How to Calculate and Improve Your Online Store’s Conversion Rate
You can pull 1,000 visitors a day and still sell nothing. I’ve seen stores with great traffic and a checkout page that quietly murders every sale before it happens. Traffic gets people to the door; conversion rate tells you how many actually buy, and it’s the single number separating stores that grow from stores that just get looked at.
Here’s the formula, a worked example, why stores bleed conversions, and seven fixes you can make this week.
The formula (it’s one line)
Conversion rate = (orders / visitors) x 100
500 visitors, 10 orders, that’s 2 percent. Typical e-commerce runs 1 to 3 percent, with well-optimized stores pushing higher. Under 1 percent almost always means something fixable is broken somewhere in the buying path.
Worked example with made-up (but realistic) numbers
Your store’s week:
| Metric | Value |
|---|---|
| Visitors | 2,000 |
| Product page views | 1,100 |
| Added to cart | 180 |
| Reached checkout | 90 |
| Orders completed | 40 |
Overall conversion: 40 / 2,000 = 2 percent.
Now the interesting part, the funnel breakdown:
- Visitor to product page: 55%. Nearly half left before seeing a single product.
- Product page to cart: 16%. They looked, didn’t want it badly enough.
- Cart to checkout: 50%. Half the interested buyers walked away here.
- Checkout to order: 44%. Over half abandoned at the final step.
The 2 percent headline hides everything. Your biggest leaks are before the product page and inside checkout, fixing either moves the number more than any amount of new traffic would.
Why conversion rates run low
- Slow pages. Every extra second of load bleeds visitors, especially on mobile data. Indonesian buyers on 3G don’t wait.
- Weak product pages. Blurry photos, vague descriptions, missing sizes, prices that make no sense.
- Surprise costs at checkout. Shipping fees appearing at the last step kill more orders than almost anything else on this list.
- Checkout friction. Forced account creation, endless forms, payment methods nobody uses.
- Zero trust signals. No reviews, no return policy, no real contact info. People don’t hand money to stores that feel anonymous, would you?
Seven fixes that actually move the number
- Compress images, cut page weight. Test your store on a phone with slow internet, not your office wifi. Night and day difference.
- Rewrite your top product page. Real photos from multiple angles, specs in a table, benefits before features.
- Show total price early. If shipping is Rp 15,000, say it on the product page, not at checkout where it becomes a nasty surprise.
- Guest checkout. Forcing account creation costs real orders. Offer the perks after they’ve bought.
- Payment methods people actually use. In Indonesia: QRIS, e-wallets, virtual accounts. Bank transfer alone isn’t enough anymore.
- Reviews and a clear return policy. Even five honest reviews change how a store feels to a stranger.
- Obvious contact. A WhatsApp button converts hesitant buyers better than any trust badge you can buy.
Measuring in Google Analytics
GA4 path: open Reports, check Sessions for visitors, track purchases as a key event. The funnel exploration lets you rebuild that table above, product view, add to cart, begin checkout, purchase. Find the step with the biggest percentage drop; that’s your priority list. Check it weekly, set a reminder, because funnels drift as traffic sources and seasons change.
Platform analytics too heavy? Your e-commerce platform’s built-in stats show orders and sessions too. Formula works either way.
What improvement actually looks like
Same fictional funnel, three fixes applied, faster mobile pages, rewritten product page, visible shipping costs. A plausible month later:
| Funnel step | Before | After |
|---|---|---|
| Visitor to product page | 55% | 70% |
| Product page to cart | 16% | 20% |
| Cart to checkout | 50% | 55% |
| Checkout to order | 44% | 55% |
New math: 2,000 visitors now produce about 85 orders instead of 40. Conversion jumps from 2 percent to roughly 4.2 percent, with zero new traffic. This is why conversion work beats buying ads for small stores: same visitors, twice the payoff.
Don’t just measure the overall number
Once basics work, check two cuts: conversion per product (a bestseller at 5 percent often hides a dead page dragging the average down) and conversion per channel (WhatsApp broadcast visitors convert nothing like cold Google traffic). Work where the number is weakest, not where it’s already strong.
A rule for where to spend effort
Small stores burn months A/B testing button colors while ignoring a 5-second load time on mobile data. Order fixes by funnel position, patch the stage losing the most people first, because a percentage point recovered early compounds through every later step.
And one habit worth building alongside: recalculate the rate every week on the same day, write it where you’ll see it again. One number tells you little; a trend line tells you whether each fix actually worked or just felt like it did.
This week’s task
Save your funnel table as the baseline. Next month’s version is how you prove the fixes worked.
Pull last week’s numbers from whatever tool you have, visitors, orders, done. Then walk your own store on a phone, homepage to completed order, noting every moment that annoyed you. That annoyance is what your customers hit daily, and every one you remove shows up directly in the percentage.
More traffic multiplies a broken funnel into a bigger loss. Fix the rate first, then buy the visitors.
