What’s new with Looker 2024?
My attempt to summarize the Looker 2024 roadmap by Google, as there were a lot of major & minor changes.
If you haven’t been able to watch the session, here is Google’s high-level recap:
The basic theme is Looker unified experience + Google’s GenAI will change the way your organization work with data.
If you have watched the session and got bamboozled by the truckload of information that come your way, welcome, me too. This is why I’m writing this because there were a lot of changes introduced that require preparation and transitioning, and Bard wasn’t too helpful in summarizing the news.
After telling Bard that I’m disappointed, I had no choice but to do the labor myself.
Looker 101
If you are new to Looker family products then a little introduction is needed before we dive in. Looker was originally a separate company until being acquired by Google in 2019-2020.
Looker by itself is not comparable to other BI tools in the market like PowerBI, Tableau, etc. It has originally been a BI platform. The differences are:
- It has a semantic layer
- It is probably the first ever that integrates with Git version control, adopting software development practices in the development workflow. This is a fundamental mindset shift that goes beyond “dashboarding” works.
After the acquisition, Google changed Google Data Studio into Looker Studio, and paved way to the vision of merging its original Studio product with Looker, bringing the best of both worlds, so to speak. As a result, you would hear of these product lines under GCP:
- Looker (it can either be Looker-hosted or Customer-hosted instances)
- Looker studio (traditionally Google Data Studio). This comes in two versions: Free and Pro.
That is until today, we have changes announced in 2024 roadmap, and hear of Looker Google Cloud Core in addition to these.
The 2024 roadmap
Summary
The vision for a unified Looker experience can be broken down into several benefits offering for different target groups:
Which Looker?
[Updated Feb 19] These changes will be available to Looker Google Cloud Core (let’s just call it Looker Core from now). With regards to Looker hosted instance (original) — the changes seem to be limited (only around performance in back-end).
The changes all point to less of a LookML learning curve for end users who have been craving the ease-of-use, fast-delivery cycle that Looker Studio offers, and accelerate the speed and ease of development for developers.
Takeaway
For me, a big change here is the promise of cohesive experience where we have a single login to access and develop everything. This can be a game changer if your organization has been struggling with the problem of organic dashboard development from non-data teams outside of Looker ecosystem.
For data teams that care about maintenance and scalability, no more “why is it not in Looker?” (which kind of defeats the purpose of single-source-of-truth that every engineers crave for and all end users complain about — while creating hundreds of reports in Looker Studio because of its flexibility advantage).
One window to rule them all, finally! At least I hope so.
Another change that was not very well demonstrated in the session (most of it was surrounding the ecosystem of unified experience and AI-powered analytics), but I really looked forward to, is performance improvement. When your organization operates by the thousands of users and hundreds of developers, and each of your explore has about hundred joins — this change matters. I wish Google had covered in some more details and we could see some benchmarking in query speed, for instance.
The features
Now that we have some grasp of the core vision for Looker in 2024, let’s take a look at the specifics.
[Updated Feb 19]
(*) These features were shown in the demo session via either Looker’s or Looker Studio’s familiar ‘front-end’. But since Google’s vision is to merge these eventually, at some point we can reasonably expect that the features would be available to all, when there is no more Looker vs Looker Studio distinction.
Alongside these features update, some important ones are:
- Looker connectivity ecosystem — where Looker modeling layer is at the center, offering integration with Thoughtspot, PowerBI, Tableau (via JDBC Driver) etc.— this will open up whole new world for companies that have long wanted to have a semantic layer separated from BI layer while being locked-in with their existing tools. I can already imagine data teams with maintenance pains (think dozens of dashboards with care-free users that often add layers of untraceable & multi-versioned formulas to define a single metrics )— will benefit from this update.
- Looker Marketplace — Looker block for Google Workspace analytics is available. Would be interesting to plug in your organizational Google Calendar and analyze how much time meetings take in a week, how fast is your customer response time (Google Chat space) or something around productivity analytics.
Transition
From Looker hosted to Looker Core
Looker hosted instances can be migrated into Looker Core “soon” (unidentified timeline), and available “if we are interested” — I suppose this means it will be up to each organization choices to transition or not, at least for now. From the session, not much detail is shared and it seems like Google is still working on this.
Benefits of transitioning?
Assuming this transition is something up to choices, and no differences in pricing, inevitable questions arise around why would an organization want to transition to Looker Google Cloud Core? These are the exclusive offering from Looker Core:
- Expanded Regional Availability: seem like a strong case if your organization user base is multi-regional, and you want to minimize latency and performance hiccups that come from multi-regional workloads. Google has completed 20+ regions worldwide and will roll out 10+ more in 2024, and will not stop there.
- Support for Private Service Connect (PSC) — if your organization security policy demands private connection, this is a no-brainer. You would definitely need to migrate.
- Control over Looker Versioning: you can decide how frequent your Looker instance is upgraded: Rapid, Default (monthly), ESR (Quarterly) — enabling self-service upgrades and downgrades for the open valid versions available for your enrolled channel.
- [Updated Feb 19] AI-enablement features: earlier version of this post — I was under the impression that AI-enablement features would be also available in Looker hosted instance (original) — however, this is updated. If you want to reap the benefits of these features, they are only available under Looker Studio Pro and Looker Core.
Pricing & offering
The origin story:
- Looker Studio used to be Google Data Studio, and it was free. As of today, Google offers Looker Studio (still free), and Looker Studio Pro (9 USD/user/project/month).
- Looker hosted and Looker Core both charge Platform pricing costs (edition based) + User license seats. There are some differences in Edition between Looker hosted vs Looker Core, same with User license types.
A note: For those that are only interested in Looker modeling layer to integrate with BI tool elsewhere, the pricing seems to be quoted separately as well.
Which Looker for which use case?
If we consider all things together, this is my takeaway for the product fit to each organizational need. It’s important to note that all AI-enablement features will be available to all Looker Studio, Looker Core, so the key driver for migration/ upgrade needs come down to sizing, PSC requirement of your organization, and AI-enablement.
It also makes sense to see that these offerings reflect the different maturity stages of your organizational analytics needs, if you consider a staged approach and have budget limitations when you first get started with Looker.
Recap
From the whole packed session:
- Google GenAI and Looker unified experience will be the key changes.
- These changes will drive self-serve capability for end users more than before, both in terms of ease-of-development and customizability, predictive analytics
- We can expect some performance improvement for heavy use cases. Hopefully some more clarity in 2024.
Looker modeling layer is now available in Preview, the Looker unified experience and GenAI enablement in usage and development will integrate well with business workflows and leverage data-driven processes. It’s a good time to reflect on your organizational analytics need and take into account these changes to leverage the benefits of Looker products.
References
https://cloud.google.com/looker-studio#pricing
Announcing my new page
Hi there! I’m moving over to Substack. You can find me in the link below and subscribe in the new space. See you on the other side :)