Note: This post, originally published in 2016, was most recently updated in November 2017.
If you’re evaluating BI tools right now, Looker has to be at the top of your list. The company is absolutely killing it: it’s raised $96mm in the past four years and has built a customer list that includes the most data-savvy companies in the world.
The product isn’t perfect (what product is?) but it’s incredibly compelling for one single reason: its data modeling layer, LookML, is a powerful data modeling layer that gives analysts tools to build sophisticated analytics, which can then be pulled into point-and-click reports by business users. This combination has been the holy grail of BI for 20+ years and Looker accomplishes it better than any other product, ever. The entire company is built on top of this fundamental innovation.
If you a) want to empower your analysts to build sophisticated analytics for your business, and b) empower your business users to customize reports on an ad-hoc basis, Looker is absolutely your best bet.
Of course, you do have options. You could use GoodData, or Domo, or Qlik, or Cognos. If you decide to go down that road, please shoot me an email in two years and let me know how that worked out for you. In all seriousness, I’ve spoken to dozens of analysts and business users who have used GoodData, Domo, and Qlik and literally not a single one of them had anything positive to say about the experience. I’ve personally implemented Cognos before and I try hard to lock those memories deep down inside. These products are classic enterprise software that are built to appeal to CIOs and IT procurement managers, not users. Avoid them like the plague.
What other choices do I have?
Looker can’t be the only thing on the market, right? It’s not—there are actually several very compelling products on the market today.
There is another class of products that has a different take on BI that I affectionately call SQL Runners. These tools let you write SQL, then run that against a database, then allow you to build charts on top of the results and share them. Simple, but powerful. I like two products in this category: Mode Analytics and Periscope. If you’re an experienced SQL user then these tools are 👍
The primary difference between this category of tools and Looker is that Mode and Periscope don’t have a modeling layer. This means that, while you can build very sophisticated reports in them, business users are often left out in the cold. If you’re not a fairly advanced SQL user, Mode and Periscope will allow you to view the reports that your SQL-loving coworkers have built for you and to do light customizations on top of them, but you won’t be able to build reports yourself. This may be OK, or it may be intensely frustrating.
Tableau is in a category of one: there is nothing else quite like it. It gives business users more powerful data visualization options than any other product on the market, and sometimes that is exactly what is needed. However, Tableau isn’t really a “business intelligence” tool.
IMHO, BI tools must plug in to production data systems and conduct useful business analytics on top of them. Tableau, for all its impressive visualization capabilities, can’t really deal with production data: it’s drag-and-drop capabilities just don’t allow users to express the complicated business logic that is required in real-world BI.
For example: Tableau cannot read the Heap event stream from your mobile app in your Redshift cluster and create anything useful out of it. It also can’t navigate your database’s relational schema to analyze customers, orders, and items. Doing this type of analytics requires more sophisticated data processing than Tableau can handle.
If you have a CSV and want to make the best chart in the world, use Tableau. If you want to ask arbitrary questions about the status of or trends in your business, don’t.
When do I use Looker?
I’m the Founder and CEO of Fishtown Analytics, a consulting company that helps venture-funded businesses implement sophisticated analytics. I’ve also been working in analytics for over a decade and a half. When I choose what BI tool to use for my clients, it’s a very simple decision: all I need to know is “who’s building the reports?”
Sometimes, we’re the only people building reports on a given client engagement, and our deliverables are dashboards our users check on a periodic basis. In this case, we use Mode. Mode is easier, faster, and more powerful for experienced analysts.
But more frequently, companies want our help to empower their business users to build reports on their own. It’s very difficult to anticipate all of the views of the data that business users might want to see, and frequently the best answer is just to empower them to self-serve. This is when we turn to Looker. Looker provides us with an amazing way to build a data model that users can then navigate on their own, building tables and charts to their hearts’ content.
Ultimately, we want to help our clients build data-driven organizations, and it makes us incredibly happy to see them getting their hands dirty creating their own reports on top of the models we’ve built. Looker enables that workflow like no other tool, ever.
Is Looker the Right Business Intelligence Tool for Your Company?
Maybe! Ask yourself the exact same question that I ask myself: “who’s building the reports?” and go from there. If you’d like to chat about our experiences working with various BI tools or specifically with Looker, get in touch.nn
⚡️Ready to improve your analytics engineering workflow? Get started with dbt today. ⚡️
Last modified on: Apr 25, 2022