How many tools are you currently using to monitor your data?
We’ve all been there: your team is meeting to discuss strategy, when suddenly someone says something like, “Wait, why is that install number from AppsFlyer 10% lower than the install number from Amplitude? Is that Total Installs or Attributed Installs?” Or – “Why is the Facebook conversion number so low compared to Stripe’s?” Suddenly, everything has ground to a halt, and everyone’s bug fixing in the virtual conference room.
There are lots of tools at your disposal to collect information on nearly every element of your business – and that’s the challenge. You have all the data you could possibly want, but it lives in silos and speaks different languages. Metrics use different titles, and each system calculates them differently. Your client-side sources and server-side sources are sending data to different destinations. These discrepancies make it hard to trust your data, and the sheer volume of information you’re processing is enough to exhaust your team.
So how do we go about creating a sustainable way to analyze your company’s performance across platforms?
1. Define Use Cases
Imagine your ideal customer. How do they engage with your business? What are their behaviors and habits across your digital ecosystem? Answering these questions is key to determining what engagements, metrics, and interactions you want to capture and report on. Once you’ve got a holistic idea of who you’re catering to, and how they’re using your services, you’ll be able to focus your analysis and determine where you should channel your energy.
2. Develop Roadmap
After you’ve identified specific use cases, you and your team should have a clear understanding of where your priorities lie. From here, you can develop your high level project plan. Based on the information you want to capture, determine how you’re going to gather and analyze it, then determine tasks based on teams, systems, use cases, or whatever makes sense for your business.
When organizing your roadmap, include estimates of how much time each task will take, and remember to add a 20% buffer to your schedule in case you run into a hurdle. Most importantly, make sure that everyone invested in and responsible for the project agrees with the plan of action. This is a crucial step for ensuring that everything gets done as harmoniously as possible!
To help make all this easy for everyone to understand, we like using swimlanes to organize roadmaps.
3. Create A Full Database Schema And Build A Data Dictionary
This is a critical part of the process. This is where you lay the groundwork that will allow you to interpret data from disparate systems, and create a unified ecosystem.
A full database schema is the logical framework for how you’re using your data. Mapping this out will help you determine the relationship between all of your data sets, and the places where they intersect. Make a list of all the engagements and actions that a customer (or potential customer) can take across your entire digital landscape. Then, think about the metrics, dimensions, and metadata you’re using to measure those actions. Is that data being measured by multiple systems? Take note of those intersections.
Once you’ve built your schema, look at the terms you’re using to refer to your data. Are you using the same terms to refer to that data across systems? Define unified terms, establish where systems overlap, and record all that information in a data dictionary. This process ensures that everyone on your team is speaking the same language, and it’ll save you time by eliminating miscommunication.
4. Fix Data Pipelines
After you’ve done a deep dive into how your systems speak to each other, you may discover that a system isn’t delivering data to you in a useful way. Maybe the system is sending data directly to a data warehouse instead of to your customer data platform (CDP). Or maybe your CDP is only updated after the data has been delivered to a data warehouse, preventing you from engaging and targeting your users in real time. Whatever you discover, assessing the way your data is flowing and changing your data pipelines will ensure that you’re able to make the most out of your information.
5. Improve Tool Implementation
In some cases, you may discover that you’re still using tools that are no longer a good fit for your company. If that’s the case, you may want to start searching for new tools based on what you’ve learned from the discovery process.
There are a few things to keep in mind as you contemplate integrating new tools. You’ll want to consider what will go into disassembling your current set-up - will you have to completely overhaul your existing ecosystem? Is replacing a tool worth the disruption that restructuring may cause? How will your new system be structured? Is there a way to re-configure your existing tools to better serve your needs, rather than shifting to a new platform? This is a lot to consider, but asking yourself and your team these questions will ensure that you’re building an ecosystem that truly serves your needs.
There’s no shortcut to developing and implementing a thoughtful strategy to gather insightful, actionable data. But fortunately, you don’t have to do it alone. Reach out to Pickaxe for more information on how we can help you make the most out of your data.