Sign In

Leighton Welch

CTO & Co-Founder @ Tracer

Leighton Welch is the Chief Technology Officer and Co-Founder of Tracer, a data intelligence platform that provides transparency into business performance.23 He graduated from Harvard University in 2013 with a degree in Computer Science and Economics.2

Welch has a diverse background in technology and engineering. His career began at SocialCode in 2014, where he worked as a Data Engineer and later as Director of Consumer Insights.12 In 2015, he joined VaynerMedia as Director of Engineering and was later promoted to Vice President of Engineering in May 2017.1

In 2018, Welch became the CTO at Tracer, where he currently works.14 Throughout his career, he has been pioneering in the ad tech ecosystem, including running the first ever Snapchat Ad and consulting on commercial APIs for major platforms.2

Welch has worked with Jeff Nicholson, another key figure at Tracer, across three companies, providing economies of scale through technology.2 His expertise lies in managing complex data to derive meaningful business insights, which aligns with Tracer's mission to turn data into decisions for both technical and non-technical teams.2

Highlights

Mar 18 · linkedin

Tracer has a unique approach to data transformations.

While the “T” in ETL (or ELT) is often thought of as a single step, it is not. There are really two key types of transformations, and they should be considered independently.

The first is the transformation of raw data into a usable format or standardized schema. For example, taking JSON responses from an API and converting it into something that can (should) be stored in a database to query against. The transformations here are going to be specific to the source, but the same across all clients. For example, the process of converting Google Ads data into a database schema is going to be different than the process for Shopify data (because the responses come back differently). But for either of those sources, it will be the same for every client (or ad account or store) which uses that source.

The second is the “transformation” into business use-cases (in quotes, because the underlying data shouldn't be modified). For example, taking that Shopify data and organizing into product lines, or that Google Ads data and organizing it into audience segments. In this case, the process of organizing is the same across all sources (because it is coming from a standardized schema), but the organization is different for every client. It is the exact opposite of the previous transformation type.

Tracer handles this first type of transformation automatically for our clients. It is effectively managed ETL so that there is usable reporting out of the box. The IP is really just the schema we convert everything to. The second type of transformation, however, is done through our Context Engine. It’s a drag and drop semantic layer where users can organize this data (rather than us trying to). This is the only scalable approach, because every user needs to organize the data differently. There is no one size fits all.

I see a lot of organizations treating these two steps as one and the same - often oversimplifying with “we use DBT.” DBT is great, but looping these two steps into one is going to create long term bottlenecks, and ultimately hinder your ability to move quickly with data driven decisions. That is because the first transformation is a purely engineering exercise. The second requires substantial business context.

Jan 3 · linkedin

Even the best #BI tools require a database. And even the best database requires data. Which means you also need connectors and a pipeline. And some people to configure said connectors, pipeline, database and BI tool. Now you have a lot of complexity and cost when all you wanted was a pie chart.

That is why Tracer exists.

Related Questions

What are some of Leighton Welch's most notable achievements at Tracer?
How did Leighton Welch's role at VaynerMedia prepare him for his position at Tracer?
What specific technologies does Tracer use, and how does Leighton Welch contribute to their development?
Can you provide more details about the AI-powered platform developed by Tracer under Leighton Welch's leadership?
What challenges has Leighton Welch faced in his career, and how did he overcome them?
Leighton Welch
Leighton Welch, photo 1
Leighton Welch, photo 2
Add to my network

Location

New York, New York, United States