Updated February 20, 2018
Score 4 out of 10
Overall Satisfaction with Birst
Birst is being used in multiple departments across the organization. Zalora is an eCommerce company with data touching all the departments. Birst is now primarily used as a data extraction tool for operational purposes, hence it's our primary tool for self-service data needs. Birst is also used by several departments for simple ad-hoc analysis. Birst is primarily used by the fulfillment center to know item or product level KPIs. Birst Live data and dashboards were used for a long time to track marketing campaign sales and discounts. Until recently we shifted to other platforms as we needed more real-time analytics. Traditionally birst sufficed most of the business problems and use cases, however due to the technological transformations we had to pivot our platform to other alternatives.
- Birst is a cloud based BI platform and hence all our data is hosted on the cloud platform. Zalora had lot of data users who used to query the database, all these loads where shifted to birst and helped our in-house databases to function smoothly and concentrate on other critical processes.
- Birst report scheduling functionality has been very useful for our company. reports are scheduled to emails to several operational heads to help them monitor action items and help to improve our operational KPIs.
- Birst has been flexible with the number of users who can access the tool. Most of the BI platforms currently charges companies based on the users who would use the BI tool.
- Birst is useful in performing time based KPIs such as YoY, WoW, MoM, Qtr-o-Qtr etc. There are several functions available which helps to perform time based calculation.
- Birst is very useful in performing aggregations at different level (and different dates) as its based on star schema.
- As all the reports are in cloud, birst reports can be easily shared with one another and acts as a report repository
- Birst helps to maintain single source of truth and helps enforce standardized definitions for all our KPIs
- Visualizer report rendering is of high quality and quite fast for simple data requests and pull
- Birst provides snapshot functionality which is extensively used in our company to maintain the price logs and changes of all our merchandise.
- Birst being a cloud based platform is a both a blessing and a curse. as the data scales the performance of the tool deteriorates. although simple KPI reports can be quickly obtained, load intensive functions such as STAT or LET becomes hard to perform.
- The customer service is rather poor when the problems at hand are tough. The turnaround time for some of the bigger problems can go on for months without an alternative solution
- The processing time for each space is not consistent. outside US servers the processing time is very time consuming. Our APAC spaces in birst gets updated data by afternoon of a working day.
- The scheduled reports might not come in time if the server is busy. We have had instance in which the scheduled reports would come a whole 24 hours later than the scheduled time.
- Birst is a very difficult setup for change. If the data in the company is changing constantly then birst is not the tool to be held with. after implementing best practices, we were able to bring down the processing time of birst live to max 25 mins.
- When the size of the data scales, powerful functions such as STAT (bring data at a different level and aggregates in the report) and LET are hard to perform. We have had instance in which STAT would run for more than an hour to render reports
- Birst designer are still in FLASH and the vizualiser which is in HTML5 lacks several critical functions like scheduling, notifications, saving expressions and managing them. this leaves the customer to use the designer module instead of the vizualiser module, however the support for designer has been removed without keeping in mind that the visualizer component is incomplete.
- Plug n play analytics is hard to perform unless you have certain technical knowledge. Data in flat files or excel files cannot be analyzed or visualized instantly like it can be done in Tableau or any other BI Platform.
- External Inputs cannot be made on the report. External parameters and data cannot be blended with the existing report. For example, if we had to increase our discounts by x% how much would our sales be? these kind of questions are hard to perform.
- The syntax of each module in birst is different. This shows lack of team work within birst. for example. In designer module the ubiquitous if statement is denoted by IF and in Admin it is denoted by IIF. this would lead to end user being confused.
- Certain important critical functions such as Report bursting, querying sample data and parameterized reporting are available only to the admins and not the end users.
- Most of the code change requests in customer service tickets are still pending to date and some of them are unresolved.
- Birst has quite a bit of learning curve as it has its own syntax. one would need to be educated to understand the different types of expressions which exist in birst.
- Birst lacks search functionality in designer and searching in visualizer is rather tedious if you have several KPI named similarly
- The Pivot Control functionality is rather poor and pathetically mimics the pivot of excel
- Advance analytics is not possible. Different statistical models cannot be implemented. this would require data to be pumped out to perform advance analytics
- The Tableau-Birst connect is limited to 500GB, hence if the data scales the ODBC connector would not be useful
- The Data logs thrown out by birst needs to be more intutive and detailed. this would help new users to better understand the problems and troubleshoot it effectively
- Usage tracking of birst is tedious which would require a separate space to be created just to track users.
- Renewal costs are expensive given the price per BI Platform is coming down in recent years with more and more products coming online.
- Speed and agility
- Ability to meet corporate data standards such as scale and security
- Ability to do ad-hoc analysis and scheduled reports
- Birst fails to cope up with scaling data, when the data volumes are low the speed and agility are on par of any good analytical platform. The flash interface in designer makes it difficult to load data faster and the data limit makes it tedious to perform item based reporting or analytics.
- Birst provides good security features, it has well defined user groups which provide for perfect boundaries within the dataset
- Scheduling is robust
- Ad hoc analysis for simple reports is good, however when external inputs, statistical math is involved it performs poorly.
- The goal of self service data usage within the company was achieved considerably. BI Data request reduced and standardized KPIs where formed because of birst
- Snapshoting feature in Birst helped to analyze pricing of our products more accurately and faster.
- Birst lacks external modular components power bi has currently. New custom visuals, charts, statistical models created by third party vendors can be added to existing dashboard and enhanced.
- Birst lacks interactive dashboard functionality and external inputs.
- Powerbi and Tableau have better CGI and rendering
- Both Powerbi and Tableau when have good backend infrastructure are very scalable solutions
- Birst lacks multiple external source intergration in comparison to th
- Birst has its own backend and as a result the company need not worry about the back end infrastructure. both power bi and Tableau lacks backend.
- Birst is good at canned reporting and emailing compared to other products
- In comparison to other BI tools, Birst can connect only to a limited number of data sources (the most prominent ones)
- Only standard star schema can be built within birst making it highly de-normalized.
- Live access is buggy and slow
- No many to many relationship can be established because of the star schema
- Performance of report rendering decreases as data scales
- Birst has good data refinement practices which helps to maintain standard definitions.
Less appropriate scenarios
- Birst is best suited for canned report or standard MIS reporting.
- Email based reporting
- Birst can also be suited for management reporting, live tracking, alerts, and operations reporting (with a limited data set)
- Financial reporting based on time (with a limited data set)
- More employees need access to data
- Poor data warehouse infrastructure
Less appropriate scenarios
- For merchandising companies where the level of detail involves a lot of rows.
- Multiple data sources which increases load at backend
- International users outside US
- Decisions had to be made real time and faster
- More dependency on customer service
- Excel or flat file based analytics
- External Input is frequently required in reports
- Interactive dashboarding
- Advanced analytics involving model building.
Birst Feature Ratings
- Poor turnaround time
- Low response rate when the problem at hand are large
- Lack of infrastructural knowledge when it comes to scaling solutions
- Lack of inter team work noticed
- Most of the tickets are open for a long period of time
Kept well informed
Support understands my problem
Quick Initial Response
Problems left unsolved
- Business Intelligence is a crowded market with many tools competing in this space. Most of them offer free premium support and hence we expect Birst also to provide premium support to long standing customers.
- The scale of the problem is large and as a result it often takes long time to resolve it, We are not sure whether a premium support would help to resolve tickets quickly
- Reporting, User management and dashboarding based tickets are generally resolved quickly
- Backend, data architectural based tickets are not resolved quickly
- High priority bugs often require high officials to be involved to be resolved quickly.
One of our team members accidently deleted a space in birst which was critical for the business. There was no means of recovering the space from our end. The support was very helpful in restoring in the space for us from scratch. This helped to save a lot of time on rebuilding the reports, rebuilding the report schedules and dashboards altogether and also manage the users. Although the support was exceptional we were also charged a premium for this. Ideally a good BI tool should have multiple protocols in place when an important space needs to be deleted.
Requires technical support
Slow to learn
Lots to learn
- Time based functionality such as YoY,MoM are very easy to perform.
- Report scheduling for several reports based on parameters
- Fixed and Routine Dashboards and Reports
- Statistical Modelling and advance analytics
- Math based functions and other Excel similar Functions are unavailable
- No search functionality and Buggy Flash interface