Website Audits
How to Run a Free Website Audit Before You Pitch a Client
You can run a useful website audit on a prospect in about ten minutes, for free, using tools you already have. The goal is not a long report. It is to find one or two problems you can prove, write them in plain words, and decide whether the business is worth contacting at all.
Below is a time-boxed process, the free tools worth using and what each one cannot tell you, a template for a one-page mini audit you can share, and the limits you should respect when you look at a site that is not yours.
Two audits, not one
Most people blur two jobs. The first is deciding whether to pitch. The second is giving the prospect something worth reading. Keep them apart.
| Private pre-pitch audit | Shareable mini audit | |
|---|---|---|
| Purpose | Decide whether to contact, and what to lead with | Show the owner one clear problem, with evidence |
| Audience | You | The business owner |
| Length | Notes, a few screenshots | One page |
| Tone | Blunt and detailed | Neutral and helpful |
| Time | About 10 minutes | An extra 10 to 15 minutes, only for warm prospects |
Do the private audit for every shortlisted business. Do the mini audit only when someone replies, or when you have a strong reason to offer one first.
The free toolkit
Pick from this list. You do not need all of it every time. The tools guide covers the wider set of options, including how to find the businesses in the first place.
| Tool | What it does | What it cannot do |
|---|---|---|
| PageSpeed Insights | Runs a lab test and, when there is enough traffic, shows real-user field data | Field data is often missing for small sites, and lab scores vary between runs |
| Lighthouse in Chrome DevTools | Audits performance, accessibility, SEO and best practices in your own browser | It is a lab test on your machine, so results depend on your connection |
| WAVE from WebAIM | Flags accessibility errors and alerts, including contrast problems | Automated tools catch only part of accessibility |
| Rich Results Test | Checks a page's structured data | It tells you what is present, not whether it helps rankings |
| The business's Google listing | Shows rating, review count, hours, category and contact details | It shows the listing, not how well the site converts |
| Browser device toolbar and Network tab | Shows how the page looks on a phone and how much it downloads | It approximates a phone, so always confirm on a real one |
| Screaming Frog SEO Spider (free version) | Crawls a site for broken links, titles and descriptions, up to 500 URLs per crawl | The free version is limited, and crawling a small site is only worth it when you suspect problems |
Two things people expect to find are not on this list. Google Search Console needs the site owner to verify access, so you cannot use it on a prospect. And Google retired its old standalone Mobile-Friendly Test, so Lighthouse and PageSpeed Insights are what to use instead.
The ten-minute plan
Work in this order, with a timer running. Each step is short on purpose.
Minutes 0 to 2: use it as a customer
Open the site on your phone. Can you tell what the business does, and how to call or book, without zooming? Tap the phone number. Take a screenshot of the first screen.
Minutes 2 to 4: the quick trust checks
Look for the padlock, read the browser tab title, and check the footer year. Search the business name and read how its listing looks in Google.
Minutes 4 to 7: run the tests
Run PageSpeed Insights on mobile and note the main content time and the top two issues. Run WAVE and note the number of errors, and any low-contrast text.
Minutes 7 to 9: read the listing
Open the Google listing. Note rating, review count, recent activity and whether the phone number and hours match the site.
Minute 9 to 10: choose the finding
Pick the single clearest problem. Write one sentence saying what a visitor cannot do, and which service it points to.
Here is the same plan as a rough time budget. It is a suggestion, not measured data.
Example (illustrative). A suggested time budget, not survey data. Adjust it to how you work.
If you get to minute four and already have a strong, provable problem, stop. You do not need to prove five things to justify one email.
What to look at, and in what order
Order matters because the fastest checks are also the most persuasive. Start with what a visitor experiences, then move to what a tool measures.
- Can a customer act? Phone tappable, booking or contact path visible, a call to action on the first screen.
- Does it load and behave on a phone? Layout, text size, speed, popups.
- Does it look trustworthy? Padlock, reviews, real photos, up-to-date content.
- Can it be found? Title and description in the search result, and consistent contact details.
- Can everyone use it? Contrast, labels, alt text.
If you want the full set of checks with what each one means, use the website audit checklist. For the common problems this process tends to surface, see 15 website problems local businesses ignore.
Stay passive and public
An audit should look at a site the way a customer would. That has practical limits.
- Public pages only. Do not try to get past a login, a paywall or any access control.
- No security probing. Do not run vulnerability scanners, try to find admin pages or test for weaknesses on a site you do not own. It is not what you are being asked to do, and it can cause harm or legal trouble.
- Go gently with crawlers. If you crawl with a tool, keep the speed low and stay to a handful of pages.
- Do not submit forms with fake enquiries. A test message costs the owner time. Read the form instead, and check that it exists and asks for a sensible amount of information.
- Be honest about how you found it. In outreach, say what you looked at and how.
Turn your notes into a one-page mini audit
Once someone shows interest, put your best finding on a single page. A mini audit that fits on one screen gets read. A twelve-page report does not.
| Section | What goes in it |
|---|---|
| Header | Business name, address of the site you looked at, date checked |
| The finding | One or two sentences describing what a visitor cannot do or experiences |
| The evidence | One screenshot, annotated, and the test result if relevant |
| Why it matters | One sentence tying it to calls, bookings or search visibility, without invented numbers |
| What I would do first | The smallest fix that solves the problem, in plain words |
| What comes after | Two or three other things you noticed, listed briefly |
| Next step | An offer to walk through it in fifteen minutes, with no obligation |
Here is what the core of one looks like, for a fictional business.
How much to give away
A common worry is that a free audit gives the work away. It does not, if you give the finding and the direction of the fix, and keep the full implementation and the priorities for a paid conversation.
- Give: the problem, the evidence and the first fix you would make.
- Hold back: the complete list of issues, the implementation plan and the pricing.
- Never give: a shaming report, a list of 40 technical warnings, or a security assessment.
Whether to send a mini audit without being asked is a judgment call. Unrequested attachments can hurt deliverability, and some owners read an unrequested audit as a sales trick. A softer route is to offer it in your first message and send it when they say yes, as the cold email guide suggests.
Mistakes that waste audit time
- Auditing on desktop only. Start on the phone, then confirm on desktop.
- Treating lab scores as real-user results. Say which one you are quoting.
- Believing every tool warning. Automated checks produce false alarms. Confirm with your own eyes before you report something.
- Reporting everything. You are choosing the one finding that helps a business owner act.
- Turning it into a project. Ten minutes is enough for most prospects. Longer audits belong to paying clients.
Automating the pre-pitch audit
Ten minutes per site is fine for twenty businesses. It is not fine for two hundred. SiteOpportunity runs the same kinds of checks for every business in a market, saves screenshots for desktop and mobile, and shows the evidence behind each finding, plus a separate Confidence score that tells you how complete the evidence was. You review the results and decide what to send. See what a result looks like or read how to interpret the scores.
Want the audit done across a whole market before you pick who to contact? Run a free search
Questions about running a pre-pitch audit
How long should a pre-pitch website audit take?
About ten minutes per business is enough to find one provable problem and decide whether to contact them. Save longer audits for prospects who have replied or become clients.
Do I need the business owner's permission to audit their website?
Looking at a public page the way a customer would does not need permission. Do not bypass logins, probe for security weaknesses or run aggressive crawls, and be honest about how you found the problem when you get in touch.
Is a free audit a good way to win clients?
It can be, if it is short, specific and neutral. A one-page mini audit with one clear finding tends to work better than a long report, and it is best offered rather than sent unrequested.
Which free tool should I use first?
Your own phone. Use the site as a customer, then run PageSpeed Insights, and add WAVE and the Google listing. Most strong findings appear in the first few minutes without a tool.
Can I use Google Search Console on a prospect's site?
No. Search Console needs the site owner to verify access. Use PageSpeed Insights, Lighthouse and a Google search of the business name instead.