That product line I launched because it felt right is now $18,000 of dead stock. My gut told me it was a sure thing. But my gut wasn’t reading customers. It was reading competitor Instagram posts.
Why most e-commerce “data habits” fail right now
I used to check my Shopify dashboard every morning. Traffic, conversion rate, session count, it felt like a disciplined start. It was security theater. The ritual never changed a single buying decision. It delayed the real work of challenging my own assumptions until the money was already out the door.
The mistake I kept making was confusing dashboard checking with thinking empirically. One is passive. The other forces you to hunt for what might prove you wrong before you spend.
What is empirical thinking?
Empirical thinking means making the next call on evidence I can observe inside my store, not on intuition, competitor mimicry, or the last podcast I heard. I write down the core assumption. Then I go looking for the smallest piece of data that would contradict it. That’s it. Not philosophy. Just a daily habit of separating belief from signal.
Before I started this, I believed I was already empirical because I tracked metrics and could quote my top-selling SKU from memory. Knowing numbers is not the same thing as challenging your own beliefs. If the pre-decision step has never been “Here’s the single number that would prove me wrong,” then I was still running on instinct with a spreadsheet open.
The reveal invert: what most people do
Whenever a hunch showed up, I’d open Shopify and look for supporting evidence. If I thought navy hoodies were trending, I’d find three recent navy orders and feel validated. I called this analysis. It was confirmation bias with a dashboard open.
A supplement store operator I know ran $4,200 in click-optimized ads because the traffic chart looked healthy. Ninety days later, returning customers drove 73% of actual revenue. The traffic obsession was a vanity metric. The spend bought noise, not profit. That’s the tax you pay for feeding gut feelings with a dashboard and calling it data-driven.
The 20% move that actually changed things for me was a five-minute pre-decision check. Before any spend over $100, I write three lines on a notecard.
- The one assumption I’m betting on.
- The smallest piece of evidence that would prove that assumption wrong.
- Exactly where in my store’s data I can find that evidence today.
The physical handwriting isn’t cosmetic. The friction separates impulse from evidence and catches expensive assumptions before they become dead stock.
A homeware store doing $35,000 a month applied this exact check to their next inventory buy. The assumption: “Customers want earth-tone ceramic mugs because Pinterest trends show them surging.” The disconfirming evidence they committed to find: a single support ticket asking for those colors, or a session recording of someone lingering on earth-tone product pages. Ninety minutes of recorded sessions later, they found zero, but spotted three users searching for stackable storage. They pivoted the buy. The stackable line sold out in 11 days. Without the check, they’d be sitting on 600 unsold mugs right now.
How do you apply empirical thinking to daily e-commerce decisions?
I pause before any inventory buy, ad campaign pivot, or pricing change and ask: “What’s the one fact I’m treating as true, and what’s the smallest counter-evidence I can find in the next 24 hours?” The job is to hunt for what would prove me wrong, not for what makes me feel clever. That flips the relationship with data from comfort-seeking to truth-seeking.
The notecard system is what I fall back on. Before a decision over $100, I grab an index card and write the three lines:
- The assumption I’m funding.
- The smallest contradictory evidence I’d accept.
- The exact report, session recording, or customer email folder I’ll check within 24 hours.
I keep the cards. I review them after the outcome unfolds. Within two weeks, I found that my “obvious” read of the market was wrong at least a third of the time. The pile became more honest feedback than any dashboard screenshot.
A WooCommerce apparel store with $80,000 monthly revenue used this for ad creatives. The assumption: “Lifestyle photos with green backgrounds outperform studio shots because they feel authentic.” They wrote the contrary evidence as “any split-test result where the studio shot beats the lifestyle shot by 10%.” They pulled Facebook Ads Manager, ran a lookback, and found three instances where studio shots won by 12% or more. The assumption had been costing them roughly $900 a month in wasted spend. The check took six minutes.
What’s the hardest part of thinking empirically?
The hardest part is the quiet emotional kick of being wrong. Nobody opens their store analytics hoping to find out they wasted money. People open it hoping for confirmation. I’ve done it hundreds of times. The temptation is to notice the numbers that agree with the gut and dismiss the rest as noise. Empirical thinking forces you to go looking for the uncomfortable fact. That’s rare. That’s also why it works.
I ran a 90-day experiment a few years ago where I tracked every business decision over $50 using these notecards. The first week felt clean. By week three, I caught myself skipping the “contrary evidence” box on cards I was emotionally attached to. I’d write “check ad performance” as evidence, then quietly look only at ROAS and ignore customer acquisition cost entirely. Real empirical thinking only took hold when I started writing the exact metric and the exact rejection threshold before opening any report. That one change, forcing a specific number that would kill the idea, shifted more outcomes than any strategy book I’d read.
The people I assumed were the most empirical in my network often turned out to be the opposite. They had polished rationalizations and dashboards that told a good story. The strongest signal of genuine empirical thinking turned out to be openly admitting ignorance. Phrases like “I don’t know that yet, but I can find out by tomorrow” separated operators who made money from the ones still defending bad buys twelve months later.
What realistic results come from a daily empirical thinking habit?
Within thirty days of consistent use, I saw a measurable drop in impulse inventory buys and a faster kill rate on underperforming ads. I didn’t need more data. I needed a tighter loop between assumption and disconfirmation. Stores that adopt a written pre-decision check typically reduce dead stock by fifteen to twenty-five percent in a single quarter, not by buying better but by stopping the worst buys before they happen.
The first week feels unnatural. Your brain wants to skip the card and act. By week four, the friction becomes a felt need, a missing step that alerts you when you’re about to move on feel. By month three, I noticed a strange thing: relief when a hunch got disproven early. Relief because I realized I’d just avoided another four-thousand-dollar mistake. That emotional flip, from defending a belief to appreciating its death, is the real transformation.
A Shopify pet supplies store at $25,000 a month tracked every notecard decision for sixty days. They found fourteen of thirty-eight assumptions were flat wrong. Two of those would have cost a combined $11,000 in inventory. The cards took less than two hours of effort. The return per minute spent was higher than any marketing channel they ran that year.
Closing
The data inside your store right now holds the next bad decision you haven’t made yet. It sits in session recordings no one watches, customer service threads no one reads, and margin reports no one pulls before ordering stock. The missing piece isn’t a new dashboard. It’s five minutes of writing down what you believe and hunting for a disproof before you spend.
Take three index cards this week. Apply them to the next three decisions over $100. Write the assumption, the contrary evidence you’ll seek, and the data source. Follow through. Keep the cards in a drawer. Pull them out in thirty days. The pile will tell you more about your thinking than any analytics tool ever will.
