It’s been a decade since Harvard Business Review hyped Data Science as the sexiest job of the 21st Century. That was the start of the decline in the digital marketing industry.
High salaries! Big Data! Open Plan Office! Unlimited Vacay!
All my real ones know what it’s actually like on the inside.
Changing Tools! Suspicious dashboards! Unrealistic expectations! Why doesn’t the data match?!
The truth is, marketing data analyst churn is about 12-18 months. That hasn’t changed in 10 years. Depending on whether you are in-house or agency, you could find yourself stuck in meetings, struggling to balance both data collection or reporting and the analysis you want to perform on it. Often analysts are spending more time maintaining and updating software platforms than actually analyzing data at all.
Data analysts, I see you.
I spoke with a few friends in the space to vent on why our profession struggles to produce insights.
There are too many marketing tech softwares.
From 2011 to today, the MarTech options exploded from 150 tools to more than 8,000. Not only are there thousands of software packages available to create visualizations, collect data and manage a business’s digital customer touchpoints, but there are also zillions of possible combinations and integrations to tie all these platforms together. Big data and the lower cost of cloud computing has made it even easier to launch a product in our space, and all these platforms; from email or ecommerce, promise insightful dashboards from your customer data.
Let’s get this straight. A dashboard doesn’t make insights. Research does. When your check engine light comes on on your car dashboard, is that insightful? No. You need to go to your mechanic to tell you that a squirrel has been living in your engine block. That’s insightful.
The problem the explosion of tools present to the data analyst, is that their marketing very often promises some type of silver bullet, and it is up to the analyst to deliver the bad news that there IS NO SILVER BULLET.
Our friend April Moore, Senior Implementation consultant at Keystone Solutions said, “Vendors are very good at selling “solutions.” But really what they are selling are tools. And no matter what tools you are using, the job is still the same. Analyzing that data, understanding what the numbers mean, where there are gaps in the picture, and then communicating that in a meaningful way that drives action – that’s where the art is.”
Expectations on data analysts are unrealistic
“Having been in the analytics field for 20 years, I often feel like it’s a mouse in a wheel. You can put him in a different size wheel, but he’ll still just be going in circles.”
Often there is a misalignment with what clients expect and what data analysts can actually produce. Deliverables are rarely well-defined at first. The data analyst will take a crack at a solution, come back to the stakeholder who then asks them for something else, be it an addition to a dashboard or further analysis. While we’re not talking about iteration, which we at Clickvoyant are fans of, we’re talking about most situations where an data analyst deliverable doesn’t feel good enough.
“I think the thing that is a constant challenge is not having a proper balance between developing your insights platform and actually coming up with insights from what you have in front of you. Companies focus so hard on what they don’t know that they forget to really build off of what they do know. Instead they wait for “the data to become available” or “the report to be built”. Like, look at what’s in front of you and make a decision already.” said Dexter Bustarde, Director of Business Analytics at GoDaddy.
Marketing Data is Dirty AF.
I promise you, no matter how much time and energy a company spent on up-front implementations, everything goes from order to chaos. The best laid plans of mice and men…. There is no such thing as “cleaning your data” because every dataset has issues. We have to spend a lot of time massaging the data into a place that can get us to the ever changing martech landscape… not to mention conforming measurement to new trends in customer behavior.
While most marketing stakeholders will accept this truth, it does not mitigate their urgency for insights or the pressure on data analysts.
Andrew Bakonyi, Senior Director of Marketing Analytics at Walmart makes this great point, “One big issue is coming in new [as analytics leader], discovering that the implementation is crap, and then spinning up a refresh that drags on for months where all of a sudden you’ve delivered zero value but a lot of promises to stakeholders that have turned off to you.”
Some companies still don’t always know how to value data.
“Data may be the new oil, but businesses built before internal combustion engines may not value gasoline. When the firm doesn’t run a business model fueled by digital analytics, digital analysis takes a backseat on the horse drawn carriage.” said Enrique Gonzales, Director of Digital Analytics at Science/AAAS.
Maybe they’ve never started or maybe they’re in the trough of disillusionment, but many companies don’t know what insights can be gleaned from their own data, much less what it takes to drive value out of an analytics practice, and that is why they need data analyst support.
But it’s often the responsibility of the data analyst to have moxie enough to take a stand and become the leader and educator of the practice.
“Digital analytics must drive success or be stuck shoveling vanity reports or worse sucking up to management validating pre-existing assumptions.”, said Enrique.
“Businesses that have found success without integrating analytics often are reluctant to change. Staff in sustaining business are often operationally the best for their roles in terms of cost and skill set, the MVP (minimum value people) of employees, and not likely to be entrepreneurial or risk inclined. Executive leadership must sponsor this change or digital data analysts will find themselves constantly fighting skirmishes over every step toward being part of marketing or product development. If management blesses analytics with sponsorship, it is up to the digital analytics team to prove the value that data can bring to the organization.”
For sho’, Enrique. While the job can be exhilarating with the right audience, more often, this is a Sisyphean task that can leave a data analyst burnt out in less than a year.
Analytics talent market is like coastline real estate.
Compounding this whole situation is the $200B gap in the analytics talent market. Demand for analysts grows while the demand for engineers declines. There simply are not enough people with a background in analytics. That scarcity makes the data analyst role expensive and competitive to fill, and so easy for us to quit. Moreover, most analysts at the junior to mid-level know that the fastest path to higher salary is to switch companies.
Companies are fighting over small pools of talent and because there isn’t enough supply in that market, analysts who do exist can demand very competitive salaries… salaries that are increasing in the post-pandemic recovery period.
What I’m saying is; if you’re one of the lucky ones to have an analyst, you’d better hold on tight.
What we have here is a good old fashioned communication breakdown.
We need to redefine what these words mean “analyst” & “insights”.
What is an analyst?
- Is it someone in pursuit of statistically significant opportunities that help a business grow.
- Is it a person in charge of building dashboards?
- Is it the admin of Google or Adobe Analytics?
What is an insight?
- Is it “Revenue is projected to be down”?
- Is it a data visualization?
- Is it, “Customers from Sante Fe, NM have had increased purchases of air filters in the last 30 days. Shifting ad budget to this DMA may generate an opportunity for $30K in the next 10 days.”
At Clickvoyant we’re taking a stand on both definitions. An analyst is the person illuminating meaning from all data sources, bringing advice to decision makers on what actions will propel a business forward in prosperity. An insight is the mathematical evidence and rationale for the advice given.
Anyone who sees the insight presentations from our AI will know that we’ve programmed her to deliver insights that look like the Santa Fe example above. She was built on our 15 years of stumbling through all the issues you’ve read in this article. 15 years of low open rates on our dashboards. 15 years of meaningful conversations with clients.
It only matters if we show the way to El Dorado.
That said, it is humanly impossible to get to all the questions a company has, and damn near impossible to scale a team under such competitive conditions. As Directors of Analytics, we built our AI to help ourselves help more people. This made our job less repetitive, got us closer to our own definition of insights faster, and reduced the cost per insight. (new metric for the industry?)
We need to start a movement toward analysis consistency.
We’ve gotten ourselves into quite a pickle. Bad communication. Unmet expectations. Lack of empathy. No analyst fortitude. This isn’t good enough this far into Web 2.0 and on the horizon of Web3.
- A Dashboard isn’t doing it. Even self-service dashboards. They create more questions than they produce answers.
- We need to calm down and include analysts in the evaluation of martech tools. It’s them you ask for analysis after the purchase is made
- We need time to let our data stack investment create ROI
- We need to give our analysts some breathing room with stakeholder patience and augmented intelligence.
- We need to come together on what an insight is actually supposed to do and trust them when they push back on vanity metrics.
This would be a great start.
