How Looker makes growing a data platform an organic experience
Mar 24, 2017
In case you hadn’t heard, Looker was deemed “Fastest to Implement” in the latest G2 Crowd Implementation Index report. Looker’s customers reported an average implementation time of 0.8 months. When you consider that other vendors implementation times are measured in whole months, Looker’s track record really disrupts the popular paradigm that a new BI platform has to be a long painful process.
Our customers’ quicker, more efficient time-to-value is due to some interesting aspects of Looker’s product and Professional Services. We opt to take a different approach to services: Rather than build the customer’s BI platform for them, we enable them to easily build their own models and reports in a safe environment. This leads to tremendously accelerated skills growth, self-sufficiency and long term value.
The primary reason Looker implementations enact so quickly is that the customers roll out of their pre-sales trial with some serious momentum. Through the trial, the customer analysts have usually already gained some insights and started to build on the LookML model the SE created. That start, plus Looker’s excellent documentation and best-in-industry chat support make it possible for virtually anyone to learn to use Looker and build a valuable BI infrastructure. Learning is of course faster and easier with some personalized attention. So our customers shorten their learning curves and get assistance making early design decisions by working online with a Looker consultant for a few hours as they build out their BI platforms. After hundreds of implementations, Looker Professional Services has learned ways to get customers moving even faster, and can help them build a better platform the first time.
It’s been said that Looker is so easy to learn, often our consultants first challenge is to get the customer to take a break from building their model and take the first step in our enablement service: The 1.5 hour Foundation Training session. The training provides a broad overview and valuable insight into what’s possible with Looker. Even the most self-sufficient analysts tend to appreciate the tutorial.
Foundation training also covers the most common features and best practices for making the models and calculations easily sharable across groups and tips on how to create a platform that is easy to scale.
Another factor that helps accelerate Looker Implementations is that each customer gets a dedicated Looker Consultant. Each consultant works with the customer to learn their business’ unique set of priorities, then customizes the teaching experience. Working closely allows flexibility on what topics are covered and on the pacing. This targeted learning builds excitement and helps people to accelerate their learning to a speed and intensity that works for them. Having dedicated personalized support and guidance throughout the enablement phases of teaching, coaching and consulting really creates an efficient use of customer time. Because goals, priorities and design decisions don’t need to be repeated, the continuity ramps up the speed of enablement.
The co-development approach
The main part of our enablement methodology is our co-development approach. We like to explain it by using a board game metaphor: If we are about to play Monopoly and I start to read you all the rules, you’re probably going to get bored and impatient. Alternatively, if I give you a two minute overview of the game then we play a few rounds together, you’ll pick up it faster and have more fun. The co-development sessions are about building your model with someone experienced in order to keep your progress moving and making good decisions.
Learning by building the production platform
At Looker, we don’t just build the model and make the customer dependent on us for maintenance and scaling. Our emphasis on co-development ensures the customer truly understands the design decisions and has the skill needed to build out out the platform for other use-cases in their company. This is instrumental in creating an environment were calculations are shared across different teams and thus “data brawls” are avoided.
With our co-development approach the customer practices their skills building their own model in a safe environment. When they face the situation of having multiple ways to get an answer, the consultant can help them weigh the options and learn any clever design patterns they need to get the best results. As a final test, the consultant reviews what they’ve built, points out specific best practices and makes recommendations. Customers part with the confidence that they’re building their model right the first time.
Data for all
Once the platform is partially built we can begin getting the end-users involved. In a single 90-minute training session, our consultants train users how to be self-sufficient in modifying their own reports by adding fields, pivots, filters and more.
Since business users can use Looker to explore data on their own, and ask new questions of the database without Analyst intervention, the Analyst doesn’t get bogged down with report change requests. They can keep exposing and integrating new data sources to be explored.
Time and again we’ve witnessed how once the business users get to experience creative exploration without waiting in the report update queue, the use of Looker seems to take off in an organic way - going viral throughout a team or department. This opens an ever-expanding world for end-users to discover and build. A world that is rewarding, exciting and valuable for everyone.
On the road to data discovery
After user training, both the analyst and data explorers are able to work independently. At that point the Looker consultant is doing less coaching on optimal use of the software and begins to focus on consulting more broadly. First the dedicated consultant listens to any new opportunities the customer wants to expand into, then he or she, makes specific recommendations for growing efficiently. Those formal recommendations conclude the enablement service but not the consultant’s relationship.
We make it our practice to formally follow up with each of our customers about a month after the enablement service wraps up. Thirty days is enough time for customers to conjure up new ways to use Looker and new questions. At that point our consultant gives them a little boost with another round of recommendations. There’s a hidden benefit of this last step to Looker Professional Services: The insights we gain on challenges beyond the normal implementation phase allows for continuous learning on our part, and leads to better, more informed advice for our next customers.
Alan Randolph is a Sr. Program Manager at Looker. He’s led hundreds of Professional Services engagements and constantly endeavors to improve support for Looker customers.