Getting Started with Product Data Management

The power of product analytics becomes apparent when any team at a company—marketing, product, design, engineering—can quickly ask and answer questions about user behavior and customer journeys.

But getting to that point requires an investment in product data management. Strong data management ensures that data ported into an analytics system is clean, accurate and reliable. That way, when teams access the data through product analytics, they have a high degree of confidence in the insights they uncover. They trust the data, and they can act quickly.

Weak data management, however, can take many forms. Perhaps duplicated data is sent to a system, resulting in inflated metrics. Or, events and properties may be named inconsistently, making it difficult for teams to match their product experience to the data. Over time, poor data management practices can create data debt–lots of untrustworthy data that results in confusion, delayed decisions, or mistrust of data altogether.

If your teams can’t trust the data, they can’t act on it—plain and simple. That’s why it’s essential to prioritize product data management no matter how advanced your company’s analytics practices are. When everyone has access to informative data—and therefore the ability to answer all questions about the customer experience—they’re empowered with insights. They know exactly how to create the best digital experiences and bring real value to the customer. Your business and your customers benefit from effective product data management.

What Is Product Data Management?

From a big picture perspective, product data management is how you collect your data and make it available for others in your organization. The point is to gather the data you need to answer questions in a product analytics platform.

Building a product data management framework involves three questions:

1. What data should I collect?

Choose what events to track in your product. Define a taxonomy—a guide for consistent naming conventions across events and properties. Clearly defining what you are tracking and why will help keep your data accurate and trustworthy.

2. Where should I store it?

Product analytics platforms like Amplitude store data on their cloud servers to enable real time querying, but many organizations keep a copy of their events in a data warehouse. You’ll also want to sync data across both locations to prevent data silos. When some of your data is cut off from your analytics platform, you can’t use it to help you make holistic product decisions.

3. Who should receive this data?

In a data democracy, teams are empowered to explore data relevant to their work, to make informed decisions. However, democracy requires governance, and a designated data governor, or team of governors, should act as stewards of these systems. Analytics governance platforms like Avo allow product managers, developers and data scientists to plan, instrument and govern analytics at scale based on each team’s needs, while ensuring product data is clean, consistent, and reliable.

Why Does Product Data Management Matter?

As companies start using product analytics, many believe that any data is better than no data. With this mindset, they rush to implement analytics tracking just before feature releases without considering the long term consequences.

Trouble comes when they get to the other side of the release, and their data doesn’t tell them much. Did we track the right events? Does this chart mean we should change Feature A or Feature B? No one knows. Some try to clean up their data retroactively, but this approach often misses errors in tracking, and it’s never scalable.

These messy data practices lead to problems that hold businesses back: fewer product iterations, questionable data, and misinformed decision-making.

The solution? Building a product data management strategy.

Implement a data governance system. Create a tracking plan to standardize what product events you’ll track and why. With trustworthy data, your teams can reliably determine how product decisions are affecting the customer.

Product data management empowers you to iterate quickly, make data-informed decisions, and create powerful products. But how do you get there?

How to Improve Your Product Data Management Strategy

Every company has a product data management strategy, the question is, how mature is it?

If you are stuck in a cleanup workflow—haphazardly implementing analytics tracking and cleaning up your data retroactively—there’s lots of room for growth. The important thing is to start small and build as you go. Getting started with product data management is like building a habit: it’s nearly impossible to master all at once, but small actions done consistently over time build a strong foundation.

1. Get Data Producers and Data Consumers Working Together

Think of data producers as the engineers who implement analytics events in code, and thus “produce” data. Data consumers, meanwhile, represent the folks who explore the data in the analytics platform and make decisions based on it. They “consume” the data.

Engineers—data producers—are often at the end of the decision-making train, implementing the tracking events that the product team already decided on. But this workflow makes analytics implementation a chore and prevents engineers from sharing insights on the larger strategy.

Take a different approach by holding a purpose meeting for every feature release. This 30-minute meeting should bring together the following people:

• An engineer from each of your development platforms (iOS, Android, web, backend, etc.)
• A product manager
• A designer
• Any other relevant stakeholders.

As a group, align on the release’s goals, key metrics, and the events you need to track to calculate those metrics. Engineers will be looped in to the goal of the release from the beginning, and therefore can make informed decisions about how to implement tracking for analytics.

2. Improve Data Accuracy

Data cleanup will always be a part of any data governance workflow. No taxonomy is perfect, and all taxonomies will need to be updated over time.

The first step to improving your data accuracy is to find out what’s wrong with your data so you can have an informed discussion with your team on what needs to be fixed. Audit the data within your product analytics platform and compare it to your data warehouse or system of record. It’s important to find your discrepancies before they cause misinformed decision-making. Wherever you find them, figure out why they’re there and decide what to do with them based on the severity.

Even incremental improvements will start increasing its trustworthiness and functionality.

3. Create a Tracking Plan

A tracking plan defines the key stages of your customer lifecycle and dictates how you will track the data and metrics that correspond to each stage.

Use this plan to standardize your product data management practices and capture accurate and trustworthy data. It should be a living document in a shareable platform—a Google document, a wiki, or an app—that your team can update and collaborate on together.

As you create this resource, make sure your tracking plan:

• defines the events you need to track to calculate your KPIs,
• explains why you are tracking a given event or property and how it contributes to your goals, and
• instructs developers on how to track these events in your codebase.

To learn how to create your tracking plan from scratch, check out Avo’s definitive guide.

Level Up your Product Data Management Strategy

The best time to start improving your product data management practices was years ago. The second best time is today. The longer you put it off, the more misinformed decisions your team will make.

Improving your product data management means building new habits. The key is starting small enough that you can see the reward quickly, and trust us—the results are worth it. If you need help along the way, Avo and Amplitude are ready to partner with you on your product analytics journey.