Two business owners are making the exact same mistake.
The first hasn't opened Google Analytics in months because she knows her tracking is broken.
The second has never installed analytics at all.
One thinks her data is too flawed to trust. The other thinks she has no data whatsoever.
They're both wrong.
They've absorbed the idea that data only counts when it's complete, clean, and collected by the right tools. I call it data shame — the quiet belief that your numbers aren't good enough to look at, so you avoid looking altogether.
Here's what I've learned from years of teaching marketing and even more years of working with real businesses: the businesses that grow aren't the ones with the cleanest data. They're the ones who act on the data they had.
.jpg)
Where "perfect data" thinking comes from
Most analytics best practices were developed for organizations with dedicated analysts, engineering teams, expensive software, and millions of website visitors. Universities naturally teach those standards because they're statistically rigorous. Marketing instruction — I've been there — involves emphasizing the importance of sample sizes and confidence intervals designed for national brands.
Small businesses inherited the standards without inheriting the resources to implement them. As a result, many small businesses become remiss about their data out of frustration. Truth is, these standards are not as essential as one might think.
Let's set a more useful set of standards.
What "good enough" actually means
Your data doesn't need to be complete. It needs to be appropriate for the decision in front of you. That's the whole standard. Three principles make it practical:
Consistency beats accuracy. Data measured imperfectly—but consistently—still reveals trends. If your website traffic is consistently undercounted by 15% because of a tracking quirk, it's off by roughly 15% every month — which means the month-over-month movement is still real. A slightly bent ruler still tells you which child grew taller this year.
The decision sets the bar. Choosing between two home page headlines requires far less rigor than choosing whether to sign a lease on a second location. Most of the decisions in front of a small business are the first kind — reversible, low-stakes, weekly. Imperfect evidence will suffice for a quick decision.
Small numbers still have shape. "We only get 40 visitors a week" sounds like nothing until you notice that's 120 a month — and that 60 of them arrived through one local directory listing you forgot you had. Forty visitors a week may not be statistically significant in a research paper. But it can be more than enough to decide which promotion to run next week.
Patterns emerge from small numbers faster than people expect. You're not trying to publish a study. You're trying to spot which direction the wind is blowing.
You're already collecting data — you just haven't called it that
This is the part for the second owner, the one with "no data." The owner has plenty of information about customers and past transactions or interactions. It's just not in a dashboard.
• The register. Daily sales by day of week. Average ticket. What sells together. That's a dataset, updated automatically, going back as long as you've been open.
• Google Business Profile Insights. How customers found you, what searches brought them there, and whether they called, clicked, or asked for directions.
• The calendar. Bookings, no-shows, the seasonal rhythm you can feel in your bones but have never written down.
• The phone and the inbox. What people ask before they buy. And the single most underrated question in marketing: "How did you hear about us?" Ask it consistently and you've built a surprisingly useful attribution model that many funded startups would envy.
• Your own eyes. Which window display makes people stop. When foot traffic swells and dies. Which regulars stopped coming, and roughly when.
A notebook updated consistently is a database. The same principle applies here as online: consistency beats accuracy. You don't need a tool to start measuring. You need a habit.
When imperfect tips into unusable
I'd be doing you a disservice if I pretended all imperfection is fine. Some problems genuinely poison the well:
• Double-counted tracking — one customer interaction being counted as two, quietly inflating every number
• Not counting important transactions — you're measuring traffic but not the thing that pays the rent
• Filters excluding real customers — a setting meant to hide your own clicks on your website that ends up hiding actual buyers too
• Cross-domain tracking broken — customers move between your website and another domain, but your analytics loses the connection, making one journey look like two.
Here's the line I draw: imperfect data is workable, unknown imperfection is not. It's fine that your ruler is bent — as long as you know it's bent, and roughly how. What you can't do is make decisions with a measurement error you've never identified. Once you know how your data is wrong, you can read around it. Until then, you're navigating with a compass that might be pointing anywhere.
Start where you are
So here's the reframe I'll leave you with: you don't need better data to start. You need to start in order to learn what "better data" would even mean for your business.
If your tracking is messy, open the dashboard anyway and watch the trends — the bent ruler still measures growth. If you've never installed a tracking tag in your life, you're not behind. You may be sitting on a pile of receipts, conversations, and rhythms that already tell a story. Grab a handful and start the process of putting them into a spreadsheet on your computer.
And if you'd like a second set of eyes — someone to tell you exactly how your data is wrong, what's trustworthy, what isn't, and what the smallest useful fix would be — that's precisely what a GA4 Health Audit is for. It works whether you have three years of messy tracking or none at all. Knowing where you stand is the whole point.
Perfect data is a luxury. Good-enough data, looked at honestly and consistently, is a strategy.
