Would you rather spend your time using data to fuel innovative new product features or sit at your desk for hours trying to make sense of messy metrics? That should be a no-brainer (unless you’re some kind of data masochist).
Unfortunately, many product managers and data analysts spend more of their time fixing bad data than they do using it to propel their products forward. In fact, according to Harvard Business Review, only 3% of companies’ data is of good quality, and Forrester estimates that making sense of the bad data takes up more than 40% of data analysts’ time (some estimates are as high as 50%). Furthermore, an MIT study found that only 3% of the data from sampled enterprises was good quality, with at least 50% containing critical errors.
There’s a word (technically two) for the time and money cost of dealing with this bad data: data debt. And debt, even good debt, sucks.
Data debt is the cost of short-sighted data decisions. One of the biggest contributors to data debt is messy analytics tracking, meaning each product team implements analytics tracking however they see fit. This creates a hodgepodge of different event names and data points that takes time (and money) to untangle later on so it can be compared.
Like other debts—student loans and mortgages—data debt is ubiquitous. And, like other debts, you need to be in control of it.
The great news is that clean analytics tracking can help you wrangle your debts, get them under control, and prevent your balances from going up in the future. The opposite of messy analytics tracking, clean analytics tracking means you’re taking an active role in measuring metrics that are relevant to your business, auditing your sources of data, and ensuring that everyone implements analytics correctly so you can trust your data.
When you invest in clean data analytics, you can understand your current debt, pay it off, and prevent it in the future.
If you’ve waded into the world of Agile and DevOps, you’ve probably heard of technical debt (i.e., the price of choosing ad hoc/short-term solutions over the more strategic route). Data debt is the debt you rack up when you make ad hoc data governance and management decisions without thinking about the long-term impact on your data health and usability.
The issue with data debt is that it creates problems you’ll have to untangle in the future. And each problem takes time to fix and creates further murk in the water you have to wade through to make decisions.
But despite this, data debt isn’t an inherently dirty word. It’s a way of understanding the cost of doing the right things versus just doing what works in the moment.
If we take a closer look, data debt typically takes on one of four forms (as shown in the graphic below):
Like their real-world financial counterparts, not all types of data debt are created equal. Some types of debt, such as immature debt, are a part of the journey to good data governance, while others, such as selfish debt, are signs that data management best practices aren’t being followed.
The trick to minimizing your data debt is a mix of good data governance and clean analytics tracking. The former prevents selfish and ignorance debt and the latter helps you uncover the sources of your debt, settle up on your balances owed, and prevent future debt down the road.
When you combine macro governance best practices with the ability to consistently implement tracking and measure metrics, you can make better business decisions and better prioritize and communicate the importance of larger, strategic product choices.
As much as we love clean analytics tracking, it’s not going to pay down your debts for you. Rather, as you put in the work to develop, implement, and support clean analytics, your data debt will naturally go down (and you’ll be able to avoid it in the future).
When your analytics tracking is messy, everyone from your developers to your product managers can trust your data. All the short-term data decisions you made in the past make your future data all that much harder to understand, leverage, and, ultimately, trust.
If you can’t trust your data, you can’t use it to make good product decisions. To get rid of your data debt and build trust in your information, the logic behind your analytics tracking needs to check out. That means creating a single source of data truth—aka clean analytics tracking.
Just like real-word financial debt, you can’t start dealing with data debt until you know how much of it is hanging above your head and how it got there in the first place. To understand how much debt you have, you need to sit down and audit every platform and application that you, your teams, and your product work with.
Then, you can figure out which applications generate the most data debt (that is, which individual tracking calls were implemented in a way that causes the most downstream work), come up with a plan for resolving those issues, and outline a framework built on clean analytics to prevent debt-creating decisions in the future.
To develop and implement clean analytics tracking, you’ll need to come up with tracking best practices through a tracking plan.
When you’re building out these best practices--such as establishing naming conventions, outlining where code should be placed, and defining which events matter most--you’ll uncover all the instances of messy tracking that you’re currently using. This process of discovery and clean up will help you understand what previous analytics choices were causing your debt, and help you systematically eliminate them.
In many cases, data debt is caused by poor tracking-analytics implementation. Often, this happens when developers have to quickly set up and integrate new tools or third-party applications and either don’t know how to follow best practices (because they’re aren’t defined) or choose not to for the sake of speed and convenience.
This was the case for a client of ours, Termius, from day one. Like most software companies, Termius serviced several platforms, and each platform had its own applications and codebases throwing their own data. It was an uphill battle to sync all of their analytics together, and their developers often made mistakes implementing analytics across the board.
If they couldn’t check the logic of their analytics, they couldn’t trust their data. So they needed better analytics to pay off their data debt and make better decisions.
Back at the beginning of 2019 (which feels like eons ago), Termius had started developing a plan for how their data should be. Based on the document, we worked with them to identify issues and set their tracking plan up in Avo. From there, they were able to implement events in their codebase and fix future issues along the way.
As a result, Termius developed a holistic understanding of all of the issues racking up debt and took the first step toward paying it off.
Clean analytics tracking helps you settle your data debt by forcing you to standardize your messy analytics. To set up clean analytics tracking moving forward, you have to resolve the implementation issues committed in the past that caused your debt.
To start, take your list of data debt sources and look at which ones have the biggest impact on your day-to-day operations. Is there an implementation error that’s more egregious than the rest? Or does one of your applications contribute more user and product data than the others but have inconsistent event and property naming? Start there by correctly implementing your analytics tracking to be in line with your central tracking plan.
Settling up data debt will prevent headaches downstream for years to come and save your data analysts (and everyone, really) time on the back end trying to make heads or tails of miscellaneous data.
You can do this work manually by having your developers look at the implementation of your analytics tracking code for each application or by using a tool like the Avo Inspector, which analyzes your data for you and flags issues in your tracking setup.
Using tools like Inspector alongside clean analytics tracking tools like Avo will help you discover data issues and see which ones are causing the most issues. From there, you can untangle the worst data offenders within your stack and begin to enjoy the fruits of your labor.
Clean analytics tracking helps you avoid future data debt by ensuring that people don’t make the same kinds of short-sighted decisions that ballooned balances in the first place. To set up clean analytics tracking, you need to adhere to a tracking plan as part of larger data governance, which will outline how all tracking decisions should be made to prevent messiness.
This benefits everyone in your company and prevents data debt in the future. Product managers now have data they can trust to back up product decisions, and developers know exactly how to implement analytics every time to prevent future debt and don’t have to try (and often fail) to track changes between releases. This central strategy for setting up how analytics are implemented—and when—prevents in-the-moment product decisions that need to be untangled down the line.
At Termius, having clean product analytics through Avo meant product managers no longer had different sets of metrics and analytics for each platform (the source of their data debt). They didn’t have to check between each application to make sure they were comparing data apples to data apples. Instead, their developers and product managers could look at the data and know that the context of it—the events and properties it was tracking—was the same no matter the source.
As long as you maintain your data governance best practices, you’ll achieve clean analytics tracking, and you’ll be able to trust your data to help you make any decision that comes your way.
Once you understand how your data debt came to be, it’s time to take the first step toward eliminating it and freeing your business from its grasp. Avo makes clean analytics tracking possible and easy, and our leading product analytics platform can help you use your data to support better business.
Start chipping away at data debt—and prevent future data missteps—today with Avo.