How Analytics Teams Are Structured at an Agency

The agency I work for helps build websites, constructs designs, guides UX strategy, and more recently, provides analytics insights to clients. I am an analyst on the analytics team, and I get to help various companies in nonprofit, manufacturing, pharmaceuticals, government, healthcare, and tech that have a high impact around the world, some of these you may have heard of.

Our team is made up of a leader and several analysts. Each analyst typically gets assigned their own client. There is rarely a situation when there's two analysts on a client, but it does happen on occasion when a project is large enough, such as when a GA4 migration requires a lot of work because of how large the existing GTM custom tracking configuration is.

This structure works well because analysts can dedicate themselves and understand the nuances of the industry of one client. They usually have the expertise to do all the work. On occasion, if they need more knowledge or help, they can turn to the rest of the team via Slack or recurring team meetings to tap their knowledge base on a variety of problems that can come up, such as Looker Data Studio's new quota limit being hit or data privacy laws or Google Data Studio's rebrand. There's often niche or nuanced issues that our work runs into, such as the Shadow DOM affected GTM tracking or how to track and collect additional data for widgets and cards when you click a link in the card (the answer has to do with the Shadow DOM).

Usually, our leader and team mates can add some insight. Some questions, often, can't be answered, so we turn to the web and forums. For the most part, this is a new department that has been emerging over the last two years. Therefore, things are constantly shifting. Best practices and wiki's get written frequently on practices in GA4, GTM, and elsewhere. New skillsets, like Tableau, are being developed, out of a need.

Despite this chaos, we seem to be doing well overall. Clients get served, the work gets done, and ultimately, our main goal of defining an effective measurement strategy to measure success of the org's goals and implementing that -- gets complete. In a future article, I may break down the phases/steps of how we plan, decide, lay out, strategize, and direct a measurement strategy if there's interest. We're not order takers who simply implement work; we need a lot of communication skills to explain what we do, its value, and how to do it best. We also do a lot of proactive guidance to audit website's current analytics performance and measurement strategy to make recommendations on how to better measure success in the future.