TLDR: Here are the 3 key skills – read on to learn why.
- A consultative mindset
- A solid technical background
- The ability to distill complex information into digestible insights
The key to profitability is predictability. It follows then, that the purpose of digital analytics is to enable business decision-makers to take action based on predictable outcomes – everything else is in service of that goal.
Businesses invest BIG $$$ in technology vendors that promise to unlock this potential – from website analytics platforms like Google Analytics to CRMs like Salesforce and data management solutions on AWS and Google Cloud. Typically this includes an investment in a reporting platform like Tableau or Domo, and a vague goal of enabling ‘full-funnel marketing attribution’ to align marketing spend to pipeline revenue.
The trouble is, as the marketplace for business intelligence continues to evolve, the data output by said technology grows exponentially. The result is often an underutilized (and poorly implemented) tech stack and a wealth of data and “insights” that are borderline worthless. GIGO – wasted potential.
Too often, organizations opt to invest in additional platforms (or reflexively migrate to a new stack), rather than investing in the expertise necessary to leverage existing platforms that may already be in place.
I’m talking of course about investing in analysts. Whether you’re hiring an analyst or looking to break into the digital analytics profession, I’m highlighting what I think are the most important qualities to possess.
1. A Consultative Mindset
“Hey Jack – give me a bar chart that shows our revenue over the last 12 months.”
– Jack’s Manager
Often, the most damaging thing an analyst can do is to do produce exactly what is asked for. Anyone can push buttons – uncovering and understanding the business question behind an ask is the first and most critical step in providing meaningful, actionable intelligence.
What is this bar chart going to enable? Not uncover – enable. Recall our opening – “The purpose of digital analytics is to enable business decision-makers to take action…”
Thinking about data in terms of action re-frames the question: is “revenue” the right metric?
After some probing, it turns out that Jack’s manager is looking to estimate how much marketing spend needs to ramp up to hit the team’s quarterly revenue goal. The trouble is, our hypothetical revenue is generated by a marketing pipeline that may have started some 90 days earlier.
In this scenario, Jack’s manager would be better served by a lead cohort visualization, so that months with increased marketing investment can be linked up to the months where those investments finally paid off with revenue.
To recap: a consultative approach is the key to uncovering the underlying business questions that insights, data, and dashboards should be answering.
2. Solid Technical Background
Before we get any further: what I’m not talking about here is expertise in specific systems and platforms. Yes, it can be helpful to have pre-existing knowledge in platform X or Y when onboarding to a new business and project – but most platforms can be learned far more quickly than foundational concepts in data.
Given that most organizations agree it typically takes anywhere from six to eight months for a new employee to become fully productive – what’s more important, specific platform knowledge, or the core concepts that can be leveraged regardless of platform?
Thinking About Aggregations
What does a single row of data in a table represent? The answer to that question forms the underpinning of any transformations, visualizations, and most importantly insights that are going to come from the data in that table.
Too often, I see folks with a self-professed “expert-level” knowledge of pivot tables in Excel make poor assumptions about the underlying representation of the data – leading to situations like aggregating “daily unique users” across a month, or reporting on an average of averages.
Expertise in Excel might represent advanced knowledge in a particular platform – but that’s not the same thing as an understanding of data. To paraphrase many, many people: any analyst worth their salt has a solid understanding of SQL.
For my money, it’s not just that SQL and it’s various iterations form the underpinnings of most data wrangling and visualization platforms. Working with SQL builds a foundation of how to think about data, what a single row in a table represents, what it means to take a sum of a metric column to aggregate a dimension over many rows vs. performing an arithmetic operation across a row.
A general understanding of data structures, what it means to manipulate them, and what they represent when visualized in a chart or dashboard is far more important than the specific knowledge of how to create a chart in a particular interface.
3. Ability to Distill Complex Information
Say it with me: dumbing things down is not distilling. Distillation is the act of extracting essential meaning – that is not the same thing as removing otherwise important context.
What good is knowing that sales are down Quarter-over-Quarter if you ignore that the previous quarter included a big product release?
The analyst should always be trying to distill a problem down to the core underlying insight, and getting that point across with as little noise as possible. Sometimes this skillset is called “storytelling”, but I find that can be a bit misleading.
“Stories” bring to mind fairytales and sweeping epics – in a business context, the story should at most be a collection of insights that form a hypothesis: “By undertaking initiative A, we have impacted our business as measured by B because of C”.
Wrapping Up
An understanding of data concepts, the ability to distill information and communicate insights, and a consultative mindset are the critical components of the digital analytics function – at least according to me 🙂
Feel like I’ve left something out? Let me know in the comments below!