Lucid Increases ROAS by 52% with Snowflake, dbt, and Hightouch

The Results

• Lucid now has a data-driven marketing tech stack built on Snowflake Data Cloud, dbt, and Hightouch.
• With dbt and Snowflake, transformations that used to take hours can be done in minutes.
• Using Hightouch and Snowflake, Lucid achieved a 52% increase in ROAS in Google and a 37% increase in new users.
• Since adopting this Hightouch and Snowflake, Lucid’s engineering team has been able to save an average of 12.5 hours of engineering time per integration.

The Challenge

With a relatively limited data stack built around AWS Redshift, Lucid had no ability to personalize customer engagements across channels. There was no 360-degree view of the customer and this meant that it was impossible to optimize ad spend/performance or create customized marketing campaigns for different audiences. Transforming the raw data in Redshift was a nightmare, and sending data to various business tools and ad platforms was even more difficult.

In many ways, we developed our tech stack with a vision for the future, but we also wanted to solve the problems we have now. The problem with any out-of-the-box analytics suite is that you are restricted to only doing what it supports natively. CDPs are not extensible to what you want them to do.

Weston Rowley

Director of Strategy & Analytics

•

Lucid

The Solution

Snowflake

Searching for a robust data platform that could act as both a data lake and a data warehouse, Lucid turned to Snowflake. Snowflake consolidates Lucid’s into a single centralized platform and lets the company ingest data from multiple sources.

With Snowflake we don’t have to leverage a data lake. We can use the same processes across all of our different data flows. Snowflake can act as the giant bucket where we dump our raw data.

Weston Rowley

Director of Strategy & Analytics

•

Lucid

Since implementing Snowflake, Lucid has been able to establish a single source of truth for all of its customer data and remove all of the barriers that came with Redshift.

Many of our metrics that run the business come through Snowflake at some point in time. Once our data is in Snowflake, everything is easy to work with and there is absolutely no friction.

Weston Rowley

Director of Strategy & Analytics

•

Lucid

Snowflake’s native separation of storage and compute gives Lucid unparalleled flexibility and makes it really easy to scale up or down as needed. With Snowflake powering analytics, Lucid can finally trust the data.

dbt

Once the data is in Snowflake, Lucid still has to transform it from its raw state and build data models so that it can be used not only for analytics but also for data activation. This is exactly where dbt comes into play.

All of the logic that goes into calculating our core metrics lives in dbt and it’s essential to everything that we do related to analytics - dbt is the underpinning of everything.

Weston Rowley

Director of Strategy & Analytics

•

Lucid

Since implementing dbt, Lucid has been able to build a number of custom data models that are purpose-built for different marketing and analytics use cases. These models include everything from:

• Offline conversion LTV valuations for different ad platforms
• Custom audience segments for ad platforms and marketing campaigns
• Recently acquired customers from channel partners
• Custom objects (like leads, subscriptions, accounts, licenses, etc.) for marketing automation emails
• Business metrics for executive reporting
• Errors and alerts for Slack notifications

Instead of having to write ad-hoc SQL queries and leverage custom code, dbt gives Lucid the ability to standardize, orchestrate, and execute SQL queries directly against Snowflake. Lucid’s Strategy and Analytics Team uses dbt for 95-98% of all data their data transformations.

dbt is so much faster than writing in another language and it’s incredibly flexible. You can run anything in dbt that you can write in SQL. What would otherwise take several hours can be done in a few minutes with dbt.

Weston Rowley

Director of Strategy & Analytics

•

Lucid

Hightouch

Before adopting Hightouch for Data Activation, Lucid built several custom data pipelines to send customer data to Marketo. However, this data came in from multiple different sources which meant pipelines were prone to failure. The constant data rate forced Lucid to keep a close eye on API limits and campaign backlogs. In addition to this, far more users were pushed into Marketo than were actually needed which meant Lucid was constantly fighting to keep the user count within the appropriate limit.

Sending data to ad platforms was also a nightmare. A single integration to Google Ads could take upwards of 12-16 hours of engineering time to build and this is not even accounting for the maintenance involved to keep it up and running; and with no built-in alerting it was nearly impossible to know when, where, or why specific failures occurred. Since pipeline breakages could have a huge impact on Lucid’s ad spending efficacy, manual CSV files were adopted instead.

Thanks to Hightouch, I no longer have to wake up on Saturdays and Sundays to upload manual CSV files. With Hightouch, we can trust that the same data that lives in our internal reports is the same data that is getting passed to our ad platforms.

Mallory Reese-Bagley

Sr. Marketing Operations Manager

•

Lucid

Since implementing Hightouch, Lucid has been able to optimize bidding in Google Ads to achieve a 52% increase in ROAS and a 37% increase in new users. Instead of taking 12-16 hours to launch and test a custom API job, a single integration in Hightouch can be set up in 5 minutes and the most complex data models can be set up in 2-3 hours, saving an average of 12.5 hours. Before Hightouch, Lucid had no ability to create first-party audiences for specific marketing campaigns. Thanks to Hightouch Audiences Lucid can now target high-value customers and launch cross-promotion campaigns for existing customers based on usage.

We just started using Hightouch Audiences, but now the marketing team no longer has to come to us to build an audience.

Weston Rowley

Director of Strategy & Analytics

•

Lucid

What’s Next?

As a PLG (Product-led-Growth) company driving product usage and accelerating adoption is vitally important to Lucid. With Snowflake, dbt, and Hightouch working in unison, Lucid’s data models will only continue to improve, enabling the data team to address more complex use cases around analytics and data activation to acquire even more customers.