Our strategy and design teams spent two hours debating whether custom illustrations performed better than stock photos. We gathered for a quarterly social media readout and compared clicks, discussed color changes, and sorted posts into categories. By the end of the meeting, we still couldn’t explain what caused the small differences in performance.
Our conclusion was to keep testing next quarter. Eventually, we thought we’d found a winning layout, so we put it into production and used it regularly. Within months, our creative work felt tired and predictable.
Chasing small engagement gains had narrowed our creative choices, and we still didn’t know whether those gains helped the business. That experience left me with a question I still come back to: When does performance data hinder making better creative decisions that actually grow the business?
The appeal of green arrows in a marketing dashboard
A dashboard full of rising numbers and green arrows gives everyone a hit of dopamine. We love it when clicks, video views, and engagement are up. It makes it easy to leave a meeting convinced the strategy is working.
Sometimes a click is just a click. It might reflect genuine interest, curiosity, confusion, or an accidental tap. The number tells us something happened — but we still have to work out what the user behavior actually means.
Digital tools offer detailed metrics across every touchpoint of the consumer journey. We can track open rates, click paths, scroll depths, and dwell times, sometimes down to the millisecond. That precision often tempts us to track numbers simply because the software makes doing so easy. We can confuse activity with financial progress, treating every recorded digital footstep as evidence of business value.
Before we add another chart to our reporting, we should know what decision it’ll help us make.
10X your SEO with Semrush for Enterprise.
The world’s most powerful SEO platform, purpose-built for Enterprise.
Request demo
Digital strategy and the tortoise approach
In the classic story of the tortoise and the hare, the tortoise’s tactic was slow, steady, and intentional. There’s value in going deliberately slow — examining what you’ve done, what’s needed next, and why. A tortoise approach lets long-term goals guide your daily choices so you don’t react to every change. Move deliberately, choose useful measures, and resist changing direction every time a dashboard ticks the wrong way. Test ideas and respond quickly, but know what you’re trying to learn before you start.
If the goal is repeat purchases, ask whether customers keep coming back. If the target is improving online sales, examine what helps people complete a purchase. If you’re building brand preference, look for evidence that people remember your brand and have a reason to choose you. Those qualifying questions will take you further than asking which post collected the most clicks.
Testing can’t substitute for strategy
Looking back at that quarterly meeting, testing wasn’t the problem. We’d lost sight of the question we wanted it to answer. We compared small differences without understanding whether they reflected the design, topic, audience, or chance. Then we treated what looked like a winner as a formula.
The cost of these decisions went beyond our time spent in the meeting. A familiar layout became easier to approve because we could point to its past performance. Fresh ideas had to compete with a chart. Past results can inform the next idea, but they shouldn’t set its boundaries or limit future goals.
I’d approach that readout differently today. I’d start with the business goal, choose a question worth answering, and get a consensus on what to do with the result. If we couldn’t draw a clear conclusion, we’d acknowledge it. Another quarter of testing would need a clear purpose.
Give data a job
Imagine an online checkout where customers regularly leave between steps two and three. Perhaps the page reveals an unexpected shipping cost, asks for information customers don’t have, or doesn’t work well on a phone. The data shows us where to begin searching for answers. We can investigate the cause, make a change, and check whether more customers finish their purchase.
Marketing data can serve the same purpose. Testing a clearer product description might show whether customers understand the offer. Comparing campaign results might determine where to spend the next dollar — or where to stop spending. In each case, the data supports a decision we can then explain.
Make room for creative judgment
New ideas spark possibility. When you have a good one, it’s exciting to see it take shape. However, a new idea won’t have a performance history. We have to judge whether it fits the brand, speaks to the audience, and gives people a reason to care. Then we can test it with a clear goal.
Data helps us learn from what we’ve done. Creative judgment helps us decide what to try next. We need both, along with enough time to act on what we’ve learned.
Audit your reporting
You don’t need to control the entire marketing budget to change how your team uses its time. Start with three questions:
- What decision will this metric change? If the number rises or falls, what will you do differently? If you can’t answer, reconsider why you’re reporting on the metric.
- Is the insight worth the effort? Add up the time your team spends gathering, formatting, checking, and discussing the numbers. Which findings justified that effort? Which charts did everyone merely glance at before moving on?
- Are we giving the goal enough time? Agree on when you expect progress and how you’ll judge it. For example, a weekly report during each project sprint should help you spot problems along the way without pushing you to change a strategy that needs months to properly assess.
Start with one metric
The tortoise approach to digital strategy is deliberate. It takes patience. It asks us to focus on the goal, even when a small spike offers a quick sense of progress.
Before your next quarterly readout, choose one metric and ask what decision it helped you make. Then do something valuable with the answer.
If nobody can name a single decision influenced by a given metric, consider leaving that metric out for a quarter. Better to choose one metric that leads to a useful decision than pulling information from 20 charts you don’t know how to act on.
