I sell a Shopify audit for 590€. And I am going to tell you that 70% of that audit is doable yourself, for free, in one afternoon.
That sounds like a strange way to sell anything. But it is the truth, and you deserve to hear it before you spend money. Most of what a professional audit surfaces comes from tools that Google and the open web give away. PageSpeed Insights, Search Console, Chrome DevTools. They are free, they are good, and they will tell you a lot about why your store is slow or why pages do not convert.
So why does the paid audit exist at all? Because finding problems is the easy part. The hard part is knowing which ones matter, in what order, and how much each fix is worth in hours and euros. That is the 30% that tools do not give you. It comes from having looked at dozens of stores and learned to read the patterns.
This article does two things. First, it hands you the full DIY method so you can run the 70% yourself this afternoon. Second, it draws an honest line around the 30% that justifies paying an expert, so you can decide based on your stage and not on a sales page.
The 70/30 split is my own estimate from running these audits, not an external statistic. Treat it as a rough framing, not a measured number. By the end you will know whether you need me at all. If you do not, good. You will have saved 590€ and learned your own store.
What a professional Shopify audit measures
A real audit is not a vibe check. It rests on four pillars, and the 590€ offer is built around exactly these four. Each one answers a different question, and a finding only counts when you can place it inside one of them and say why it matters.
Performance and Core Web Vitals
The first pillar is speed, measured the way Google measures it. The three Core Web Vitals are Largest Contentful Paint (loading), Interaction to Next Paint (responsiveness), and Cumulative Layout Shift (visual stability). These are not abstract scores. They reflect what a real visitor feels when your product page takes three seconds to become usable.
LCP measures when the biggest element above the fold finishes rendering, which on a Shopify store is almost always the hero image or the main product photo. INP measures the delay between a tap or click and the screen actually responding, so it captures the lag you feel when the add-to-cart button hesitates. CLS measures how much the layout jumps while loading, the kind of shift that makes you tap the wrong thing.
An audit looks at both lab data (a controlled test) and field data (what real Chrome users experienced), because the two often disagree. A store can pass in the lab and fail in the field. The audit also breaks the numbers down by page template, because a fast homepage tells you nothing about the product page where the money is actually made.
Technical SEO
The second pillar is whether Google can crawl, understand, and rank your pages. This covers indexability, duplicate content from collection and tag URLs, metadata, internal linking, structured data for products, and the sitemap. Shopify generates a lot of this automatically, which is both a help and a trap, because the defaults are not always right for your catalog.
The crawl side checks that Google can reach every page worth ranking and is not wasting its budget on filtered or tagged URLs that duplicate each other. The understanding side checks that titles, descriptions, and product structured data are present and unique, so a product can show its price and availability directly in search results. The ranking side checks internal linking, because a product nothing links to is a product Google quietly ignores.
This pillar is where Shopify stores leak the most silent value. The platform spins up /collections/all, tag pages, and vendor pages that all compete for the same keywords, and most owners never notice until an audit lays them side by side.
UX friction and conversion
The third pillar is where visitors stall, hesitate, or leave. Checkout steps, mobile navigation, product page clarity, search behavior, form friction. Baymard Institute has spent years documenting these patterns through usability research at baymard.com/research, and most stores repeat the same handful of mistakes.
The audit walks the real buying path on a phone, because that is where most of your traffic is and where most of the friction hides. It looks at whether the price is visible without scrolling, whether the variant selector is obvious, whether the cart drawer makes the next step clear, and whether the checkout asks for anything it does not need. Each point of friction is a place where a visitor who wanted to buy quietly decided not to.
Deliverable: a prioritized list with effort estimates
The fourth pillar is not a measurement. It is the output. A pile of problems is useless. The deliverable is a prioritized list of fixes, each tagged with expected impact and the effort it takes, so you know what to do Monday morning and what can wait.
This is the pillar that turns the first three into action. Anyone can hand you forty findings. The value is in saying these five are worth your week, these ten can wait a quarter, and these twenty-five are noise you can ignore. Without that ordering, an audit is just a longer to-do list you will never finish.
The 70% you can do yourself with free tools
Here is the work you can run today without spending a euro. Four tools, in order. Give yourself an afternoon, open a blank document, and write down every finding as you go, because the list itself is the point.
PageSpeed Insights
PageSpeed Insights is the front door. Paste a URL, get a report. Run it on your homepage, your best-selling product page, and one collection page. Do not test only the homepage, because product pages behave differently and that is where money is made.
The report gives you a lab score and, when enough real traffic exists, field data from the Chrome User Experience Report. The field data is the one that matters for ranking. You can read more about how this real-user dataset works in the Google CrUX documentation.
Look at the three Core Web Vitals first. If LCP is above 2.5 seconds or INP is above 200 milliseconds, you have a real problem. The Opportunities and Diagnostics sections below the scores tell you what is dragging it down: oversized images, render-blocking scripts, unused JavaScript. Write each one down with the URL you tested, so you can tell later whether a finding is store-wide or stuck to one template.
Run each page two or three times rather than once. A single run is a snapshot and can swing by ten or fifteen points for reasons that have nothing to do with your store. If you see the same opportunity appear across every run and every page, that is a real pattern worth chasing.
For the INP details, which is the trickiest of the three to fix on Shopify, read my dedicated article. Google's own explainer on the metric lives at web.dev/articles/inp and is worth ten minutes.
Google Search Console
Search Console is free and tells you how Google actually sees your store. If you have not set it up, do that first. It takes ten minutes and you verify ownership through your domain provider or a Shopify setting.
Three reports matter. The Core Web Vitals report shows you field data grouped by page type, so you can see whether the problem is your product template or your blog. The Page Indexing report shows you which URLs Google refused to index and why, which often surfaces duplicate-content issues that Shopify creates on its own. The Performance report shows you which queries bring traffic and where you rank.
Spend twenty minutes here. In the Page Indexing report, open the "Why pages aren't indexed" list and look for entries like "Duplicate without user-selected canonical" or "Crawled - currently not indexed." On a Shopify store these almost always point at tag and filtered collection URLs eating your crawl budget. In the Performance report, sort by impressions and look for queries where you rank on page two, because those are the pages one small fix can push onto page one.
Chrome DevTools
Open your store, press F12, and you have a full diagnostic lab in the browser. Two tabs do the heavy lifting.
The Network tab, with the cache disabled and throttling set to a slow 4G profile, shows you every file your page loads and how heavy it is. Sort by size. A 2 MB hero image or a 400 KB third-party script will jump out immediately. Check the total page weight at the bottom of the panel, and watch the waterfall for files that load one after another instead of in parallel, because that serial loading is often what stretches your LCP.
The Performance tab records what happens during load and interaction. It is denser, but even a quick recording shows you long tasks blocking the main thread, which is the usual cause of poor INP. Record a page load, then click around the way a buyer would, and look for the red-flagged long tasks in the timeline. Shopify apps are frequent culprits here, and each one usually owns a recognizable chunk of script you can trace back by name. For apps specifically, here is how to measure their cost on your store.
Screaming Frog or Sitebulb
The last tool is a crawler. Screaming Frog is free up to 500 URLs, which covers most small and mid-size catalogs. Sitebulb is an alternative with a friendlier interface and clearer explanations of each issue.
A crawler walks your whole site the way Google does and builds a map. It flags broken links, redirect chains, missing meta descriptions, duplicate titles, thin pages, and orphan pages that nothing links to. On a Shopify store the duplicate-title and duplicate-content findings are usually the most valuable, because the platform generates filtered and tagged collection URLs that compete with each other.
Point the crawler at your homepage, let it run, then sort the results by issue type. Look first at the duplicate titles and duplicate descriptions, then at the response codes for any 404s and long redirect chains, then at the pages with zero internal links pointing in. Export each list to a spreadsheet. Run the crawl, export the issues, and you have a technical SEO punch list. This is the single most tedious part of the 70%, and it is also the part a tool does best.
The 30% that justifies an expert
You now have a stack of findings. Here is where the free tools stop helping and experience starts paying for itself. Four limitations the tools cannot cross.
Impact-effort prioritization
A tool gives you a list. It does not rank it. PageSpeed will tell you that you have eighteen opportunities, and it sorts them by estimated savings in milliseconds, which is not the same as business impact. A 300 ms saving on a page nobody visits is worthless. A 100 ms saving on your top product page might be worth thousands.
Knowing which fix to do first means weighing impact against effort, and that weighting depends on your traffic, your margins, and your team. No tool sees those. Take a real example: PageSpeed flags both "serve images in next-gen formats" and "reduce unused JavaScript" on the same page. The image fix is an afternoon and lifts every page. The JavaScript fix might mean untangling an app you depend on for half your revenue. Same report, opposite verdicts, and only context decides.
Measurement bias and interpretation
Tools lie if you read them wrong. A single PageSpeed run is a snapshot and can swing fifteen points between two runs on the same page. Lab data and field data routinely disagree, and people draw the wrong conclusion from the wrong number.
A CLS score can look fine in the lab and be terrible for real users on slow connections, because the lab does not load the third-party ad or the cookie banner that shoves your layout down. The reverse happens too: a scary lab LCP that field data shows real users never actually feel, because they arrive on warm caches. Knowing when a number is noise and when it is signal is interpretation, and interpretation comes from having been fooled before. The most expensive mistake in a DIY audit is spending a week fixing a number that was never the real problem.
Pattern recognition
When I open a store, I am not starting from zero. I have seen the same theme bloat, the same app conflicts, the same checkout leaks across many stores. A finding that looks isolated to you reads, to me, as the third symptom of a known cause.
That recognition is the difference between fixing a symptom and fixing the root. A tool reports the symptom. It cannot tell you that three separate findings, a slow LCP, a layout shift, and a render-blocking script, all trace back to one badly configured review app loading its widget on every page. You would chase three fixes. The pattern says there is one. That single insight often saves more time than the whole audit costs.
Estimating effort and impact in hours and euros
This is the one that matters most and the one no tool attempts. "Reduce unused JavaScript" is a line in a report. Is that two hours or two days? Does it need a developer or can your theme handle it? What revenue lift can you reasonably expect?
Turning a technical finding into a number of hours and an expected euro return is judgment built on having quoted and shipped these fixes before. It is what lets you decide whether a fix is worth doing at all. A finding that costs three developer days to ship and moves a metric nobody feels is a finding you should refuse, and only an estimate in hours and euros lets you say no with confidence. This is the heart of the 30%, and it is what the deliverable in a paid audit actually sells.
Decision matrix: DIY or pro audit?
The honest answer depends on your stage, not your budget. Here is how to decide.
| Situation | DIY is enough | Pro audit is worth it | |---|---|---| | Monthly revenue | Under 5k€, early stage | Above 10k€, every point of conversion counts | | Your time | You have a free afternoon and enjoy the digging | Your time is worth more than 590€ a day | | Technical comfort | You can read a Network tab without panic | DevTools and crawlers are a foreign language | | What you need | A list of problems to learn from | A ranked plan with hours and expected return | | Decision at stake | Curiosity, general health check | A redesign, a migration, or a real budget to allocate |
If you are early, curious, and not yet bleeding money on slow pages, run the 70% yourself. You will learn your store, and that knowledge compounds. The free tools are genuinely good.
To make it concrete, picture three stores. The first does 3k€ a month, the owner has Saturday free and likes tinkering: this is a clear DIY case, and paying 590€ would buy a plan they have time to build themselves. The second does 15k€ a month and is about to commit to a theme redesign: here the audit decides what the redesign should fix first, and getting that order wrong costs far more than the fee. The third does 8k€ but the owner has no technical comfort and no spare afternoon: the DIY path exists in theory, but in practice the audit buys back a week they do not have.
If you are past the point where guessing is cheap, where a week spent fixing the wrong thing costs more than the audit, then the 30% is what you are paying for. Not the findings. The prioritization, the interpretation, and the costed plan.
DIY or expert: not a money question, a stage question
Whether you run the audit yourself or hire someone is not really about 590€. It is about where your store is.
Early on, the constraint is knowledge, and the DIY path is the better teacher. You run PageSpeed, you read Search Console, you crawl the site, and you build an instinct for your own store that no report can hand you. Do it. Seriously. The afternoon you spend in DevTools now is the afternoon that makes every future decision faster.
Later, the constraint flips to time and confidence. When the cost of fixing the wrong thing first is higher than the cost of the audit, the math changes. That is the moment the expert earns the fee, by handing you a plan you can act on without second-guessing it. You stop paying for a list of problems and start paying for the certainty of knowing which one to solve on Monday.
Either way, start with the data. If you only do two things this week, run the INP method I wrote up and measure what your apps cost you. Those two alone will tell you most of what you need to know about whether your store is healthy or quietly leaking money.
Decide from the data, not from a sales page. That is the whole point.
