be able to identify certain content or apps as certified.
 
Eventually, I look forward to a future where we can more officially designate content as being certified. That identification would extend all the way to datasets within the actual data catalog. That official IT-certified stamp would give the business users confidence that they can use either that data or report to further analysis.
 
However, that functionality hasn’t arrived at this time. In the meantime, the way we often manage differentiation between Self-Service and IT-Managed analysis is by having separate apps. In our case, even though we can't officially put a certified stamp on it, we do have Skyline-certified apps. Certain apps within the Power BI workspace are guaranteed to be accurate by our IT team. We're the ones responsible for creating and maintaining that content and ensuring all the calculations are right. We can (and do) put our logo on those apps so users know which apps are provided by our team.
 
In Self-Service, users can choose whatever logo and content they want. In our Community of Excellence (which I recommended in the previous blog to share knowledge within your organization), we also list our certified apps. We specifically say which apps are ours, and we tried to follow a naming standard: using the word “Insights” that indicates that an app is part of our official collection (for example, Project Insights and Practice Insights).
 

What Are the Pros of a Sandbox Environment for IT-Managed and Self-Service Analytics?

Another best practices question that we’ve received in the past is how using a sandbox environment (for single-use questions or exploration of content) fits into IT-Managed and Self-Service. We've worked with the sandbox concept with several clients, and there are a few good reasons for using one.
 
To start, one of the key causes of that duplication in the Self-Service world is often that people aren't working in a shared space. I really like the idea of creating sandboxes for experimentation and to increase collaboration. What you originally thought was a one-time question might be relevant longer-term and may be relevant to other areas, which you can’t know if you are in your own bubble.
 
I really like the idea of exploring that in a shared or collaborative sandbox shared by a department. If you have a smaller organization, you may just need a single sandbox. A sandbox doesn't need to necessarily fall into the IT-Managed or Self-Service world: either would be a viable option.
 
So then, how do you go from what started as a sandbox experiment or prototyping and move that into a final product? Content in the sandbox could be directly shared with a small group for feedback or deployed in its own sandbox app. Once content reaches a certain threshold of significance, you may consider moving the content into a certified dataset or app. This may involve an IT review process and hardening of the model. The goal would be that the sandbox is never the end point for content that has proven value. It is intended to be an incubator where business users or IT can collaborate to refine analytical insights.
 
Analytics

 

Love our Blogs?

Sign up to get notified of new Skyline posts.

 


Related Content


Blog Article
How to Handle Refactoring in a Self-Service Analytics World
Scott HietpasScott Hietpas  |  
Sep 10, 2019
In this blog series, Scott Hietpas, a Principal Consultant with Skyline Technologies’ data team, and Matt Pluster, Data Analytics and Data Platform Team Director, explore the advantages and disadvantages of different analytics management approaches. For a full overview on this topic, check...
Blog Article
Self-Service Analytics: Pros and Cons
Scott HietpasScott Hietpas  |  
Aug 13, 2019
In this blog series, Scott Hietpas, a principal consultant with Skyline Technologies’ data team, explores the advantages and disadvantages of different analytics management approaches.  For a full overview on this topic, check out the original IT-Managed vs Self-Service Analytics...
Blog Article
IT-Managed Analytics: Pros and Cons
Scott HietpasScott Hietpas  |  
Jul 30, 2019
In this blog series, Scott Hietpas, a principal consultant with Skyline Technologies’ data team, explores the advantages and disadvantages of different analytics management approaches. For a full overview on this topic, check out the original IT-Managed vs Self-Service Analytics Webinar...
Blog Article
Data Lake vs Data Warehouse: Avoiding the Data Swamp
Scott HietpasScott Hietpas  |  
Jul 16, 2019
In this blog series, Scott Hietpas, a principal consultant with Skyline Technologies’ data team, and Matt Pluster, the Data Analytics and Data Platform director, respond to some common questions on data warehouses and data lakes. Previous installments covered pros and cons of each solution,...
Blog Article
Data Lake vs Data Warehouse: What’s the Best Configuration?
Scott HietpasScott Hietpas  |  
Jul 02, 2019
In this blog series, Scott Hietpas, a principal consultant with Skyline Technologies’ data team, responds to some common questions on data warehouses and data lakes. For a full overview on this topic, check out the original Data Lake vs Data Warehouse webinar.   In my previous blog...