I talk to a lot of people about analytics and insights, and most people I speak to already have a BI tool that they use, typically Power BI, Tableau or Qlik. Invariably at some point in our conversation they ask but how isThoughtSpot different to my existing platform. It's an important question, because why would someone consider changing if the platforms are essentially the same? This is why in general an organisation will typically have only one of the PowerBI, Tableau or Qlik in use and not multiple tools. I’ll run through the key differences that I see in ThoughtSpot, and the reason I left my dream job as the Director of Workforce Analytics to lead Harper Stone and to become a ThoughtSpot Premier Partner.
Built for business users
When you first see ThoughtSpot, you will notice that the User Interface is very different to other BI tools. It is more akin to Google or YouTube in terms of its cleanness and simplicity. Just like Google, the prominent search bar is the intended way to interact with ThoughtSpot. Users have a strong level of comfort asking questions through search. With my experience working with clients in less than a 1 hour session, business users have enough confidence to start using ThoughtSpot by themselves to ask their own questions and get the immediate data insights they need to make business decisions.
The simplicity of ThoughtSpot is really a standout and the speed at which it can surface answers across massive datasets is remarkable. This really removes so many barriers for business users. There is no longer a delay in getting an answer where they must wait for an analyst to produce a report or dashboard. ThoughtSport encourages exploration. All charts in ThoughtSpot are interactive meaning that a user can right click on any point and drill into the data by any dimension, there are no predefined drill paths or limitations, if the data exists then the user can drill down.
Built for scale
For the last 15+ years the leading BI platforms have been Tableau, Qlik and Power BI. These stalwarts lifted BI out of the depths of IT and brought it to the business-masses and powered a data-driven revolution. Being developed in the 90’s, these tools have been architected for data scales of a different time. To work around these limitations, extensive effort is required to reduce the dataset size through the use of cubes and aggregations.
ThoughtSpot as a company was established in 2012, with the goal of bringing the simplicity of Google-like search to BI, regardless of the size of the dataset. ThoughtSpot have built an in-memory calculation engine that utilises massively parallel processing, query performance optimiser, and data sharding and replication to enable ThoughtSpot to handle transaction level data for datasets in the size of TB’s and return search-like performance to queries.
ThoughtSpot Embrace was launched with v6 in January 2020. Embrace is the live query interface that allows the ThoughtSpot BI and Visualisation Engine to live query the leading Cloud Data Warehouses. At the time of writing, Embrace supported live query to SnowFlake, Amazon RedShift, Google Big Query, Microsoft Synapse, SAP Hana, and Terradata, with more CDW’s coming online with each new ThoughtSpot release. The advantage of Embrace is the complete elimination of all data prep and migration activities as well as a significant saving in storage costs as only one copy of the data is stored in the CDW.
The questions you didn’t ask
SpotIQ is the AI engine within ThoughtSpot that is constantly scanning data to surface insights, the answers to questions you didn’t think to ask. SpotIQ can also be run by a user against a specific set of data. SpotIQ identifies outliers, trends, and correlations within the data and provides these as charts with simple descriptions that business users can understand.
SpotIQ can also be a very effective productivity enhancer for analysts exploring a new dataset. Identifying potential areas for further analysis.
Isn’t it the same as Tableau Ask Data, Qlik Bot, PowerBI Q&A …
These are all search functionality that have been bolted on to an existing platform. ThoughtSpot has been built from the ground up to search across data, it’s a relational search engine. These bolt-on searches can only provide answers to simple questions and can only search across pre-built content or simple datasets. They do not scale and cannot search across large datasets. It’s sort of like comparing Google to that weird paperclip that was part of Windows back in the 90’s.
Data curators not report monkeys
In many organisations, analysts have become report monkeys. They are on a treadmill of servicing ad-hoc requests, creating just one more dashboard to answer a slightly different question. With ThoughtSpot the role of the analyst shifts from report monkey to data curator. As the curator they are focused on preparing data for search rather than content (dashboard/report) creation. Preparing data for search involves the configuration of metadata that simplifies search for users. Working with Embrace, it is possible to get search up and running on complex datasets in hours, seriously.
One of the biggest and most common blockers to BI adoption and self-service is the licence model. Paying $2,000pa for a licence for a business user who ends up consuming pre-built dashboards is just not an acceptable ROI for many businesses, so to keep a cap on costs we limit our user licences.
With ThoughtSpot the value is in the data. So the licence model is based around data size - ie number of rows being searched. We don’t care how many users are accessing ThoughtSpot, in fact, the more the merrier. In this way, ThoughtSpot doesn’t become a victim of its own success, where so many people find value in the platform that it becomes too expensive and we limit access.
So it’s the panacea?
Well while that would be great, no it’s not the solution for everyone and every case. PowerBI, Tableau, and Qlik have been built for analysts and they are mature products with extensive capabilities, and can produce pixel perfect visualisations. But if you are looking for true self-service for business users, or if you have very large datasets that have hit the performance limits of Tableau, Power BI or Qlik, then ThoughtSpot is worth a look.