Should I focus on getting better at Google Analytics?
What did it mean to be “good with data”?
So while I dabbled over the years, I failed to really make progress on my rather vague goal of becoming a more quantitative marketer. Then about two years ago, my boss and CEO said very clearly to me (and very kindly) that if I wanted to advance to the next level of my career (going from Director to VP) I had to focus on my quantitative skills. The time had come.
Fast forward, I was promoted last year. While I would certainly not brag about my quantitative skills, I have learned a ton about how to use data to make sense of complexity. That journey is what I want to talk about today.
Four attitudes toward data
I read a wonderful article recently that created clarity on my own journey with data: Beyond One and Zero . In it, Stephen Bailey focuses our attention not on the technical aspects of data, but on overall attitudes toward data. I love this framework and immediately recognized that I spent time in each of these quadrants.
Stephen Bailey’s representation of the four attitudes towards data.
Phase 1: The belief that data holds The One Correct Answer
When I first joined a data startup in 2013, I was an Optometrist. Stephen describes this person as being an advocate for measurement who doesn’t have a grasp on the complexities of working with data.
They are more concerned with building consensus than understanding complexity. When I was an Optometrist, I liked to say things like: “We’ll just see what the data says!” or “Let’s just test it.” I believed there was One Correct Answer, and that the data could reveal it.
When I was given a metric like “Marketing Qualified Lead” or “Pageview”—I trusted it completely. In fact, I felt frustrated when someone explained the nuances of how the data was created—why did the details matter? They had pulled it from The One Correct Answer place, right?
Seth Rosen’s tweet on data .
Over time, I found myself increasingly frustrated by the inability to get at the data that had The One Correct Answer. The data always seemed to have an issue or required all this work to get it. Why was it so hard? Why was data so unreliable? I stopped believing that it was worth the effort.
Phase 2: The belief that data isn’t worth the effort
Crushed by how data failed me, I became a Nihilmetricist. I now knew the data didn’t have The One Correct Answer, but I also knew I couldn’t just reject data.
Stephen describes the Nihilmetricist as someone who uses data for show. They appear with an abundance of data, dashboards, and ever-changing metrics. But the abundance of data only further obfuscates. Rather than creating clarity, the Nihilmetricist overwhelms you with data. And that’s the point.
Fine, I thought. If I can’t get the data I need to understand how my work impacts the business, then I will just look at all the data. This month I’m reporting on Pageviews and next month I’m bragging about Time On Site. Right now we care about Total Mentions and next month I’ll be scolding you that we can’t forget about the “quality” of those mentions.
A lot of marketers that appear “quantitative” at first glance are really Nihilmetricists, more concerned with proving the value of our job than creating clarity about how to drive growth. Jimmy wrote a post a few years ago called “ Forget About Traffic and Start Using Content to Drive Leads and Sales .” This is a post to the Nihilmetricists, the marketers out there blindly chasing an impressive stat, with zero interest in being held accountable to business goals.
While existing as a Nihilmetricist works just fine as an individual contributor, they make for frustrating managers. As I began to manage people , I felt the urgent need to create clarity. It wasn’t useful to just throw a bunch of data at my team—we needed focus. People needed to know if they were doing a good job.
Phase 3: The belief that qualitative data is all that’s really needed
This is when I entered my Skeptimetricist phase. A Skeptimetcist understands that quantitative data is challenging for many reasons, so they advocate for a more qualitative understanding of the world. I used research! But it was less quant-focused.
For example, when I was building my little communications team at HubSpot, I used LinkedIn to understand the size of a communications team vs. the overall size of the organization at companies like ours. I did a lot of interviews with other communications leaders I could learn from. I was able to build a sense of order using more qualitative tools.
Honestly? This isn’t a bad approach! Qualitative research is enormously helpful in making sense of complexity, and this mode of operation served me well for quite some time. Still does!
But Skeptimetricism only goes so far. Stephen writes: “Skeptimetricism is understandable, especially for a business leader. The world he sees is complex, but prolonged skeptimetricism is untenable. At some point, he must take ownership of how his team views the world. It is too inefficient to organize otherwise.” Quantitative data is a powerful way to make sense of the world. As teams grow, data becomes an incredible tool to help large groups of people create a shared sense of reality.
This is exactly where I was when my CEO said, “You have to work on your data skills.” My team was growing. I had hired the first six members of the marketing team, and would more than double the team in the following year. My qualitative approach worked for a small team, but it didn’t scale well. It required a lot of explanation from me.
But most importantly, it failed to create accountability for my team. Was the marketing team doing a good job or a bad job? No way to tell! That is not a recipe for reporting to the CEO, earning the trust of your board, or building a high-performing team .
Phase 4: The belief that data is absolutely critical, while also understanding all of its shortcomings
As I started to engage with this process, I entered my Pessimetricist phase.
A Pessimetricist advocates for measurement while also knowing that data has flaws and potholes. The Pessimetricist cares deeply about “making a complex reality knowable,” while also understanding this work isn’t ever really complete. We are always working with imperfect data.
Here’s an example. Our North Star Metric at dbt Labs has always been Weekly Active Projects. This is roughly the total number of companies using our product, dbt, every week. This is the chart my boss showed me over three years ago when he convinced me to join the company, and I have looked at it every week of my life since I joined.
Weekly Active Projects at dbt Labs.
I know all (or most of) the flaws with this chart. I know that a project isn’t exactly 1:1 with a company because a single company can have multiple projects. I know the issues of tracking open source adoption. I know all the underlying data quality issues that go into this chart. But in spite of these flaws, this chart is still the single best way for my team to understand the overall health of dbt’s adoption today and the potential for future commercial growth.
We can segment this chart by region, by integration type, by our open source product versus our commercial product and gain a rich understanding of how, where, and who is finding success with dbt.
To an outsider, this appears to be a nice, clean chart showing up and to the right growth. To an insider, this chart is deeply flawed, suspect, and yet STILL 10 times better than any other approach to understanding traction.
This pattern appears in every situation.
Yes, we need to improve our attribution model, even though we know that attribution will never give us the complete picture of how marketing activities impact revenue.
Yes, we need to rely on our campaign reports to understand performance, even though we know that the accuracy of those reports is dependent on humans executing a process perfectly—from proper UTM tagging to SDR follow-up—and humans never execute perfectly.
Yes, we need to benchmark our key performance metrics against other companies even though we know those benchmarks are imperfect and we lack the full picture of how they were calculated.
It is through the process of working with data—understanding how a metric is calculated, an opportunity is bucketed, a piece of data is collected—that we build our internal understanding of reality. There is no One Correct Answer that exists in the data. There is no perfect data set that we can just pull from. There is no test we can run that will tell us what decision to make. There never will be. The process is the point.
When we engage with data, build metrics, group things, find patterns—we are not trying to find The One Correct Answer. We are trying to make sense of a complex reality.
Impactful quantitative marketers help their teams make sense of a complex reality.
What’s your attitude toward data? What would you like it to be?
In order to progress in your career—to go from an individual contributor optimizing a single metric in Google Analytics, to leading a team or a company—you have two choices:
Join a company that shares your attitude toward data. There are lots of companies in the world that successfully operate in all of these zones. You don’t need to work at a Pessimetricist data company like I do.
Figure out how to use data to make sense of a complex reality. Some solid spreadsheet skills will serve you well here. You likely don’t need to learn SQL (though it won’t hurt!). But what you MUST do is put in the time to understand where your data came from, how it has been cleaned and transformed, and how it maps to the reality you live in every day.
Data at its most impactful is about creating a shared sense of a complex reality. Shared is the key word here. There is no perfect data set that contains The One Correct Answer. In other words, becoming a more quantitative marketing leader requires both quantitative skills and collaboration skills to help people understand a new way of thinking about the world.
Here’s the good news—you’re a content marketer! You spend your days collaborating with subject matter experts and bringing clarity to ideas. You likely already enjoy the work of making sense of a complex reality. All you’re really trying to do is add another tool to your existing skill set. I’ve found my own journey toward becoming a more quantitative marketer to be extremely rewarding. I hope you do as well.