Great Sales Month, Worst Business Decision?

This is a composite example based on patterns I have seen across several companies. Numbers and details are illustrative.

Record traffic. Orders pouring in. A team working overtime on shipments. By every visible metric, the biggest sale of the year at a mid-sized furniture e-commerce company had been a massive success.

"So why do I feel like we just made a mistake?" the owner asked.

The numbers tell one story and your gut whispers another. What looks like success doesn't feel sustainable. If you run an e-commerce business, there's a good chance you'll recognise yourself somewhere in this.

The surface: everything looks great

Traffic was up significantly. Conversion rates had held steady, which with increased traffic felt like a win. Marketing spend had been higher than normal, but that was planned. They had invested heavily in paid social, run coordinated email campaigns and tested new influencer partnerships.

On paper, the playbook was working exactly as intended.

Then came the uncomfortable details:

• Customer service was handling unusual volume.

• The warehouse shipped clearance items that had been sitting for months. Good for moving inventory, but not their money-making products.

• Cash flow projections for the next quarter were tighter than they should have been after such a strong period.

"Just the normal post-sale slowdown," they reasoned.

But the feeling persisted.

The investigation: what the data actually said

This is where I start: looking at what they actually measure beyond revenue.

Do you track where your customers come from? Yes, they had Google Analytics.

Do you know how much it costs to acquire a customer by channel? Sort of. They had a blended average.

Do you measure the loyalty of new customers? Pause.

"We haven't really looked at that yet."

And there it was. They had optimised for the event without any way to measure what happened to brand and customer loyalty.

I pulled their Shopify transaction data into BigQuery: every order, every customer ID, every timestamp, every discount code. Then we layered in marketing spend by channel and their GA4 behavioural data.

The first query was simple: cohort analysis. Group customers by the month they first purchased, then track their behaviour over time.

The pattern appeared immediately. Customers acquired during the sale had a repeat purchase rate of about 12% within 90 days. Customers acquired in regular months? Closer to 35%. Sale customers almost never interacted with the brand: no reviews, no social likes, no newsletter signups. Among the other customers, 42% did.

Customers from regular months behaved differently. They browsed new arrivals. They bought complementary items, a side table to match a chair or a rug to go with the sofa. Some even upgraded, buying a higher-priced item the second time than the first.

The hidden cost: marketing that works against you

When we connected marketing spend to the customers it actually brought in, the picture sharpened. During the sale, paid social acquisition cost had jumped to nearly three times the usual level. Not because the ads performed poorly, but because every furniture retailer was running a sale and bidding up ad costs. The customers they attracted were comparison shopping for the best deal.

This is where the Kamben Insight Model becomes useful. Looking at the case through that framework revealed the core tension: short-term revenue growth versus long-term customer value. High traffic and conversion during the sale masked a deeper problem. These weren't customers with a relationship to the brand. They were deal hunters, and the relationship was purely transactional.

The framework: what you should measure

Customer acquisition cost (CAC) by channel. Blended CAC hides the truth. Some channels bring customers who become advocates. Others bring customers who churn after one purchase. Divide a channel's total spend by the new customers it acquired, then track what those customers do next.

First purchase versus repeat purchase. During the sale month, 78% of revenue came from first-time buyers. In a regular month it was closer to 50/50. New customers are expensive to acquire. Repeat customers are profitable.

Repeat purchase rate by cohort. Track it by cohort, the month or campaign a customer came from. For furniture, 90 to 180 days is reasonable. For faster-moving goods, 30 to 60 days works better. It tells you whether you are getting better at attracting the right customers.

Engagement signals beyond purchase. For furniture especially, the purchase cycle is long. Look for earlier signals: newsletter engagement, social follows, time spent on new arrivals rather than sale items, complementary purchases, reviews and referrals. They predict future value before the second purchase happens.

Full price versus discount. Customers whose first purchase was at full price or a minimal discount had a repeat purchase rate three times higher than those who came in through deep discounts. That doesn't mean never run sales. It means understanding that the customers you acquire through sales are fundamentally different.

The outcome: strategy over tactics

With this visibility, the conversation changed. The question was no longer "Should we run another sale?" but "What kind of business are we trying to build?"

If the goal is sustainable growth with healthy margins, acquisition strategy has to shift. Smaller, more targeted sales for loyal customers only. More investment in content and organic channels that attract browsers, not hunters. Referral programmes that activate your best customers. Email segmentation that treats discount buyers differently from full-price buyers.

Or, and this is the harder conversation, it means accepting that some customers aren't worth acquiring. That a smaller sale with fewer orders from higher-intent customers is better business than a massive blowout that fills your warehouse with one-time buyers.

The data gives you the clarity to make that decision consciously.

What you can do tomorrow

You don't need a complex data warehouse to start. Export your customer transaction data. Sort by customer ID. Count how many customers have made two or more purchases. That's your baseline.

Filter by acquisition date. Compare customers from your last major sale with customers from a normal month. If there's a big gap in repeat rates, you've found the signal.

If you want to go deeper with BigQuery, cohort models and predictive lifetime value, that's where I come in. But you can start seeing the patterns now with the data you already have.

Closing

The record month wasn't the company's worst business decision, but it was a warning sign dressed up as success. Without the data visibility to read that signal, they would have doubled down on bigger sales and more paid ads. Instead, they're building something different. Something more sustainable.

The question for you: what signals is your data sending that you're not reading yet?

---

Curious whether your campaigns are building the business you think they are? Test your site with Kamben Kompass and get a free analysis in two minutes. Or reach out directly.