The Online Store Metrics That Matter (There Are Six, Not Sixty)
The analytics panel has two hundred numbers and your store is governed by six. The difference between looking at data and using data is knowing which ones trigger a decision — and which just decorate.
August 18, 20236 min readIn this article
You open the analytics panel and two hundred numbers are waiting: sessions, bounce rate, pages per visit, new versus returning users, time on page. They all look important and none tells you what to do on Monday. Meanwhile, the question that actually matters — does this store make money, and will it make more next month? — is answered by six numbers almost nobody looks at together.
This article is that six-metric dashboard: what each one measures, how to calculate it without exotic tools, what range is normal and — most importantly — which concrete decision it triggers when it moves. The general metrics of any website are in website metrics that matter; here we go to the ones specific to selling.
The dashboard: six numbers, one weekly look
| Metric | Formula | Healthy reference |
|---|---|---|
| Conversion rate | Orders ÷ visits × 100 | 1-3% typical; <1% signals a problem |
| Average ticket | Revenue ÷ number of orders | Depends on catalogue; watch the trend |
| Acquisition cost (CAC) | Marketing spend ÷ new customers | Must be below first-order margin — or lifetime value |
| Repurchase rate | Customers repeating within 90 days ÷ total | 20-30% is solid for most |
| Net margin per order | Ticket − cost − shipping − fees − CAC | Positive. Sounds obvious; almost nobody calculates it |
| Cart abandonment | Unpaid carts ÷ carts created | 60-75% is normal; >85% screams broken checkout |
Conversion: the general thermometer
Conversion rate is the aggregate verdict on everything: product, price, photos, trust, checkout. That's why it's the dashboard's first number — and the worst for diagnosis, because when it drops it doesn't tell you where to look. The key is segmenting it:
- By device: if desktop conversion doubles mobile, your problem has a name and it's your store on a phone.
- By traffic source: low conversion from one specific campaign flags an ad promising what the store doesn't deliver.
- By product: the page that gets visits and doesn't sell has a problem of its own — price, photos or information.
- Never compare against external benchmarks without context: a high-ticket store converting at 0.5% can be a great business; a consumables store at 1% can be dying.
Ticket and repurchase: where the money grows
Doubling visits costs a fortune; raising the ticket 20% and repurchase five points usually costs almost nothing — and they multiply together. They're the two silent levers:
- The ticket rises with bundles, free-shipping thresholds and add-ons in the cart. The concrete tactics are in how to increase your store's sales.
- Repurchase rises with post-purchase email, consumable reminders and loyalty. It's the metric that separates a business from an acquisition machine.
- Multiply to understand: 1,000 visits × 2% × $35 × 1.3 purchases/customer = your revenue. Every factor you move moves the total — and some are much cheaper to move than others.
- Lifetime value (LTV) is the advanced version: what a customer leaves in 12 months. Once you have it, CAC gets judged against LTV rather than the first order — and suddenly you can outbid your competition for acquisition.
CAC and margin: the part that prevents self-deception
Here live the two uncomfortable metrics — the ones that turn "we sold so much" into "did we earn anything?". Calculating them means leaving the analytics panel and entering your real numbers:
- CAC includes all acquisition spend: ads, first-purchase promotions, the welcome discount. Total monthly spend ÷ new customers that month.
- Net margin per order subtracts everything: product, packaging, subsidised shipping, gateway fee, prorated returns and CAC. What remains is the business.
- The classic trap: $12 CAC against $10 gross margin per order = losing $2 per new customer... unless repurchase rescues it. That's why the six metrics get read together, never alone.
- If net margin is negative and repurchase is low, more marketing isn't the answer: it's accelerating the loss. Fix the margin or the repurchase first, then step on the accelerator.
Abandonment: the metric with money buried in it
60-75% abandonment is normal — people comparing, calculating shipping, saving for later. What matters is the excess above normal and its exact location:
- Measure by step: cart → checkout start → details → payment. The abnormal leak between two specific steps is your free diagnosis.
- A leak at the cart: surprise shipping costs or product doubts. Fixed by anticipating costs.
- A leak at the details: a long form or forced registration.
- A leak at payment: missing methods, a failing gateway, or mobile not cooperating.
- And recover the leavers: the abandoned-cart email sequence rescues 5% to 15% — money that was already nearly collected.
The vanity metrics (and their proper use)
It's not that followers, likes and sessions are useless: they just don't govern. They're top-of-funnel indicators — useful for deciding which content to repeat or which channel to explore, useless for knowing whether the business works.
- Sessions without conversion = reach without business. Celebrate traffic only when the six-metric dashboard can absorb the arrival.
- Followers don't pay invoices: the question is how many become your own contacts and buyers.
- Bounce rate lies in stores: whoever opens a product page, sees the price and calls on WhatsApp counts as a bounce — and is a sale.
- The general rule: if a number can't trigger any concrete decision, it's decoration. Review it quarterly or never.
Frequently asked questions
What tools do I need to measure all this?
Fewer than you think: your store platform's panel (orders, ticket, abandonment), your web analytics (visits and sources) and a spreadsheet where you gather the six numbers weekly. The spreadsheet is the key tool, because it's the only one that crosses sales data with spend data — something no panel does on its own.
What's a good conversion rate?
Between 1% and 3% for most stores, but the honest range depends on your ticket, category and traffic source. The useful comparison isn't against the industry but against yourself: this month's conversion against the previous three, segmented by device and source. A sustained drop there is an actionable alarm.
How often should I review the metrics?
The dashboard's six, once a week in a short session; net margin and CAC calmly once a month, when the expenses are complete. The error at both extremes: watching daily (noise that craves hasty decisions) or only when something's wrong (you're already late). The weekly series is the balance.
My traffic is rising but sales aren't. Where do I look?
It's the classic symptom of wrong traffic or a store that doesn't convert. Segment conversion by source: if the new traffic comes from one specific campaign or network and converts near zero, the ad's message doesn't match what the store delivers. If every source converts worse than before, the problem is inside: speed, product pages, prices or checkout.