What is operational analytics?

Operational analytics, as a data philosophy, aims to address many of the limitations of the stacks of data past (and, truthfully, many modern data stacks). Namely: This means turning data from insights that fill ad hoc dashboards and reports into action that can unite the efforts of every team within the modern business. 

This is about more than just powering analytics with rich behavioral data. It’s about empowering the business teams who consume that data analysis directly with all of the data available to your organization. 

By empowering teams to make more informed day-to-day decisions with data, a company can improve how efficiently it runs. Seems simple, right? The truth is that as obvious as the operational approach sounds, it’s not an approach many companies have pursued yet—though the needle is moving. 

The question remains, however, if operational analytics is so effective, why hasn’t everyone been doing it all along? Well, let’s take a look. 

History: The journey to doing more with your data

Not long after the advent of the data warehouse, and the subsequent explosion of analytics and business intelligence functions, organizations realized different teams had varying pictures of their users. This was because each business team was working off a different, siloed dataset about customers. 

When these disparate data sources were combined in a warehouse, data teams (and the business teams they supported) unlocked a deeper understanding of their users’ behavior, which started the race to get as much customer data as possible into the warehouse as possible. But it turns out, you can have too much of a good thing, in a way. 

Organizations were so focused on getting all of their data into one place that they disconnected the data from the people who needed it the most.

Data quickly outgrew BI tools’ ability to meaningfully represent it as both quantity and complexity scaled. These dashboards and reports helped people understand the data, but there was a lag between when data was created and when a decision was executed. This lag could make or break the work of frontline teams when it came to things like prioritizing support tickets or targeting user segments. 

Organizations were now left wrestling with finding a way to seamlessly sync data from the warehouse to operational tools like Salesforce, Marketo, and Braze. Some built DIY integrations. Others used P2P tools like Zapier (on a small scale). Then, thankfully, reverse ETL came along and made operational analytics possible with near real-time data syncing from the warehouse to business tools companies relied on (all without a ton of engineering favors and brittle code). 

This real-time availability of accurate data makes operational analytics possible and unlocks a suite of behavioral data-driven use cases such as: 

• Ticket prioritization by product usage
• Abandoned cart recovery and prevention 
• Audience segmentation for marketing email drip campaigns 
• Customized lead scoring across touchpoints, devices, and channels 
• Detection of early churn indicators & most engaged customers
• 360-degree customer views 

By creating a unified customer view, frontline teams like support or marketing have the information and context they need to engage customers in just the right way at just the right time. While the movement to warehouse-as-source-of-truth promised a similar centralization of customer data, it’s the practice of operational analytics that lets you get the full potential out of all the data you’re collecting, your data team, and the business tools you’re paying for each year. 

Why do you need operational analytics?

TL;DR: Operational analytics is the only way to effectively democratize your modeled data from your warehouse at scale. Operational analytics lets you scale the complexity and sophistication of your data without overloading your data professionals or eroding trust in your data along the way. 

Without operational analytics, both business and data teams run into trouble: 

• Data teams are stuck in a never-ending cycle of answering ad hoc data inquiries from business teams, leaving them without the time to build their skills or do more advanced modeling to sharpen the company’s competitive edge. 
• Business teams can only access the data available in their tools (and often this data is incomplete or out of date since it’s out of sync with the warehouse). Additionally, frontline teams often feel like they’re operating with one hand tied behind their back, unaware of what other customer-facing teams are doing. 

Operational analytics saves everyone in your company from the brain-draining tasks that keep them from doing more, letting frontline teams and your data teams alike have a greater impact on your core business objectives. 

For example, with operationalized data, marketing teams can use Marketo to create hyper-targeted email campaigns based on sales outreach and advertising data. This means they no longer need to wade through miles of spreadsheets or pre-generated reports not suited to a specific task: the data they need is just where it needs to be.

Traditional analytics vs operational analytics

To truly understand the importance of operational analytics for data-driven companies, it’s worth breaking down the differences between it and traditional analytics. 

• Traditional analytics was used to understand what happened in the past of business operations using data presented in BI dashboards or regularly-generated static reports. Traditional analytics put the burden of data gathering on the data team. 
• Operational analytics is the practice of using data, collected, transformed, modeled, and synced via the modern data stack, to fuel future action with democratized access to fresh data across the company. Operational analytics gives business teams the ability to access reliable data themselves, freeing the data team to work on more advanced projects. 

This shifts the focus off of inferring based on historic data to strategizing with the most current data available, all synced from one central source of data truth (the warehouse). 

Traditional vs operational analytics in action: Customer success 

Let’s look at a common example: The customer success team at a growing startup needs data to understand where users are running into issues in the product. 

With traditional analytics, their data team could pull together a BI dashboard that tracks the kinds of tickets users submit, as well as how long it takes to resolve those support requests. This sort of information would help the team see how much capacity they have, but leaves all the problem-solving of how to tackle the backlog up to the human in the loop. 

Instead, when these same customer success teams and data teams use reverse ETL to operationalize data, the burden of refreshing the dashboard’s data and prioritizing tickets disappears. Now, data from other sources–like the CRM–is pulled together with each support ticket so each incident can be automatically prioritized based on predetermined key characteristics. 

When customer success teams can operationalize behavioral data about their customers, they’re able to demonstrate a coherent, end-to-end understanding of each individual person they’re serving. This makes it easier than ever to answer key questions about those customers and delight them at each touchpoint. 

Furthermore, this brave new world of operationalized data makes it possible for data teams to focus less on refreshing stale dashboards and more on advanced modeling, as well as giving customer success teams all the information they need to have the most impact possible with customers. 

Operational analytics and the modern data stack  

It’s no secret: The modern data stack is the gold standard of data-driven business today. However, while cutting-edge data companies continue to push the envelope of what’s possible with data every year, the inability to action or operationalize the data companies have worked so hard to collect in the warehouse leaves many data initiatives coming up short. 

The secret is an updated vision of the modern data stack as the nervous system of the modern business. By investing upfront in key core infrastructure, organizations can seamlessly move along their data maturity journey. The new modern data stack includes the familiar components–loading, storing, transforming, and syncing–alongside operational tools like reverse ETL. We generally see it break down into four tool categories:

1. Data warehouse and data modeling tools act as the brain of your business, processing and sharing data in and out of the body.
2. Behavioral data creation and ETL tools act as a relay system between the core parts of the business to keep systems up to date and consolidate vital information. This includes Snowplow, of course. 😊
3. Operational tools like reverse ETL (Census) transmit and contextualize important data from the brain out to the frontline tools at the edge of your business. 
4. Marketing, sales, and support tools act as the limbs of the business body, fielding interaction with the world and customer relationships within it.  

In the above example, you can see how all four categories of modern data tooling make up a healthy, functional body of data-driven business. This ecosystem, fueled by enriched, trustworthy data, is able to seamlessly move throughout the world, collecting input from, interacting with, and learning about customers at every touchpoint. 

Snowplow and operational analytics: Helping you do more with your data

Snowplow plays a critical role in creating and transporting data from the limbs to the data warehouse, collecting data from web and mobile applications and loading it into your data warehouse. 

Because the data Snowplow creates is incredibly rich and granular, it is well-suited to fuel an operational analytics engine, especially when paired with a reverse ETL tool like Census. Snowplow’s wide array of data collectors let you start tracking event-level data across a variety of different programs and applications right out of the box so you can tie all of your data sources together in your warehouse. 

By combining the functionality of the Snowplow Behavioral Data platform with a powerful operational tool like reverse ETL, companies can automate data integration across the last mile of the modern data stack. 

This is where reverse ETL tools like Census shine, taking all of the brilliantly modeled data your data team has worked on and transmitting it throughout your organization so every tool tells your frontline teams the story, whether they’re relying on Hubspot, Zendesk, or Marketo. 

Take the first step toward operational analytics 

While having a fully built modern data stack certainly speeds up the use cases you can power when you operationalize your data, you don’t have to have every tool under the sun to get started. 

Instead, focus on getting the data that’s key to your business objectives into a central source of truth like a data warehouse. From there, you can begin improving how frontline teams work by adding a reverse ETL tool to your stack (freeing up your data team to build out other areas of your stack in tandem).  

To start, your data team should work closely with the frontline team most in need of fresh data to understand what information, exactly, those customer-facing teams need most. This ensures you operationalize the data with the highest impact from day one before moving onto more advanced use cases. 

Why wait any longer? Find out how Snowplow and Census can help you operationalize your data and make your teams more effective. You can sign up for a demo of Census here and with Snowplow here to get started today.

Written by Allie Beazell, Director of Developer Marketing @ Census