You nailed your proof of concept project. Your initial group of users are happy, and they’re getting comfortable mining the first batch of data for insights that may actually have an impact on your company's bottom line.
The Higher Ups are happy too. They knew in theory your company needed a data warehouse, but they’re relieved to see it's actually paying off.
So, kudos to you: getting your feet wet went well. But a vast ocean of data and analysis awaits you. It's time to start swimming—but to where?
In this article, we'll explore how to figure out how and where to go deeper with your company's data.
Although there are many ways to build out your data warehouse, here are four strategies that are especially effective at helping your company thrive.
One common strategy for going deeper is to focus on getting more fine-grained information about the data you've already been analyzing.
For example, suppose you started with basic, out-of-the-box Salesforce data. In the next round, you might target custom fields. When setting up your Salesforce data syncs and reporting, you probably skipped these fields because they get into a degree of complexity that isn't useful for addressing high priority problems. Or maybe custom fields are used for edge cases and your Marketing director didn't want her staff to start spending time analyzing exceptions until she felt like they had a firm handle on the basics. But now that the core fields are in good shape, it's time to bring in the custom ones.
Or perhaps you're ready to start analyzing cohorts. It wouldn't be surprising if you held off on cohort analysis in the first round. Cohort analysis can be extremely valuable, but tricky to pull off. There are so many ways to slice and dice your customers into cohorts—by timeframe, by marketing campaign, by customer type (e.g., tiers like premium and lite), etc. And how precisely you define a cohort can require navigating differences over strategy and other political tensions. But now that you've racked up some wins, it's time to dive in.
Another good strategy for going deeper with data is to add more departments into the mix.
Let's say your first data infrastructure project focused on information from Shopify or another ecommerce tool. It made a lot of sense to start there because the Higher Ups were most concerned about revenue data—e.g., average order size, abandoned cart flows, or the efficacy of loss leaders.
Now that you've nailed the revenue data, it might be time to bring in customer service data from a CRM like Zendesk. For example, perhaps your CEO is seeing an uptick in returns and wants to get a better handle on customer frustration. Your Shopify data can provide some high-level answers like which products are returned most often, but it can't slice and dice the data to differentiate between user personas or reasons for return. That’s a perfect job to tackle at this point in your data journey.
In your initial data infrastructure setup, odds are you tried to limit your scope to discrete, manageable slices. That's a good way to go at the beginning. In a company where information has been traditionally siloed, trying to pull it all together from the jump is like walking while tossing banana peels in the direction you're headed.
But now that you've made some progress, it's time to start building an end-to-end view of key facets of your company that are critical to your business' success. For example, now might be a good time to build a conversion funnel that will let your users dig in to identify missed opportunities and wasted resources.
When you're first getting started, it’s important to focus on problems that are on fire. But once you've proven yourself, you may want to carve out at least a little time to build out your company's ability to monitor potential fire hazards.
For example, suppose that the number of long-term active users has started to slowly but steadily decrease. It may not count as “churn” yet, but if it continues, that trend could end up hammering your company. This could end up becoming a real problem with annual subscriptions, where users may not have churned just yet but there are signs that they won't re-up—for SaaS companies, that might look like dwindling activity level...or massive data dumps in preparation for a migration away from your platform. By enhancing your data solution, you can help managers identify those warning signs (and those at-risk customers) before a spark turns into a raging fire.
Sometimes where you need to focus your efforts next is a no-brainer. For example, if there's a hot mess that Higher Ups desperately need to fix and their staff need easy access to more data to solve the problem, that's your new project.
But most of the time, the next step isn't so simple. So what you need to do is figure out how to prioritize the four strategies—or how to mix and match them.
To help you out, here's a handy chart about how to think through the trade-offs you're facing:
Your ultimate goal should be to get the best bang for the buck for your company. So for each potential project, ask yourself how much work the project will take versus how much payoff can you reasonably expect. Most of the time, you'll want a mix of quick wins and tasks that take more work but have a higher ROI.
There are no hard and fast rules here. In general, you'll want to create a portfolio of work where the odds of substantial success are high and the effort is relatively low. But there are occasions when it's worth taking on game-changing projects that are a lot harder to pull off. After all, if you can manage to pull Excalibur out of the stone, your company will become the One True King of your market.
You don't have to spend a ton of time deciding where your data quest should take you next. If you just take a step back, think about the paths before you, and deliberate a little on their risks versus versus rewards, odds are you'll find yourself on a journey that helps your company succeed and helps you build a successful career.