Check five things, in this order, and stop at the first one that fails: (1) is tracking and checkout working, (2) is the traffic the right traffic, (3) does the product page earn an add to cart, (4) does the cart survive to checkout, (5) does checkout complete. The four funnel numbers in your Shopify analytics (sessions, added to cart, reached checkout, completed checkout) tell you which step to start at. Fixing a later step while an earlier one is broken wastes the work.
Shopify’s own conversion breakdown report shows exactly these steps and calculates each as a share of total sessions (Shopify Help Center). Open it for the last 30 days before reading further.
Which number is broken
Divide each step by the one before it. The step with the biggest gap against your own earlier weeks, or against another traffic source in your own store, is where to start. We do not give benchmark percentages: the ones circulating in “traffic but no sales” posts rarely name a source, and a cold TikTok audience and a branded search audience should never share one benchmark.
| What you see | Start at | Section |
|---|---|---|
| Zero orders and you have never seen a test order complete | Step 1 | Tracking and checkout work |
| Sessions, but visitors leave without viewing a product | Step 2 | Traffic quality |
| Product views, few or no adds to cart | Step 3 | Product page |
| Adds to cart, few reach checkout | Step 4 | Cart |
| Reached checkout, few complete | Step 5 | Checkout |
Step 0: is the sample big enough to say anything?
Zero orders is not always a verdict. If the true conversion rate were 1 percent, the chance of seeing no order in 300 sessions is 0.99 to the power 300, about 5 percent. At 100 sessions it is about 37 percent. So 100 sessions with no sale is ordinary bad luck at 1 percent; 500 sessions with no sale (about a 0.7 percent chance) is a signal. This arithmetic is ours, from the binomial formula: the probability of zero orders is (1 minus rate) raised to the number of sessions.
Rule of use: below roughly 300 sessions from one traffic source, do the technical checks (steps 1, 4, 5) and stop changing copy. Above 500 sessions with zero orders, something structural is wrong.
Step 1: rule out broken tracking and checkout
Before diagnosing behaviour, prove a customer can pay.
- Place a real order on your own store on a phone, on mobile data, using the same landing link an ad uses. Refund it afterwards.
- Check the payment methods a customer in your target country actually sees. See card declined and payment failed for causes on the payment side.
- Check the analytics. Shopify counts a session as ended after 30 minutes of inactivity and has changed how “reached checkout” is counted (Shopify Help Center), so old screenshots of your funnel may not compare with today’s. If you buy ads, confirm the ad platform sees purchases: Meta tracking for Shopify.
- Look at your installed apps. Every script an app adds can slow the mobile page; the app cost calculator lists what a typical stack costs so you can see what you are paying for.
Step 2: traffic quality
If most sessions never view a product, the store may be fine and the audience wrong.
- Split the funnel by traffic source in the acquisition report. One good source and three bad ones means you have a traffic problem, not a store problem.
- Compare what earned the click with the first screen of the landing page. The product, audience and promise should be recognisable in two seconds. If the ad shows a desk gadget and the page opens on a brand slogan, that mismatch is the first fix.
- Separate curiosity traffic (short-form entertainment) from intent traffic (search, retargeting). Judge each on its own progression to product view and add to cart, not on clicks.
- Very small stores with almost no traffic sources beyond friends should read how to get your first 100 visitors first.
Step 3: product page (views but no adds to cart)
A buyer needs answers to five questions without scrolling far: what is it for, what exactly arrives, will it fit or work for me, when will it arrive, what if I dislike it.
| Buyer question | What the page must show | Weak version |
|---|---|---|
| What does it do for me | One concrete use and result | A lifestyle slogan |
| What arrives | Contents, size, material, quantity | Supplier image, no specifications |
| Will it fit | Sizing, compatibility, scale in a photo | One studio shot |
| When | Ships from where, processing plus transit range | “Fast shipping” |
| If I dislike it | Return window and who pays | A footer link |
Use the full product page blueprint and trust signals guide to work through it. Only add proof you actually have: no invented reviews or fake scarcity. If your theme cannot express the layout you want, a page builder such as GemPages is one route, but it cannot supply proof or fix a weak offer.
Step 4: cart (adds to cart but few checkouts)
The usual cause is a cost that appears late. Baymard’s aggregate of 50 studies puts the average documented cart abandonment at 70.22 percent, and lists extra costs such as shipping, tax and fees as a reason for 40 percent of abandoners, ahead of slow delivery (20 percent) and forced account creation (18 percent) (Baymard). Those are survey figures across many sites, not a benchmark for your store, but the ranking of reasons is a useful checklist.
- Open the cart in a private window, in the target country, and note every amount and the delivery range shown.
- Does the shipping cost first appear at checkout? Show it, or fold it into the price. See shipping settings for dropshipping.
- Is a discount code required to see a sensible price? Test at the regular price too.
Step 5: checkout (reached checkout, few complete)
Buyers here have shown intent, so the cause is usually cost surprise, a missing payment method, or an error. Baymard lists security worries (19 percent), a complicated flow (17 percent) and site errors (17 percent) alongside the cost reasons. Check the target customer’s wallet and card options, taxes and duties shown before payment, and errors in the checkout. Go deeper with checkout optimisation and cart abandonment fixes.
What we could not source
We give no conversion-rate benchmarks because we found no neutral, Shopify-specific source that separates traffic type. Numbers such as “add to cart under 8 percent is a problem” appear in vendor posts without a source; use your own weeks and your own sources as the baseline instead. To read the funnel numbers in more detail see funnel metrics: what each drop-off tells you.
Common questions
How much traffic do I need before I can diagnose?
Enough that zero orders is unlikely by chance. At a 1 percent true rate, 500 sessions with no order has under a 1 percent chance, so it points to a real fault. With 100 sessions, luck alone explains it about a third of the time.
Should I buy more traffic?
Not until steps 1 to 5 pass for the traffic you already have. More visitors into a broken step just multiply the waste.
Should I install a conversion app first?
No. Prove tracking and checkout, then fix the step that fails. Apps add cost and scripts; check the app cost calculator before adding one.
Method: researched from Shopify Help Center documentation, Baymard’s published statistics and the top-ranking guides; checked 29 Sept 2026. Probability figures are our arithmetic.