If you run an online shop without a CTO, storefront problems rarely announce themselves as “conversion issues”. They show up as abandoned carts, “is this in stock?” emails, payment drop-offs, and warehouse firefighting after a sale has already been promised.
You do not need a redesign mood board to find most of that. You need a walkthrough of the live shop — as a new customer would experience it — and an honest look at the ops behind it.
Use this checklist yourself. Tick what you can prove, not what you hope is true.
How to use this checklist
- Open the site on your phone and on a laptop
- Pick one popular product and one awkward one (bulky, made-to-order, low stock, or multi-variant)
- Buy — or go as far as you safely can — using a real shipping address style your customers use
- Note every moment of doubt, delay, or “I’ll check with the team”
- Rank issues by lost sales or support load, not by how embarrassing they feel
A short written list beats a vague feeling that “checkout is a bit clunky”.
Catalogue and discovery
- Can a new visitor find a popular product in under three clicks from the homepage?
- Do search, filters and category pages show products that are actually available?
- Are out-of-stock or low-stock items honest — or do they invite an order you will later cancel?
- Do product pages answer the questions that already create support tickets (size, fit, lead time, what’s included, delivery area)?
- Are variants (size, colour, pack) clear before add-to-cart, with price changes visible?
- Do related products and bestsellers help someone continue — or dump them into a dead end?
If discovery is weak, paid traffic and pretty banners will not save you. People cannot buy what they cannot find or trust.
Product truth and stock messaging
- Does the site’s stock message match what the warehouse would pick today?
- What happens when stock hits zero mid-day — does the page update, or can customers still order?
- For pre-order, back-order or made-to-order items, is the promise dated and obvious before payment?
- Are bundle or multi-location stock rules understandable to a human, not only to the system?
Stock lies are expensive. They create refunds, bad reviews and ops chaos that looks like a “website problem” but started as a data problem.
Cart, checkout and payments
- How many steps from first “Add to cart” to a paid order?
- Can someone see shipping options and likely cost before they enter card details?
- Are guest checkout and account checkout both workable, or does one feel punished?
- What happens on card decline, 3-D Secure failure, timeout, or address validation error — clear recovery, or a dead end?
- Do promo codes, gift cards and VAT display behave as a customer would expect?
- Is there a final review step that shows item, delivery promise and total without surprises?
Every unclear step is a quiet “no”. You will not hear most of them; you will only see the conversion number sag.
Delivery promises and returns
- Is the delivery promise specific enough to plan around — or vague enough to create inbox mail?
- Are restricted postcodes, remote areas, or collection options explained before payment?
- Is the returns path visible, fair, and aligned with what ops will actually do?
- Can a customer find tracking or “where is my order?” without emailing you?
Trust is not a badge pack. It is whether the promise on the page matches what happens after the order lands.
Mobile reality
- Does the mobile purchase flow match how your customers actually buy (thumbs, one hand, poor signal)?
- Are tap targets, sticky bars and pop-ups fighting the add-to-cart button?
- Do images, variants and delivery info load in a usable order on a mid-range phone?
- Have you tried the flow on a real device — not only a resized browser window?
For many UK SMEs, mobile is not a secondary channel. It is the channel. If mobile is awkward, the desktop polish is mostly vanity.
Ops behind the storefront
Storefront symptoms often start elsewhere:
- Stock, pricing or catalogue updates re-keyed between systems
- Orders landing in email or spreadsheets before the warehouse sees them
- Courier rules that the site cannot express, so CS invents workarounds
- Peak-day processes that only work when one person is in the building
If several of those are true, fixing the theme will not fix the business. See also Business Systems and the guide on choosing systems without vendor lock-in.
How to prioritise what you found
Score each issue roughly:
- Does it stop payment or create a wrong promise? — fix first
- Does it create repeated support or warehouse rework? — fix next
- Does it annoy without killing the order? — schedule
- Is it taste, not friction? — park it
Resist the urge to start with a homepage redesign because it is more fun than fixing decline recovery or stock honesty.
You can do a useful pass without buying a review
Print this checklist. Walk the shop. Write the top five issues with evidence (screenshot, URL, what you clicked). Give that list to your agency, developer or internal owner with a clear order.
That alone often beats a vague “we should improve conversion” conversation.
If you want an independent walkthrough with annotated findings and a prioritised fix list for your team or agency, an E-commerce Health Check is a structured way to do it. It is a means. The point is fewer lost sales and fewer operational surprises — not a nicer slide deck.
Next step
Book a Discovery Call if the checklist already shows friction you cannot ignore — or if the storefront looks fine and the pain is really in stock, fulfilment or the systems behind it.