The 5 ecommerce metrics that actually matter (and the 10 that don’t)
Your dashboard has 50 metrics. Five of them drive decisions. The rest are noise. Here’s how to tell the difference.
Jakob Sperber
Director
Bussines
/
Every ecommerce platform, every analytics tool, every agency report is packed with metrics. Revenue. Sessions. Conversion rate. ROAS. CTR. CPC. Bounce rate. Average session duration. Email open rate. The list goes on.
More data should mean better decisions. In practice, it means more confusion. Founders stare at dashboards full of numbers and still can’t answer the basic question: is this business getting healthier or sicker?
The problem isn’t a lack of metrics. It’s that most metrics are descriptive (they tell you what happened) rather than diagnostic (they tell you what to do about it). Here are the five that actually drive decisions — and the ten you can safely ignore for daily operations.
The 5 that matter
1. CM2 per order (contribution margin after variable costs)
Your margin per order after COGS, shipping, payment processing, and fulfilment. This is the foundation. If you don’t know CM2, you don’t know what you can afford to spend on acquisition, and every other metric is floating in space.
Why it matters: It sets the ceiling for every acquisition and growth decision.
How often to check: Monthly, by product and by channel. Update when costs change.
2. New customer CAC (not blended)
The actual cost of acquiring a first-time customer through paid channels. Not the blended number that includes returning customers — the real acquisition cost of someone who’s never bought from you before.
Why it matters: Compared to CM2, it tells you whether acquisition is profitable. Compared to 60-day LTGP, it tells you whether your payback is sustainable.
How often to check: Weekly, by channel.
3. CAC payback period
How many days until a new customer’s cumulative gross profit covers their acquisition cost. This is your growth speed limit — the metric that connects marketing spend to cash flow.
Why it matters: It tells you how fast you can reinvest. A 30-day payback lets you cycle capital 12x a year. A 90-day payback cuts that to 4x.
How often to check: Monthly, using cohort data.
4. Repeat purchase rate (60/90 day)
The percentage of customers who make a second purchase within 60 or 90 days. This is the clearest indicator of product-market fit and the single biggest driver of long-term profitability.
Why it matters: Repeat customers have near-zero acquisition cost. Every percentage point increase in repeat rate compresses payback and expands margin.
How often to check: Monthly, by acquisition cohort and channel.
5. MER (with allowable MER as the benchmark)
Total revenue divided by total marketing spend — but only useful when compared to your allowable MER derived from margin structure. MER tells you if the overall system is working within the constraints your business can sustain.
Why it matters: It’s your top-level health check that sidesteps attribution problems. Above allowable = room to scale. Below allowable = tighten up.
How often to check: Weekly.
The 10 that don’t (for daily operations)
These metrics aren’t useless — they’re just not decision-driving. They’re either too granular, too easily gamed, or too disconnected from profit to warrant daily attention.
1. Revenue
Revenue without margin context is meaningless. A $100K month at 8% CM3 is worse than an $80K month at 15% CM3. Revenue is an input, not an outcome.
2. Blended ROAS
Mixes high-margin and low-margin products into one number. A 4x blended ROAS can hide a campaign that’s losing money on every sale. Use POAS instead.
3. CTR (click-through rate)
A high CTR with poor conversion is worse than a moderate CTR with strong conversion. CTR is a creative diagnostic, not a business metric. It tells you if the ad is interesting, not if it’s profitable.
4. CPC (cost per click)
Cheap clicks that don’t convert cost more than expensive clicks that do. CPC is a platform mechanic, not a profitability indicator.
5. Bounce rate
Highly dependent on page type, traffic source, and user intent. A “high” bounce rate on a well-designed landing page that converts might be perfectly fine. Context matters more than the number.
6. Average session duration
Longer sessions don’t necessarily mean higher intent. Someone confused by your site structure spends longer browsing too. Focus on conversion rate, not time on site.
7. Email open rate
Apple Mail Privacy Protection has made open rates unreliable since 2021. Track click rate and revenue per recipient instead.
8. Social media followers
Followers don’t buy things. Engagement rate and attributed revenue from social are what matter — and even those are secondary to unit economics.
9. Blended CAC
As we covered in detail: blended CAC dilutes your new customer CAC with returning customers who cost nothing to reactivate. It makes acquisition look cheaper than it is.
10. Projected LTV
A 12-month projection based on historical cohorts. Useful for strategic planning, dangerous for operational decisions. Use 60-day LTGP for anything that touches spend or cash flow.
Building your metrics dashboard
Strip your daily dashboard back to the five that matter. Everything else goes into a monthly review or a diagnostic deep-dive when something looks off.
Your weekly check should take five minutes and answer three questions:
Is MER above or below allowable? (Overall health)
Is new customer CAC within target? (Acquisition efficiency)
Is repeat purchase rate holding or declining? (Retention trajectory)
If all three are green, keep scaling. If any is red, dig into the diagnostic metrics (CTR, CPC, conversion rate, email click rate) to find the root cause. But don’t let diagnostic metrics drive strategy — that’s how you end up optimising click-through rates while the business bleeds margin.
The bottom line
More metrics doesn’t mean more clarity. The five metrics above connect directly to profitability, cash flow, and growth capacity. Everything else is either a supporting indicator or noise. Build your operating rhythm around the five, and use the rest only when you need to diagnose a problem.


