We’ve always been customer-obsessed at Levels, from talking to thousands of them one-on-one to learn more about their wants and needs to having everyone take a shift with our support team as part of their onboarding. So, when I first joined the company in the summer of 2021, one of my most essential directives was to come up with ways we can better collect feedback and data about how our members are using products then organize and disseminate it to the team in a way that informs product decisions.
With 25+ years of experience doing business intelligence, analytics, and user insights work for companies like Intuit, Adobe, Zynga, and Google, I certainly have plenty of best practices I’ve picked up along the way. I could have marched in here, implemented all the tools and systems I developed at those other companies and rested on the laurels of work I’ve done before — which is exactly what I see plenty of people do when they move from company to company.
I think that’s absolutely the wrong approach if you truly want the most valuable insights for your brand. Every company has a different DNA, different qualities that make it operate well. Moreover, every audience has different approaches that will resonate more. Rather than copying what you’ve done before at other companies — or what you’ve seen competitors doing — often the better path is to figure out what methods best engage your customers and best allow your team to take the data and drive action from it.
Sometimes that means letting go of metrics and processes that have worked well for you in the past. Sometimes that means going in the face of the status quo. And, yes, it means a lot more work experimenting and iterating than if you were to go the tried-and-true route. I promise it’s worth it for how it will help your company grow in the right direction.
We get asked by other companies all the time about the metrics we gather, and while we’re happy to share them, I’ll do you one better. Read on for the exact process I use for figuring out the proper analytics process at each company I join and some of what that process has taught us at Levels.
1. Start Sharing the Data You Have, Stat
When building out an analytics strategy, there’s a common approach I see companies take: spend months and months asking every employee what metrics they want, create a huge requirements document, and then take months or more to build out every ask before implementing anything.
I think this is the wrong way to go about it for several reasons. First, there’s a little bit of Steve Jobs’ “people don’t know what they want until you show it to them” at play. While, yes, sometimes I will get a specific data question from teammates, they’re often not sure how to answer the question of what data they want and how they want it presented. After all, they’re looking to me as the expert of what would be most valuable. Second, there is nothing more frustrating than a team member giving you a specific metrics request and you replying that you don’t have access to that piece of data. Finally, if you take too long to deliver data requests because you’re trying to collect input from everybody, people may forget what they asked for by the time you present it — or have different needs entirely.
I take a different approach. I like to start by looking at the data we already have access to and the various ways we can make sense of it. This is great because it allows me to make an impact quickly upon joining a new company (something that’s especially critical in the world of startups). But, more importantly, sharing whatever insights I’m able to gather from the data in front of me helps me understand through experimentation what will be most helpful to the team moving forward. This leads me to my next step:
2. Analyze How Teammates Interact With Your Analytics
In the same way that you’re constantly analyzing how users interact with your product, you should analyze how internal stakeholders react to the data you’re collecting and reporting. Are they engaging with it? Do they have questions about it? Are there ways that it’s not serving their actual needs?
I don’t actually survey my teammates about how they feel about my data or anything formal like that. Instead, I’m paying attention to micro bits of feedback in how they’re interacting with the insights I’m giving them. So, for instance, if I produce a deck and present it to the team via a Loom video, I’m looking for feedback like:
- How much of the company views it, and which teams are they on?
- What type of questions are they asking? Do they seem confused?
- Where do they give a thumbs up?
- What areas of the presentation seem to have no reaction at all?
Even something as simple as emoji reactions can help me understand what areas my teammates find valuable.
With this information, I can start to make minor adjustments to what I’m doing, doubling down on the areas that seem to have the most engagement and reducing or even cutting out metrics that seem less valuable (for instance, maybe only reporting that metric once a month instead of weekly).
I think of it as being a chef at a restaurant. You don’t ask people who walk in, “What do you want to eat? I can make anything!” You curate a menu based on your expertise and the ingredients available to you, and then adjust your menu based on what people are ordering.
Another great part about just starting to deliver data and seeing what resonates is that it helps my teammates better envision what other insights they want. They can reflect on what’s missing, where it might be valuable to dig deeper, and the like — giving me real-world insights into other tools, surveys, or data-collection methods that would benefit this company.
3. Never Settle With Your Data
And then, you just keep on doing this. Especially in today’s fast-moving world, I don’t think you can ever build one master dashboard of metrics and be done forever. Your company is a living, breathing thing. When you pivot, add new products, get the budget to buy a powerful new tool, or just have something new the team wants to learn about its members, you have the opportunity to iterate on what you’re doing to deliver better insights.
Sometimes that even means going in the face of the status quo. For instance, we’re currently debating whether net promoter score (NPS) is even helpful for us. Just because it’s well established doesn’t mean we should use it. We have to ask the question: Does it work for Levels? Is there a better way for us to gauge the information we need? If that means it comes down to a thumbs up or thumbs down as the metric we use, I think it’s better to embrace that than cling onto a popular process that isn’t the right fit. And then we have to keep watching it to make sure it’s the metric that works weeks, months, or years down the line.
There is no one right metric — only the right metric for your company at the current moment. The journey of staying curious and exploratory with data, constantly experimenting and iterating, and continually trying to push the envelope is half the fun of the job.
