End to end BI solution for not complex scenarios
Updated February 12, 2019

End to end BI solution for not complex scenarios

Anonymous | TrustRadius Reviewer
Score 3 out of 10
Vetted Review
Verified User
Review Source

Overall Satisfaction with Birst

We adopted BIRST to be used as the enterprise BI solution so all departments with enterprise BI needs would use BIRST.

We use BIRST Cloud solution.

  • Being a cloud solution the platform upgrades are not a pain point anymore.
  • Once the data model is ready, building basic reports/dashboards are really straightforward.
  • One single tool may cover all your business intelligence needs, from data ingestion to reporting so it can be cost-saving in terms of licensing.
  • Building an analytical ready model does not require highly skilled developers.
  • Lack of access to underlying database and use of non standard languages (propietary code) BQL and BIRST scripting.
  • Very poor and not user-friendly administrative area. Big dependency on command line tasks (a few drag and drop) and lack of out of the box administrative views (reports and dashboards showing platform's response).
  • Small community. Really difficult to find skilled resources in the market and few technical blog articles or videos available.
  • Data ingestion capabilities do not scale. Being capable of ingesting data is an added value however, the solution is mainly based on SQL plus scripting (no out of the box monitoring capabilities, difficult to reverse-engineer...).
  • Speed and agility
  • Other
Speed and agility: For non complex scenarios we have increased the speed by using BIRST. However, whenever the complexity increases, the time required grows exponentially. Additionally, we have been able to build an ODS layer which enables some reusability. Reusability of the data by third-party tools is really limited, there aren't data federation capabilities within the analytical layer and we have found it difficult to use networked BI (some limitations have been identified).
Other: There are things that do not work as expected such as "Networked BI" (re-usability of data across different subject areas). There are many limitations in terms of what we can monitor on the solution's behavior (usage tracking data is not complete neither accurate). It is really difficult to reverse engineer the solution developed by another person and not having direct SQL access to the underlying database prevents us from having the right level of data-traceability/data-lineage.
The response time when cases are created does not meet our expectations.
Tableau does not have a common semantic layer like BIRST does so we use Tableau for data discovery purposes and BIRST for enterprise BI.
Within BIRST, we can increase the level of data governance.
BIRST is capable of ingesting data from many different data sources. Regarding ETL capabilities, it can do more than any other BI tool I'm aware of. But if you do really have data integration needs you better have a data integration tool for that purpose.
Some "modern"connectors do not fully work and they are not just "plug and play".
There are connectors in ALPHA version for more than a year (i.e. BOX).

Depending on the complexity of your company's data and business needs it may or may not be a good fit. I would say it can be a good BI ecosystem for small and mid-size companies with low complexity so they can have data ingestion and reporting capabilities with one single tool.

If you need to integrate data from many different data source;, your environment does change quite frequently, you have many different business areas, the data volume you have is quite big... then it is less appropriate.

Birst Feature Ratings

Pixel Perfect reports
Customizable dashboards
Report Formatting Templates
Drill-down analysis
Formatting capabilities
Integration with R or other statistical packages
Not Rated
Report sharing and collaboration
Publish to Web
Publish to PDF
Report Delivery Scheduling
Pre-built visualization formats (heatmaps, scatter plots etc.)
Location Analytics / Geographic Visualization
Predictive Analytics
Not Rated
Multi-User Support (named login)
Role-Based Security Model
Multiple Access Permission Levels (Create, Read, Delete)
Responsive Design for Web Access
Dedicated iOS Application
Dedicated Android Application
Dashboard / Report / Visualization Interactivity on Mobile

Additional Questions on Birst Capabilities

From the end user point of view (both interacting with dashboards already built or creating your own reports with Visualizer as self service BI) it is user friendly.
From the admin/developer point of view we do not find it user friendly, a lot of commands for administrative tasks (few drag and drop), a lot custom code required for data ingestion purposes including Python, batch command line, BQL (BIRST SQL language, it is difficult to reverse engineer, there is a lack of a super-administrator role which could have access to everything.
Personally I would say it could be a good fit for mid-size companies with low complexity (few data sources, few STAR-schemas and few conformed dimensions).
There are 2 points which have prevented us from being really agile with BIRST: difficulty of reverse engineer what has been built by another team/developer and lack of skilled resources in the market.
Additionally, there are capabilities which BIRST does come with: ADR (Automated Data Refinement), Networked BI and Data Mashups which should have helped on bringing additional speed and increased re-usability but we have found many limitations around those.
We do use BIRST Mobile to access reports and dashboards.
BIRST via web browser does have "Admin", "Visualizer" and "Designer" tabs. Accessing via mobile app the capabilities are limited so only dashboards are available.
When designing dashboards for mobile it would help if on the development area we could preview how the dashboard will look in a mobile device.
The common semantic layer of BIRST (like other enterprise oriented BI tools in the market) is something you do not value until you miss it. We do make use of it however the learning curve is not really fast. It is difficult to reverse engineer as there are many places where things can be changed.
Lineage is quite straightforward as you can select an object and its dependencies however the graphical part gets really messy when you have a complex model behind.
One really important thing BIRST should put a focus on is moving away from Flash and fully adopt HTML5. The administrative/development area still requires flash.
BIRST does release new features and bug-fixes every three months which quite enough.
My guess is that the biggest effort BIRST is doing is moving away from Flash and adoption HTML5, in this area they are behind other vendors.
Regarding new capabilities, the main focus lately is on the self service data preparation area with PRONTO which is quite aligned with what other vendors are doing.
What we are missing are capabilities other vendors are announcing like natural language processing or a bit of artificial intelligence. We do not anticipate much progress on this areas.

Using Birst

7000 - We do follow a centralized/de-centralized approach around BI with BIRST covering the data-warehousing and semantic layer needs and another BI tool for data discovery and piloting purposes.
Current usage of BIRST in our company is not focused on a specific business area or function but all areas with enterprise BI needs.
50 - We use BIRST as SaaS so in-house we do not support the platform/technology but the applications built on top of it.
The support team we have is basically granting access to users, and supporting the different applications including: monitor the execution of ETLs, building new reports/dashbords, troubleshooting issues...
  • Enterprise wide reporting with big target audience having a common semantic layer in order to ensure KPI definitions are consistent
  • No innovative ways we have used BIRST
  • Depending on BIRST new capabilities. Currently we're trying to adopt Pronto for self service data preparation purposes
  • Product Features
  • Third-party Reviews

Birst Support

There are tickets we've created which seem simple but not answered in over a year.
In many cases we are provided with workarounds but not real solutions.
Multiple times we are given the answer "code change" requested so we have to wait for the next release (every 3 months) and in some cases we have to wait for more than one release.
Slow Resolution
Poor followup
Less knowledgeable
Problems left unsolved
Escalation required
Yes - The tool has been selected as the Enterprise BI solution so having it 24/7 is a must
Yes - Multiple times we are given the answer "code change" requested so we have to wait for the next release (every 3 months) and in some cases we have to wait for more than one release.
There are other cases where a minor release has been provided (not having to wait for 3 months) but this is not the usual case.

Using Birst

Unnecessarily complex
Requires technical support
Not well integrated
Slow to learn
Feel nervous using
Lots to learn
  • Self service BI to build simple reports with Visualizer is easy
  • Interacting with existing reports and dashboards is really easy
  • Data ingestion process is quite complex
  • Data modelling and making changes to the existing model can be really tricky. It is hard to reverse engineer and apply changes on a model built by another developer/team and it is hard to change something if you do not regularly apply changes (i.e. add a dimension to a model you built 5 months ago).
  • The administration parts are not user friendly. It is Flash based and a lot of things are done via commands. There is no super administrator role so not possible to have a look at everything which has been built.

Evaluating Birst and Competitors

BIRST was the innovative and disruptive technology of
the BI Industry in 2015 and we were looking for a SaaS cloud solution what we did not consider is the high complexity of our data-ingestion/integration needs neither the low re-usability of the data ingested by BIRST.
During the selection process make sure the tool does fully cover the capabilities needed and clearly position it within your toolkit so each tool is used for the purpose it was selected.