Shopify Conversion: The Seven Things We Fix First

When a store owner asks for a redesign, we look at the numbers first. About half the time the design is fine and the store is losing money to four things that have nothing to do with how it looks.

Here's the order we work in, and roughly what each one has been worth on real stores.

1. Audit the apps, individually

This is the cheapest large win on almost every Shopify store we see.

One store we took on had eleven apps: reviews, chat, three separate upsell tools, loyalty, a currency switcher, and two analytics apps that both loaded the same tracking library. Together they added 3.2 seconds to first contentful paint on a mid-range Android. Two of the upsell apps fired on the same trigger and occasionally rendered on top of each other.

How to do it properly: measure each app's load cost on its own. Disable one, measure, re-enable, repeat. Tedious and unambiguous. Then ask three questions per app:

• Is anyone using the data or feature it provides? Check, don't assume.
• Could a theme change do the same thing? Reviews, currency switching, and simple upsells usually can.
• Is it worth its load cost against measured revenue?

On that store, four apps came out and their functions were rebuilt into the theme. Mobile LCP dropped from 5.4 seconds to 1.9. Nothing about the design changed.

2. Show shipping cost before the address form

The largest single drop-off we've measured in a Shopify checkout was the step where shipping cost first appeared — step three of five, after personal details.

People had already committed and then found out the total. Moving shipping onto the cart page, with a threshold indicator ("$6 more for free delivery"), lifted checkout completion 31%.

If your rates are complex, show a "from" price with the calculation available. Any number is better than a surprise.

3. Take steps out of checkout

Shopify's checkout is good and there's still room. What we change:

• Surface wallet payments first. Shop Pay, Apple Pay, and Google Pay above the card form, not below it. On mobile these convert dramatically better because they skip typing entirely.
• Never force account creation. Guest checkout, with an optional account offer on the confirmation page.
• Turn on one-page checkout if you're on a plan that supports it and your rate logic allows.
• Cut the fields that don't route or price the order. Company name and second address line are optional for most stores and rarely used.
• Instrument every step. Without per-step funnel events you're guessing about where people leave. With them, you fix the actual step.

4. Structure the product data properly

This is the least interesting item and frequently the highest-value one.

An outdoor retailer we worked with had 3,400 SKUs where buyers decide on specifications — waterproof rating, fill power, pack weight, temperature rating. None of it existed as data. It was written into description paragraphs, differently by whoever added the product, over nine years.

So the only filters available were price, brand, and size. Someone looking for a jacket rated above 20,000mm had to open products one at a time and read.

Four of the project's ten weeks went into building a per-category attribute schema in metafields, extracting what we could from descriptions, and having their product team fill the rest. Afterwards, faceted filtering on real specifications became straightforward, and search-to-purchase rate nearly tripled.

The same data powers Product schema, which makes you eligible for rich results with price and availability. Two returns from one piece of dull work.

5. Put content on collection pages

Collection pages catch the highest-intent searches — "waterproof walking jackets", "men's chelsea boots" — and on most stores they contain a product grid and one sentence.

Which is why a guide on a competitor's blog outranks your own category page for your own category.

What we add:

• A short buying-guidance section above the grid — the two or three things that decide the choice
• Subcategory links, which help both shoppers and crawlers
• Sizing, materials, or compatibility detail below the grid
• An FAQ block answering what people ask before buying in that category
• Breadcrumbs with BreadcrumbList markup

Don't dump 800 words above the products. Shoppers came for products. A tight section above and the substance below works for both audiences.

6. Get faceted navigation under control

Filters generate URLs. Left alone, three filters across a decent catalogue produce tens of thousands of crawlable near-duplicate pages, and Google spends its crawl budget on parameter permutations instead of your products.

Decide deliberately which combinations deserve indexing. Category plus one meaningful filter is often a real search term worth its own indexable page. Three stacked filters almost never is — canonical those back to the base category and keep them out of the sitemap.

On one store this alone took Googlebot from crawling product pages weekly to daily, which showed up as faster indexing of new stock.

7. Fix the product page decision layer

Once someone is on a product page they have a specific set of doubts. Answer them there rather than in a policy page they won't visit.

• Delivery date, not delivery method. "Arrives Tuesday 3 March" beats "standard delivery 2–4 days".
• Returns, stated plainly, near the button. One line, no link required.
• Stock reality. Genuine low-stock indicators help. Fake urgency doesn't — we tested a countdown timer and it moved conversion 0.4%, inside noise, while support questions about whether the offer was real went up.
• Reviews on the page, including the critical ones. A product with only five-star reviews reads as filtered.
• Sizing help for anything worn. A size guide reduces returns as well as hesitation.
• Images that answer the question the description can't. Scale, texture, and what it looks like in use.

The operational thing that isn't conversion but costs more

If your site and warehouse disagree on stock, you're either overselling or hiding inventory. One retailer with two warehouses running separate counts, reconciled overnight, was overselling roughly 40 orders a month. Each one a refund, an apology, and often a lost customer.

Unifying that into one source of truth with same-hour discrepancy reporting brought it under two a month. It doesn't show up in a conversion report and it was worth more than most things that do.

The order, and what to expect

1. App audit — days, and usually the biggest speed win
2. Shipping transparency and checkout steps — a week, and the biggest conversion win
3. Product data structure — weeks, and it unblocks everything else
4. Collection page content and facet rules — weeks, results over months
5. Product page decision layer — ongoing

Conversion changes show in two to four weeks on reasonable traffic. Organic changes take three to six months, and on a large catalogue add time for recrawling.

A tune-up engagement covering the first three items runs two to three weeks and costs a fraction of a rebuild. If the design genuinely is the problem we'll say so — but we'd rather check the numbers first, and we'll tell you when the cheaper answer is the right one.

Our Shopify development work starts with this audit.