Is Being Data-Driven a Good Thing? Data-Driven vs. Data-Informed Decisions

Updated: 5 days ago
We all want to assume that we have a data-driven organization. But what does that mean, and more importantly, is that a good thing? In a 2022 Adverity survey of 300 CMOs at small and midsize businesses in the US, UK, and DACH region, 67 percent said the volume of marketing data available had become overwhelming, and Adverity's own CMO, Harriet Durnford-Smith, said CMOs had become so bogged down by data challenges that they lost sight of their core purpose, meeting customer needs (Adverity). Being so buried in data that you aren't focused on customer needs should be considered a crisis for any marketer.
Here is the short answer. Using data is good. Letting data make the decision for you is risky. I'd rather a company be data-informed: clear about the decision first, disciplined about the few numbers that bear on it, honest about what the data can and can't tell you, and willing to combine the evidence with customer knowledge and judgment. "Track everything" is the opposite of that, and it's one of the most common ways companies fool themselves into thinking they're data-driven.

1. What does data-driven actually mean?
Being a data-driven organization means making business decisions based on data and analysis rather than relying only on intuition or anecdotes. Done well, it touches marketing, sales, operations, and finance. It helps you make more precise decisions based on evidence, find inefficiencies in your processes, understand customer behavior and preferences so you can improve their experience, spot new revenue opportunities and pricing problems, and see market shifts before competitors do.
Notice that customer needs sit right in the middle of that list. Data is supposed to bring you closer to customers. When it pulls you away from them, something has gone wrong.
2. What is the difference between data-driven and data-informed?
In a strictly data-driven process, the number decides. The dashboard says cost per lead went up, so the campaign gets cut. The test says variant B won, so variant B ships. That sounds rigorous, and sometimes it is. But it assumes the data measured the right thing, that the sample was big enough, that nothing else changed at the same time, and that the metric reflects what you actually care about.
In a data-informed process, people decide, and data is the strongest input. The team asks what decision they're making, what evidence would change their mind, how much uncertainty is in the numbers, and what the data can't see. Then they combine that evidence with what they know about customers, the market, and the business.
As a statistician, I lean data-informed for a simple reason. Every dataset answers only the question it was collected to answer, under the conditions it was collected in. Every model rests on assumptions. And once a metric becomes a target, people start optimizing the metric instead of the outcome, which is the idea economists call Goodhart's law. A data-informed team keeps those limits in view. A purely data-driven team tends to forget them.
3. What could possibly be bad about being data-driven?
The first risk is losing the human element. Experience, customer conversations, and judgment carry information that never shows up in a platform report. If the organization learns to defer to whatever the dashboard says, that knowledge goes unused.
The second risk is cost. Collecting, cleaning, storing, and analyzing data takes time and money, which is a real challenge for smaller organizations with limited resources.
The third risk is overload. A "track everything" mindset buries the few useful signals under thousands of fields no one reads, and it slows the team down instead of speeding it up. Collecting data is essential, but it only pays off when you prioritize what is relevant to your business goals. The way to avoid overload is to be clear about what data you need for each goal before you start collecting.
4. Why does "track everything" feel safe but backfire?
Frequently, CMOs say "everything" when asked what they want to monitor, without a clear link to their business objectives. That default usually comes from not yet knowing what they want or need. Tracking every possible data point produces an overwhelming amount of information that is seldom examined or used. It clutters the view, hides the important insights, and makes it harder for the team to contribute anything valuable.
The "track everything" mentality puts wants ahead of needs, and it usually comes from one of four places:
No clear key performance indicators (KPIs)
Trouble telling vital data from nice-to-have data
Fear of not capturing enough data
Not knowing what the data could do or what analytics capabilities exist
Curiosity is admirable in analysis, but restraint matters when you decide what to collect. Indulging every interest leads to large amounts of irrelevant data that never touches a KPI. As companies evolve and their digital setup changes, those nice-to-haves go stale, while the cost of collecting, processing, and storing them keeps adding up. Ironically, tracking everything "just in case" often leaves the critical questions unanswered, because the effort to dig through it all is too great, or the pile produces a distorted picture.
The goal is balance. Track too little and you can't analyze what matters. Track too much and you end up with analysis paralysis.
5. How many KPIs should you actually track?
Turning business outcomes into measurable KPIs is hard, and busy teams rarely have the bandwidth to do it well. I start from the top. Identify your primary, secondary, and tertiary objectives, then the goals that move each objective, such as increasing online purchases or form submissions, then the one or two metrics that tell you whether each goal is on track. For every KPI, there must be specific data or analytics that supports it. If KPIs and analytics are misunderstood, the match between them will be too.
My rule of thumb is 1 to 2 metrics per goal, 2 to 3 goals per objective, and roughly 5 to 7 key objectives. Here's the arithmetic. At the low end, 5 objectives times 2 goals times 1 metric is 10 metrics. At the high end, 7 times 3 times 2 is 42. That full set belongs with the teams who run the work. The executive view should hold only the handful at the top, and if your count is drifting toward the high end, treat it as a signal to cut.
A marketing KPI tree is the cleanest way to lay this out, because it shows exactly how each channel metric connects to revenue. And remember that KPIs change as industry conditions and business needs change, so tracking should evolve with them. You can't predict every future data need, and you don't have to. Focus on the key areas now and refine as you go.
6. Why do research, analytics, and marketing need to talk?
The research team is often separate from the marketing team, and the two rarely sit down together to set KPIs. I once consulted for a market research training company. When editing their materials, I noted they had no educational content about figuring out which metrics to track. The feedback shocked me: "Please remove that. It is not relevant." I still can't believe that response, yet I know how common the attitude is. Researchers collect data and feel their job is done. There's little back and forth about how the data will be used or which KPIs matter most right now.
The same gap shows up between marketing and analytics. The analytics industry keeps changing. The move from Universal Analytics to GA4 is a good example of how quickly familiar tools can shift under a team. Without understanding what each platform can do and the jargon it uses, organizations underuse some features, overuse others, and fall back on "track everything" because they don't know what to ask for. Knowing the terminology in your own platforms helps you decide when tracking something is appropriate and how to read it once it's there.
The fix is a standing conversation. Whoever designs the research, whoever builds the tracking, and whoever makes the marketing decisions should agree on the decisions the data is meant to support.
7. Where do qualitative data and testing fit?
Core platforms such as GA4, Meta's ad tools, and Amazon Seller Central are the backbone of most companies' data, but they shouldn't be the whole picture.
Qualitative data puts the numbers in perspective. Surveys, polls, ratings and reviews, NPS questions, and customer interviews tell you why people behave the way the quantitative data shows. Good market research often reveals that you need to track less than you thought, because it clarifies which behaviors actually matter.
Testing should be routine in your ad campaigns and on your site. A/B tests show which designs and messages people respond to and support personalization. Just make sure each test has a pre-set success metric and enough volume to reach a reliable answer, or you're back to reading tea leaves.
Combining qualitative input with focused quantitative tracking leads to a stronger understanding of your customers and their experience with your brand.

8. How do you know your company is data-informed?
Tracking everything usually signals that the harder work of deciding what matters hasn't been done yet. Here is a quick self-check:
Every major decision starts with a written question and the metric that will answer it.
Each KPI has an owner and a clear line to a business objective.
The executive view fits on one screen.
Reports show ranges and comparisons alongside the headline numbers.
Customer research and qualitative feedback appear next to the platform data.
Leaders can explain when they overrode the data and why.
Tracking gets reviewed and pruned at least once a year.
If you check most of those boxes, you're using data the right way. If you don't, start with the KPIs and work down. My post on data-driven marketing strategies walks through what to measure first and how to decide where the next dollar goes.
If your team is drowning in data but still can't answer the questions that matter, I can help you decide what to track, what to cut, and what to model. You can book a free 30-minute introductory call here.
Dr. Linda Orr, PhD, is the founder of Orr Consulting, a fractional CMO and marketing analytics practice serving brands from $5M to $500M in revenue. She built Upwork's first enterprise marketing mix model, covering an $88M annual media portfolio, and spent about 20 years as a marketing professor before moving fully into consulting.






