Where do your visitors stop?
Before you change anything, find out where people drop out. Shopify has a report for exactly this.
In your Shopify admin, go to Analytics > Reports, click the Category filter and choose Behavior. Open Conversion rate breakdown. It shows four steps: all sessions, sessions with cart additions, sessions that reached checkout, and sessions that completed checkout. Look at the last 30 days, and if you can, the same weeks last year.
The step where most people leave tells you which checks to start with:
| Where they stop | What it usually means | Start with checks |
|---|---|---|
| Few add to basket | Wrong visitors, slow pages, product pages that don't answer the question, prices | 1 to 6 |
| Many add to basket, few reach checkout | A surprise in the basket: delivery cost, delivery date, returns | 7 to 9 |
| Many reach checkout, few finish | Something in checkout: an account wall, a long form, a missing payment method, an error | 10 and 11 |
| Nobody does anything at all | Check that the store works: place a test order | 11, then 1 |
No shop gets everyone through. Baymard's average across 50 studies puts basket abandonment at 70%. You're looking for the step that is much worse than the others, or worse than last year.
1. Are the right people visiting?
How to check: In Analytics > Reports, filter by Acquisition and open Sessions by referrer. This shows where visits come from, but not who bought. For that, Shopify's Help Center points to the Total sales by referrer part of your sales reports. Put the two side by side.
What bad looks like: One source sends a lot of visits and almost no sales. Often that's an ad or a social post that promised something your shop doesn't quite sell (a different price, a product that's sold out, a style you don't stock).
The fix: Make the ad and the page it lands on say the same thing. If an ad shows a red jacket at €49, it should land on the red jacket at €49, not on your home page. If a source sends people who never buy, spend less there before November and more where buyers come from.
2. Are they ready to buy, or just looking?
How to check: Look at the pages people read and how often they come back. In November, many visitors are browsing for ideas, comparing your price with two other shops, or waiting for the sale to start.
What bad looks like: It isn't bad. A visitor who is waiting for Friday isn't a fault in your shop. It only becomes a problem if you treat every visitor the same, for example with a discount code for people who would have bought anyway.
The fix: Give each group a different next step. We wrote a whole article on this: how to tell browsers, comparers and people waiting for the sale apart, with a table of the signals to look for. The rest of this list is about the visitors who wanted to buy and something stopped them.
3. Does the shop work on a phone?
How to check: In Analytics > Reports, filter by Behavior and open Sessions by device. Then, if your conversion rate report lets you split by device, compare phones and computers. Then do the simple test: on your own phone, on mobile data, go from a product to a finished test order.
What bad looks like: Phones bring most of your visits but far fewer of your orders. Or, on your own test: a menu that covers the "Add to cart" button, a size picker that's hard to tap, a cookie banner and a popup on top of each other, text you have to zoom to read.
The fix: Fix whatever got in your way on the test, starting with anything that covers the buy button. Remove popups that appear before someone has read anything: Google warns against them on mobile, and shoppers find them annoying. See what annoys shoppers less than a Black Friday popup.
4. Are your pages fast enough?
How to check: In your Shopify admin, go to Online Store (the Themes page). It shows a summary of your Core Web Vitals over the last 30 days. For more detail, search for "contentful" in Analytics > Reports to find the Largest Contentful Paint reports, including by page type. Data can be up to 36 hours behind.
What bad looks like: Google's thresholds, which Shopify uses too, measured at the 75th percentile of visits:
| Good | Poor | |
|---|---|---|
| Largest Contentful Paint (how fast the main content shows) | 2.5 seconds or less | over 4 seconds |
| Interaction to Next Paint (how fast a tap responds) | 200 ms or less | over 500 ms |
| Cumulative Layout Shift (how much the page jumps) | 0.1 or less | 0.25 or more |
Product pages that are "Poor" on a phone are the ones to fix first.
The fix: The usual causes on a small Shopify store are big images and apps that add scripts. Compress the main product image and use a modern format (Google's guide suggests WebP or AVIF). Don't lazy-load the main image at the top of the page: Google's guide says "Never lazy-load your LCP image". Remove apps you no longer use, because some leave scripts behind. Make theme changes in the next two weeks, so you can test them well before Black Week.
5. Does the product page answer the question?
How to check: Read your five most visited product pages as if you were a stranger. Then read last November's emails and chats: every question people asked is one that others didn't ask, and left.
What bad looks like: No size or measurements, one photo, no answer to "will it fit", "is it washable", "does it work with X". The same question in your inbox every week.
The fix: Put the answer on the page, near the price and the buy button: size and fit, what's in the box, materials, a photo of the product in use. If one question comes up again and again, it deserves a line on the page, not only an FAQ at the bottom.
6. Can people find what they're looking for?
How to check: In Analytics > Reports, filter by Behavior and open Top online store searches with no results and Searches with no clicks.
What bad looks like: People search for things you sell but call by a different name ("trainers" when your product says "sneakers"), or for things you don't sell at all.
The fix: Add the words people use to product titles, descriptions or tags. If many people search for something you don't stock, that's useful too: it's a product idea, or a reason to say clearly what you do sell.
7. Do costs show up late?
This is the most common reason people leave a basket that you can actually fix. In Baymard's survey of US online shoppers, 40% of those who left a basket for a reason other than browsing said extra costs (shipping, tax, fees) were too high, and 12% said they couldn't see or work out the total cost up front.
How to check: Add a product to the basket on your own phone. Is the delivery cost visible before checkout? Is it on the product page? In Analytics, check whether many people add to basket but few reach checkout.
What bad looks like: The first time a visitor sees the delivery cost is on the last step of checkout. Or the cost is high compared to the price of a small product.
The fix: Show the delivery cost on the product page and in the basket, in plain words: "Delivery €4.95, free over €50". To set a free delivery amount, go to Settings > Shipping and delivery, open your shipping profile, edit the rate and choose Offer free shipping with a minimum order amount. Or set the rate type to Order amount with a top range at 0. Leave the Maximum empty on the highest range: Shopify warns that otherwise larger orders get no rate and an error at checkout. For more ideas on free delivery as a Black Friday offer, see Black Friday without big discounts.
8. Is the delivery date clear?
How to check: On a product page, can you see when it will arrive, without clicking anywhere?
What bad looks like: "Ships in 3–5 working days" in the footer, or nothing at all. In November and December that's a real worry: in the EU, 20% of people who bought online in 2025 had a delivery that was slower than expected (Eurostat), and Baymard's US survey has 20% leaving a basket because delivery was too slow.
The fix: Put an expected delivery day on the product page and in the basket ("Order today, with you by Thursday"), and from late November the last day to order for Christmas. Carriers publish their Christmas cut-offs in December; the 5-week Black Friday plan has last year's dates as a guide, and how much margin to leave.
9. Can people find your returns policy?
How to check: In Settings > Policies, check that you have a written return and refund policy and a shipping policy. Then look for them as a visitor: on the product page, in the basket, in the footer.
What bad looks like: No written policy, or one people only find in the checkout footer, after they've decided. In Baymard's US survey, 13% left a basket because the returns policy wasn't good enough. In the EU, 8% of online buyers in 2025 had trouble finding information on guarantees and their legal rights (Eurostat).
The fix: Write the policy in plain words and add one short line near the buy button: "Free returns within 30 days", or whatever your real terms are. In the EU, people can cancel an online purchase within 14 days of delivery without giving a reason (Your Europe), so you're likely offering this anyway; say so. For gifts, consider a longer return window until January. Shopify links your policies in the checkout footer once you add them, and you can add them to your menus in Content > Menus.
10. Would you trust this shop?
How to check: Ask someone who doesn't know your shop to look at it for two minutes, then ask them who runs it, where it's based and how to reach you.
What bad looks like: They can't tell. No address, no contact details, no "about" page, no reviews, a theme that still has demo text in it. In Baymard's US survey, 19% left a basket because they didn't trust the site with their card details, and 17% because the website had errors or crashed.
The fix: Add who you are, a real contact email and phone number, your business address, and a few real reviews. Fix any broken links and error pages. Show the payment logos people know at checkout. None of this takes more than an afternoon.
11. Does checkout get in the way?
How to check: Place a test order. Shopify lets you test with a test payment gateway or Shopify Payments in test mode, and test orders aren't charged. Then go to Settings > Checkout and look at the Customer contact method and Customer information sections.
What bad looks like:
- An account wall. Require customers to sign in to their account before checkout is switched on. In Baymard's survey, 18% left a basket because the site wanted them to make an account. Shopify also notes that with this setting on, buttons like Apple Pay aren't shown in the cart.
- A long form. Company name or phone number set to Required when you don't need them. Baymard counts an average of 14.88 form fields in US checkouts, and says 7 to 8 can be enough. 17% left because checkout was too long or complicated.
- Few ways to pay. 9% left because there weren't enough payment methods.
The fix: Turn off the sign-in requirement unless you really need it. Set company name to Don't include or Optional, and the phone number to Optional unless your carrier needs it. Turn on the express payment buttons your customers use (Shop Pay, Apple Pay, Google Pay, PayPal): Shopify shows them at the start of checkout, and some of them also on the product page.
12. Are your prices close to the shops you compete with?
How to check: Search Google Shopping for your five best-selling products, by name or model number. Note who sells the same or a similar product, and their price with delivery.
What bad looks like: The same product is a lot cheaper elsewhere, delivered, and nothing on your page explains why yours is worth more.
The fix: You don't always have to be the cheapest. Show what the other shop doesn't: faster delivery, an easy return, advice, a gift wrap, a guarantee. If you do lower prices for Black Friday, in the EU the "was" price must be your lowest of the last 30 days: see the EU 30-day price rule for Black Friday.
Which checks first, with four weeks to go
You won't do all 12 this week, and you don't need to. A rough order:
- This week: the funnel report, a test order on your own phone (checks 3, 7 and 11), and your returns and delivery lines (8 and 9). These are small changes with the biggest effect.
- Next week: speed and product pages (4 and 5), while there's still time to test theme changes.
- Mid November: traffic sources and prices (1 and 12), so your Black Friday ads point at pages that work.
- The week before: test everything again on a phone, with the sale switched on.
Big changes to your theme or checkout in Black Week itself are risky. If something breaks on the busiest days, you lose more than you gain.
A worked example
This is an illustration, not a customer story. A one-person shop sells wool blankets. The conversion rate breakdown shows plenty of cart additions, but most of those visitors never reach checkout. That points to the basket.
On her own phone, the owner adds a blanket to the basket. The delivery cost (€9.50 for a heavy parcel) only appears on the last step of checkout, and the product page says nothing about when it arrives. Her inbox from last November has the same question five times: "Will it arrive before Christmas?"
She makes three changes in an afternoon: the delivery cost and "free over €120" on the product page, an expected delivery day next to the buy button, and a line about returns. Two weeks later she checks the same report again, to see if more people now reach checkout.
Where Signal fits
Most of these checks are fixes on the page, and you don't need a tool for them. Where Signal helps is with the visitors who still hesitate. A small script on your shop reads why someone is visiting from the pages they read, where they came from and whether they've been before. When one question would change what to show ("Is it a gift?", "When do you need it by?"), it asks it, and shows the answer or the offer you wrote for that reason. If you connect Shopify, it can show three matching products from your own catalogue.
Shoppers can also ask their question right on the product page, and you answer from your inbox, or by email if they've already left. Over a few weeks those questions show you what your product pages are missing (check 5). The script loads after your page, so it doesn't slow your product pages down, and it never stores IP addresses. See how it works, or start in shadow mode: it shows what it would have done on your real traffic, without visitors seeing anything.
When this isn't the right approach
- Nobody visits at all. If you have a few visits a day, these checks won't change much. The work is in getting people to the shop: search, social, your newsletter, the places your customers already are.
- The product is the problem. If people come, read, and leave across every source and every device, ask a few of them why. Sometimes the answer is the product or the price, not the page.
- You're reading this in Black Week. Don't rebuild anything. Do a test order, put the delivery cost and date on the product page, and answer last year's questions. The 5-week plan has a short list for late starters.
- Your numbers are tiny. With ten orders a month, one good or bad week moves the conversion rate a lot. Look at what people do on the pages and what they ask you, not only at percentages.
