Looker vs Tableau: What Data Communities Tell Us

Which one do you choose?

Another regurgitated blog post - treading the same water, Looker this, Tableau that - or a blog post sourcing opinions from real Looker and Tableau users?

If you're here, congrats - you made the right choice. Here's everything the data communities talk about, when we talk about Tableau vs Looker.

I. If you want self-service analytics, then Looker is a better choice.

The overwhelming consensus view is that Looker outshines Tableau when it comes to self-service analytics, thanks to Looker's data modeling layer.

Data modeling in Looker is done via Looker’s own modeling language called LookML. LookML enables technical users to define dimensions, metrics, aggregates, and relationships. When a user starts creating reports, Looker then constructs the SQL based on the defined LookML and queries the database. The results then appear as tables, charts, graphs, etc.

Looker vs Tableau: Self-service analytics

Having said that, most data practitioners also pointed out that Looker's price point is too hefty in comparison to Tableau, or most other BI tools.

So if you want a self-service BI tool with 80% of Looker's capacity but at 20% of Looker's price point - checkout Holistics. Both Holistics and Looker are architecturally similar:

Code-based modeling layer with self-service data exploration.
100% cloud-based with a centralized data modeling approach for BI teams.

Holistics is 100% bootstrapped, while Looker is VC-funded so its approach to product development is more customer-friendly:

Gentler learning curve towards data modeling setup so you can save time on training and quickly spring to actions.
Friendlier price point with pay-as-you-go pricing so you can scale without sweating money.

Holistics - The Best Looker Alternatives

III. If you're a big fan of fancy visualization, then go for Tableau

Most people agree that Tableau comes with the best variety of data visualization types that are highly interactive and beautiful in nature.

Tableau vs Looker: Visualization

Looker vs Tableau: Visualization

III. Tableau vs Looker: Key Pros and Cons

The best thing about Tableau: Beautiful visualization. You can create excellent dashboards on Tableau.

Tableau Limitations:

Tableau can cause request queue frustration as everything has to go through your data analyst. Non-technical users are not able to create reports and get insights by themselves. (When people try to have Tableau do more than Viz, it fails)
Tableau lacks data modeling and data dictionary capabilities for Data Analysts - which means that you’ve to separately maintain your metrics definitions elsewhere (that's too much work for data teams!)
It also lacks version control and collaboration when building data logic and dashboard. Without proper data governance, data consumers eventually lose trust in the dashboards you have put hours into preparing.

The cons of Tableau

The best thing about Looker is: Code-based modeling layer with self-service data exploration. Analysts can define analytics logic centrally (no metric knife fight!) and usiness users perform self-service data exploration. You can also have Git version control.

Looker Limitations:

Eye-watering price point. The entry price point is around $35000/year - not many companies can afford that.
Limited visualization.
Steep learning curve. LookML is not easy to learn.

(For more Looker alternatives that are open-source and affordable, check out this list)

IV. Conclusion: Tableau or Looker?

It goes without saying that you can't just choose your BI tool based on what people say on the internet (even this one).

Having said that, senior data practitioners provided a few helpful pointers when choosing the right BI tool:

Talk to your data consumers and data explorer. Create a list of key dashboards you want to build at an 80/20 level (don't have to make it perfect, just do the 20% of the work that gets the dashboard to 80% of what the vision is)
Come up with a small UAT group who will try interacting with the dashboards to give you feedback
Try out 3 tools you are debating between and evaluate the following questions:
Does the tool accomplish the goal of the dashboards?
Was the tool super annoying making you not want to use it more?
Was it hard to get setup? Will it be hard to keep running?
Are your UAT group happy with the results?

(All of the screenshots used in blog posts are sourced from expert's discussions dbt slack and Locally Optimistic slack - we encourage you to join these communities to see full conversations. )