Sisense is one company gathering new momentum in the self-service business intelligence (BI) space.
In September 2018, the company announced a new $80 million investment from New York-based venture capital (VC) firm Insight Venture Partners.
If you are familiar with BI tools, then you'll likely be impressed with Sisense (which is priced only by custom quote).
It's an attractive product with substantial power.
Still, Sisense lacks the brand recognition of other BI heavyweights such as IBM Watson Analytics and Microsoft Power BI.
But with its intuitive user interface (UI) and the significant depth of its data visualization capabilities, Sisense is seriously worth consideration.
While its UI and commands are nowhere near as familiar as those of Microsoft Power BI, it is a serious threat to Tableau Desktop given its top-shelf functionalities, such as in-chip rather than in-memory processing, and natural language commands and queries you can use inside third-party apps such as Microsoft Skype and Slack.
Seriously, you can ask a question in Skype and Sisense will answer you in Skype.
That's enough to make even IBM Watson Analytics sit up and take notice.
No worries yet, Watson, as not everything in Sisense supports natural language, which is part of the reason it isn't one of our Editors' Choices.
On the downside, Sisense is still a bit too clunky to be ready for prime time in a fully data-democratized organization where you want people using data in their job decisions regardless of their skill level in data science or statistics.
You know, like everyone in any given organization can use Microsoft Word ($128.00 at Amazon) without having to know how to write code or even how to spell correctly.
True, being able to access Sisense's analytics simply by popping a natural language query into a third-party app goes a long way toward making the platform universally useful.
However, the rest of the platform's UI is still necessary and it just doesn't match that level of user-friendliness needed to satisfy users who aren't data knowledgeable.
Still, the company is working on this weakness and does a creditable job providing online training and learnings with a well-organized support section and a well-maintained blog.
Even so, this is a fine—dare I say, badass—app that average and highly skilled business analysts are sure to appreciate.
It lifts most of the burden from skilled staff, without having to buy additional tools.
It's a full-stack tool so there's less dependency on IT or highly skilled resources.
Sisense also plays nice with other analytics and apps, which explains why the company makes half of its revenue from embedded use in other products.
Sisense has yet to reach critical mass in the market, but it's likely to hit that milestone soon.
Meanwhile, the company is mum on prices so you have to ask them for a quote.
That's a drawback, too, considering it keeps touting a low total cost of ownership (TCO).
It's hard to do the math on that claim without knowing the price first.
Getting Started
Think of Sisense as consisting of two parts: There's the intuitive web interface and then there is ElastiCube, Sisense's proprietary analytical database.
ElastiCube must be downloaded and run locally, something I didn't have to do with other players.
After the download, I went to the Windows Start menu and opened Sisense ElastiCube Manager.
If you want to do the tutorial first with sample data already in the system, then select File > New ElastiCube File, and name the file "tutorial," "testing," "messing around," or something that will later mean that this isn't the file you need for anything else.
Then, follow the prompts to dip your toe in before you dive into the deep end of the pool.
Having enough familiarity with data science, I jumped straight into the deep end.
I did watch the tutorials later, and they are well-done and easy to follow.
It's smarter to watch those first as the UI isn't quite as intuitive as it should be and something of a disappointment after all of the natural language goodness.
Anyway, there I was with Sisense open in my browser and ElastiCube Manager open on my desktop.
I went straight for "Open File" on ElastiCube.
Nope, that's not the path to my data apparently.
It brought up local files but wouldn't let me open my Comma Separated Values (CSV) files.
Next, I clicked on "Connecting to Data" and that took me to a guide listing the connectors, of which there are plenty.
There I learned the CSV connector is one of several that are preinstalled.
One more click on "Working with Data," and there was a prompt pointing to a "+" button where I could upload my local CSV data.
I'll call that three click bumbles and no serious fumbles, which means that, if you are an experienced business analyst, then exploring the system sans tutorial isn't hard.
But if you aren't, then you'll likely find yourself totally lost, fast.
There's a significant learning curve here so watch the tutorials and take notes.
But in short, click on Add Data, select your data sources, and enter your login credentials as needed to connect.
All available tables are presented in each data base, and you then select the ones you want to use.
You can preview and mashup multiple data sources before adding to your schema.
Creating a join is done by drag-and-drop.
Large data sets can be combined in a single cube.
After that, I could analyze the data and create dashboards using the web interface.
All of that sounds easy and it is if you've worked with BI apps before, but not so much if this is your first foray into working with data.
You can pull in data from multiple data sources, including eBay, Facebook, QuickBooks, and PayPal.
It also integrates with cloud storage platforms like Box.
In addition, you can incorporate data from Database-as-a-Service (DBaaS) platforms like Google BigQuery.
Once the data was loaded, I followed the prompt to the Build command where I could configure and build an ElastiCube.
There I was presented with two options: Build Schema Changes and Build Entire ElastiCube.
Once again, first timers and lay users might stall out and become confused.
I chose the first option, and the system took about four minutes to complete the build and import process.
Then I was ready to design a dashboard and run any number of ad hoc analytics.
I also now had an ElastiCube file on my desktop.
The Discovery Process
Click on "Dashboard" in ElastiCube Manager and you're automatically moved to Sisense's web interface in your browser.
Alternately, at a later date, I had no need to open ElastiCube Manager first.
I simply went to the web interface to use data already in ElastiCube.
Once on, I was asked to select a data set (from those already in ElastiCube), aka a Cube.
I could also give the new dashboard a name here before I click Create.
Under the subheading "Widget" on the next page, it asked me to select data again.
But this time it didn't mean for me to choose an entire data set as I'd already done that on the prior page.
Rather, it meant for me to choose fields from the tables in the selected data sets.
If you select fields from different tables that you haven't already joined, then you'll get an error message—yet another point in the process at which business users may stumble.
There is a "Try Again" command but that doesn't do anything if the tables aren't joined.
I popped back over to ElastiCube and joined the tables there.
From my Cube, I chose the fields "brand" and "device" and then a visualization: treemap.
Clicked the "Create" button, typed a title into the title bar, added a couple of filters, and voila: I had an interactive visualization to explore.
If you're a regular Tableau Desktop (Visit Store at Tableau) user, then you're going to think this process is cool and super-efficient.
If you're more of a "Tell me like it is" user of IBM Watson Analytics, then it's going to take you awhile to understand enough about this app to truly appreciate it.
Data Visualizations
Unlike other self-service BI apps, the value of visualizations in Sisense is not in the number of designs and formats from which you can choose but in the depth of insights they expose.
In short, the multidimensional widgets render interactive, "drill to anywhere" visualizations that provide lots of insights by simply scrolling the mouse over them or clicking on different sections.
Sisense also lets users reposition and resize visualizations on dashboards prior to sharing so that they are easier to read in either email or feed modes, making it ideal for viewing on multiple devices.
I would argue, however, that the available depth of its visualization might be the most significant differentiator for Sisense.
These mean that an analyst or a lay user could easily discover more than they originally expected from any given analysis, without any further effort.
But for this benefit to be fully realized in a data democratized organization, Sisense first needs to make it easier for lay users to get to this point.
Cons
Perhaps a bit complex for a self-service business intelligence (BI) tool.
Analytics process needs work.
Natural language features have limitations.
The Bottom Line
Sisense will easily appeal to seasoned BI users with its comprehensive features, but it may frustrate novice users.
Sisense is one company gathering new momentum in the self-service business intelligence (BI) space.
In September 2018, the company announced a new $80 million investment from New York-based venture capital (VC) firm Insight Venture Partners.
If you are familiar with BI tools, then you'll likely be impressed with Sisense (which is priced only by custom quote).
It's an attractive product with substantial power.
Still, Sisense lacks the brand recognition of other BI heavyweights such as IBM Watson Analytics and Microsoft Power BI.
But with its intuitive user interface (UI) and the significant depth of its data visualization capabilities, Sisense is seriously worth consideration.
While its UI and commands are nowhere near as familiar as those of Microsoft Power BI, it is a serious threat to Tableau Desktop given its top-shelf functionalities, such as in-chip rather than in-memory processing, and natural language commands and queries you can use inside third-party apps such as Microsoft Skype and Slack.
Seriously, you can ask a question in Skype and Sisense will answer you in Skype.
That's enough to make even IBM Watson Analytics sit up and take notice.
No worries yet, Watson, as not everything in Sisense supports natural language, which is part of the reason it isn't one of our Editors' Choices.
On the downside, Sisense is still a bit too clunky to be ready for prime time in a fully data-democratized organization where you want people using data in their job decisions regardless of their skill level in data science or statistics.
You know, like everyone in any given organization can use Microsoft Word ($128.00 at Amazon) without having to know how to write code or even how to spell correctly.
True, being able to access Sisense's analytics simply by popping a natural language query into a third-party app goes a long way toward making the platform universally useful.
However, the rest of the platform's UI is still necessary and it just doesn't match that level of user-friendliness needed to satisfy users who aren't data knowledgeable.
Still, the company is working on this weakness and does a creditable job providing online training and learnings with a well-organized support section and a well-maintained blog.
Even so, this is a fine—dare I say, badass—app that average and highly skilled business analysts are sure to appreciate.
It lifts most of the burden from skilled staff, without having to buy additional tools.
It's a full-stack tool so there's less dependency on IT or highly skilled resources.
Sisense also plays nice with other analytics and apps, which explains why the company makes half of its revenue from embedded use in other products.
Sisense has yet to reach critical mass in the market, but it's likely to hit that milestone soon.
Meanwhile, the company is mum on prices so you have to ask them for a quote.
That's a drawback, too, considering it keeps touting a low total cost of ownership (TCO).
It's hard to do the math on that claim without knowing the price first.
Getting Started
Think of Sisense as consisting of two parts: There's the intuitive web interface and then there is ElastiCube, Sisense's proprietary analytical database.
ElastiCube must be downloaded and run locally, something I didn't have to do with other players.
After the download, I went to the Windows Start menu and opened Sisense ElastiCube Manager.
If you want to do the tutorial first with sample data already in the system, then select File > New ElastiCube File, and name the file "tutorial," "testing," "messing around," or something that will later mean that this isn't the file you need for anything else.
Then, follow the prompts to dip your toe in before you dive into the deep end of the pool.
Having enough familiarity with data science, I jumped straight into the deep end.
I did watch the tutorials later, and they are well-done and easy to follow.
It's smarter to watch those first as the UI isn't quite as intuitive as it should be and something of a disappointment after all of the natural language goodness.
Anyway, there I was with Sisense open in my browser and ElastiCube Manager open on my desktop.
I went straight for "Open File" on ElastiCube.
Nope, that's not the path to my data apparently.
It brought up local files but wouldn't let me open my Comma Separated Values (CSV) files.
Next, I clicked on "Connecting to Data" and that took me to a guide listing the connectors, of which there are plenty.
There I learned the CSV connector is one of several that are preinstalled.
One more click on "Working with Data," and there was a prompt pointing to a "+" button where I could upload my local CSV data.
I'll call that three click bumbles and no serious fumbles, which means that, if you are an experienced business analyst, then exploring the system sans tutorial isn't hard.
But if you aren't, then you'll likely find yourself totally lost, fast.
There's a significant learning curve here so watch the tutorials and take notes.
But in short, click on Add Data, select your data sources, and enter your login credentials as needed to connect.
All available tables are presented in each data base, and you then select the ones you want to use.
You can preview and mashup multiple data sources before adding to your schema.
Creating a join is done by drag-and-drop.
Large data sets can be combined in a single cube.
After that, I could analyze the data and create dashboards using the web interface.
All of that sounds easy and it is if you've worked with BI apps before, but not so much if this is your first foray into working with data.
You can pull in data from multiple data sources, including eBay, Facebook, QuickBooks, and PayPal.
It also integrates with cloud storage platforms like Box.
In addition, you can incorporate data from Database-as-a-Service (DBaaS) platforms like Google BigQuery.
Once the data was loaded, I followed the prompt to the Build command where I could configure and build an ElastiCube.
There I was presented with two options: Build Schema Changes and Build Entire ElastiCube.
Once again, first timers and lay users might stall out and become confused.
I chose the first option, and the system took about four minutes to complete the build and import process.
Then I was ready to design a dashboard and run any number of ad hoc analytics.
I also now had an ElastiCube file on my desktop.
The Discovery Process
Click on "Dashboard" in ElastiCube Manager and you're automatically moved to Sisense's web interface in your browser.
Alternately, at a later date, I had no need to open ElastiCube Manager first.
I simply went to the web interface to use data already in ElastiCube.
Once on, I was asked to select a data set (from those already in ElastiCube), aka a Cube.
I could also give the new dashboard a name here before I click Create.
Under the subheading "Widget" on the next page, it asked me to select data again.
But this time it didn't mean for me to choose an entire data set as I'd already done that on the prior page.
Rather, it meant for me to choose fields from the tables in the selected data sets.
If you select fields from different tables that you haven't already joined, then you'll get an error message—yet another point in the process at which business users may stumble.
There is a "Try Again" command but that doesn't do anything if the tables aren't joined.
I popped back over to ElastiCube and joined the tables there.
From my Cube, I chose the fields "brand" and "device" and then a visualization: treemap.
Clicked the "Create" button, typed a title into the title bar, added a couple of filters, and voila: I had an interactive visualization to explore.
If you're a regular Tableau Desktop (Visit Store at Tableau) user, then you're going to think this process is cool and super-efficient.
If you're more of a "Tell me like it is" user of IBM Watson Analytics, then it's going to take you awhile to understand enough about this app to truly appreciate it.
Data Visualizations
Unlike other self-service BI apps, the value of visualizations in Sisense is not in the number of designs and formats from which you can choose but in the depth of insights they expose.
In short, the multidimensional widgets render interactive, "drill to anywhere" visualizations that provide lots of insights by simply scrolling the mouse over them or clicking on different sections.
Sisense also lets users reposition and resize visualizations on dashboards prior to sharing so that they are easier to read in either email or feed modes, making it ideal for viewing on multiple devices.
I would argue, however, that the available depth of its visualization might be the most significant differentiator for Sisense.
These mean that an analyst or a lay user could easily discover more than they originally expected from any given analysis, without any further effort.
But for this benefit to be fully realized in a data democratized organization, Sisense first needs to make it easier for lay users to get to this point.
Cons
Perhaps a bit complex for a self-service business intelligence (BI) tool.
Analytics process needs work.
Natural language features have limitations.
The Bottom Line
Sisense will easily appeal to seasoned BI users with its comprehensive features, but it may frustrate novice users.
