Every successful business accumulates more and more data, and there are many solutions to cover reporting needs. We outline this in our Data Intelligence article. But the value of a Data Team goes much further, especially thanks to Operational Analytics.
Producing actionable insights launches your business strategy to the next level and directly impacts everything from revenue to company goals and next steps. It’s a no-brainer.
“This is the challenge, here’s the guaranteed solution, done.”
Here’s how it works: When a business uses a data warehouse, there are several data sources from the company and its overall environment — including SaaS applications (Salesforce, HubSpot, etc.). Therefore, the warehouse has a lot to work with and is able to make a variety of decisions based on extensive context.
Taking that context, Analytics Engineers prepare statistics about the correlation between specific events within Google Analytics (or a similar service) and then create a condition that determines the next, most suitable process.
In other words: “Your poor results in the X space are due to Y, so the system is automatically activating process Z — which will solve the issue, as proven by factual data.”
That is Operational Analytics. To make it even simpler:
Imagine the Data Stack (or Data Team) as a human body.
- The senses (sight, sound, smell…) are data sources that provide you with various data.
- The brain is the data warehouse, where the data is gathered. It’s able to process all those senses (data) and send a signal (condition) for you to move your arm, for example. The condition is created based on the “dashboard” generated in your brain.
- The reflexes are operational analytics. You don’t have to make a conscious decision to perform an action — it happens automatically, like breathing in and out. It’s operational.
A Data Analyst transforms your business into a simple equation that guides future operations.
There’s one approach that applies to every business at nearly all times: Operational Data Models. No single dataset or dashboard has ever been able to deliver value at this scale.
Coined by Benn Stancil, Operational Data Models work based on a Data Analyst thinking about a business like a series of models. Using a simple equation and basic metrics of the business, he/she creates frameworks that directly contribute to a business’s revenue.
Take Facebook. The equation is: (the number of daily active users on the platform) x (the minutes on the platform per daily active user) x (revenue per minute) = Facebook’s revenue.
Because a Data Analyst understands the business perspective and the overall data patterns, he/she is able to create an operational model — thereby explaining to the stakeholders exactly which of these metrics they should invest in, when and why.
If your wallet’s tight, should you cut Operational Analytics?
How Operational Analytics can help you largely depends on which stage your business is in — more on that in Tristan Handy’s Startup Founder’s Guide to Analytics. (Spoiler: Once a team has 20–50 people, it no longer makes sense to do things manually and using internal reporting within Google Analytics and others is a headache. It’s then time to invest in a warehouse.)
The fact is, utilizing Operational Analytics will 100% save you costs down the line. But it could take a year — which, for a startup with very limited investments, may feel just a tad too risky.
However, for more established companies that not just want to but have to step up their game, this is the only way to go.
A Data Team under the same roof as other engineers = gold.
At STRV, our Data Team collaborates with our Backend Engineers and all the other teams, which lets us define, implement, test and deploy solutions with zero delays — and without having to explain certain processes to someone less experienced in the field.
The above is especially useful in the case of Data Acquisition, which is the process of digitizing data from the world around us so it can be displayed, analyzed and stored in a computer. When all engineers work together, the Data Analyst keeps data acquisition in check every step of the way, making sure all data is created/accumulated in the right format. Zero hiccups down the line.
How has STRV utilized Operational Analytics for our partners?
One of our clients asked that we build a talent audition app with bite-sized content that resembles TikTok. For our Data Team, that meant delivering a recommendation engine and a search engine.
These two solutions could have been solved solely with backend API — but by using Operational Analytics, we made them much more powerful thanks to extensive content.
The Search Engine
We take data from the backend, combine it with behavioral data — a.k.a. events (how users use the app) — and find relevant results. We combine this data and load it into OpenSearch (a highly responsive search engine with advanced features for ML), allowing the user access to the Data Team’s output directly in the app. In other words: a part of the Data Stack is used in Operations.
The Recommendation Engine
The user opens up the app, and OpenSearch decides which new content the app recommends — which is done by combining data from the backend and events. Normally, recommendation in apps starts with a curation team that handles it manually. Instead, we chose to build the MVP on data from the get-go, thereby negating a future lengthy migration process from manual to AI/ML.
So, you want to go for it. What can you expect from STRV?
We have plenty of experience with AWS, which is the go-to cloud platform for most companies. Our experience also spans across all relevant data technologies, including dbt, Snowflake, AWS, OpenSearch and other open-source (for the budget-conscious) as well as more niche tech.
In closing, please know that Operational Analytics isn’t a walk in the park. It takes not just skills but a specific mentality to do it right. You’ve always got to be improving, triple-checking your work and documenting every step.
But do all that right and you’ve got a competitive edge that’s very hard to beat.