How to prepare for the LookML Developer Certification exam?
Update (Oct 2022): Google announced the consolidation of its Business Intelligence products under the Looker brand name as part of Cloud Next, the annual GCP event. As part of this announcement, Data Studio, the free visualization tool, has been re-branded as Looker Studio. This also marks the origin of the new premium tier called Looker Studio Pro, which has more organizational friendly capabilities. And the Looker Platform is deeply integrated with as well as complements the Studio products. I can imagine a new (set of) BI certifications from Google Cloud testing a breadth of skills across the entire Looker portfolio will come out soon. I expect this article to be still relevant when that happens. :)
Note (Apr 2022): Google has sunset the LookML Developer certification exam that this article is about on Apr 1st, 2022.
Are you thinking of taking the LookML Developer exam and getting Looker Certified? Look no further. Use this article to plan your preparation.
Looker has been officially part of Google Cloud since Feb 2020. The certification swag is really cool: I got myself a Looker branded collapsible duffel bag. I expect it to be quite useful. :)
I passed the exam just last week. I’m eager to share my experience and preparation journey with fellow enthusiasts and eager seekers. First things first, I did not find this exam too difficult. It’s challenging, but rightly so. Of course, it depends on one’s background and prior experience with SQL and data models in general. Even though I’m new to using Looker (have just been playing around with it for a few months now), I have extensive prior experience with other BI tools like Tableau, Power BI, etc, and also strong expertise in SQL, data modeling, and data warehousing.
The exam has 50 questions to be answered in 100 minutes. Plenty of time to do a couple of passes, as needed. The questions are usually not too long and are straightforward enough.
As with all certification exams, the official exam guide is the first place to start. Focus areas tested include:
- Model management (troubleshooting, security, content validation, and more)
- Customization (creating and modifying dimensions, measures, explores)
- Optimization (caching, derived tables, query optimization)
- Quality (version control, code quality, data validation)
Looker’s official training on their new training platform Looker Connect is a great resource. You need to create a free Looker Connect account (which is different from your Looker account) in order to access the training content. Focus on the LookML Developer path for the purposes of this exam. I also diligently went through all the documentation links provided under the Study Resources section on the official certification page, which helped a lot.
I cannot emphasize enough the importance of having hands-on practice on the platform creating dimensions, measures, explores, and various parameters. The Google Cloud Skills Boost LookML Quest is a great start. LookML IDE provides instant feedback on any mistakes and provides a lot of contextual help and information while developing. This makes it easier to create the much-needed muscle memory using various settings and parameters. Given the large number of parameters involved, it could get confusing after a while to remember how different parameters behave within different objects (model, datagroup, explore, view, derived table, etc.)
Some key topics to pay special attention to:
- Dimension types and parameters (tiers, location, distance, dimension groups, etc.)
- Measure types and parameters (aggregatable vs non-aggregatable)
- When to use ephemeral derived tables vs persistent derived tables
- Caching policies: when to use persist for vs sql_trigger_value vs persist_with vs max_cache_age vs sql_trigger; datagroups
- Joins: from clause vs view_label
- Symmetric aggregates and fanout problem
- Filtering data: sql_always_where, sql_always_having, access_filter, always_filter, conditionally_filter
- Data security: access_filter, access_grants
- Git integration options
- Project files
- Content Validation — what happens when object names are changed, when an explore is moved from one model to another etc.
- LookML best practices; Dos & Don’ts
- Optimize Looker performance
- Create a Positive Experience for Looker Users
