If you're already asking the question, the answer is probably yes.
Most websites don't fail loudly. They bleed quietly. Load time creeps up half a second a year. Mobile conversion stays flat while mobile traffic doubles. The stock photography starts feeling three years old before anyone says so. None of it breaks. All of it costs you.
Here's the part that should concern you. A B2B site that loads in 1 second converts roughly 3x better than the same site loading in 5 seconds (Portent, 2022). The site still "works". Traffic still arrives. You're just losing two out of three deals before anyone notices.
The seven checks below are the same ones we run before deciding whether a site gets optimised, rebuilt, or left alone. You can do them yourself in an afternoon.
The short version
A B2B site loading in 1 second converts 3x better than the same site at 5 seconds (Portent, 2022). Only 48% of mobile pages pass all three Core Web Vitals (HTTP Archive Web Almanac, 2025). 94.8% of the top 1 million home pages failed WCAG 2 accessibility checks (WebAIM Million, February 2025). 61% of B2B buyers now prefer a rep-free buying experience (Gartner, 2025). If your site can't sell for you, it isn't a website. It's a brochure.
The foundation fails before the paint does. Audit before you redesign.
Why "it still works" is the wrong bar
Most teams audit their site the way you'd inspect a car for a road trip: does it start, does it brake, does the heater work. The problem is that a website isn't a vehicle you occasionally drive. It's a storefront that's open every second of every day to a buyer who has already decided whether they trust you within 50 milliseconds.
"Working" is a qualitative question. "Holding you back" is a quantitative one. The answer isn't whether the contact form submits. It's how much revenue the site is quietly leaving on the table between now and your next redesign.
B2B buyers now spend just 17% of their total purchase time with any one potential supplier (Gartner, 2025). That's less than a fifth of the consideration window. The rest is self-research, peer research, and comparing vendors on their websites before ever opening an email thread.
Trade-off honesty: not every site with imperfect metrics is "holding you back". If your pipeline, revenue, and retention are all on plan, you probably don't have a website problem. You have a bandwidth question. This audit is for the other case, where the numbers stopped following the strategy and nobody can point to why.
How fast does your site load, really?
Slow enough that you're losing money you could recover with engineering, not marketing.
The Portent analysis of 27,000 B2B and e-commerce landing pages found conversion rates drop sharply with each additional second of load time. The Deloitte/Google "Milliseconds Make Millions" study measured the same effect in the other direction: a 0.1-second speed improvement lifted retail conversions by 8.4% and average order value by 9.2% across 30+ European and US brands (Deloitte Digital, 2019 A/B study).
On mobile it gets worse. Google's deep-neural-network analysis of 11 million mobile sessions found that when load time goes from 1 second to 3 seconds, bounce probability jumps 32%. At 5 seconds it's up 90%. At 10 seconds, 123% (Think With Google).
What to check, in order: pull your site through Page Speed Insights, then into the Chrome User Experience Report (CrUX) for 28-day real-user data, then into Google Search Console's Core Web Vitals report. Field data from actual visitors beats lab scores every time. If your mobile Largest Contentful Paint is above 2.5 seconds, that's where you start.
Does your site pass Core Web Vitals?
Probably not on mobile. Only 48% of mobile pages and 56% of desktop pages passed all three Core Web Vitals in the 2025 Web Almanac crawl of roughly 16 million sites (HTTP Archive). That's the ranking floor Google uses as part of its page-experience signals, and more than half the mobile web fails it.
The three metrics worth auditing:
Largest Contentful Paint (LCP). How long until the biggest above-the-fold element loads. Target: under 2.5 seconds.
Interaction to Next Paint (INP). Replaced First Input Delay as a Core Web Vital in March 2024. Measures how snappy your site feels on every interaction, not just the first. Target: under 200 ms.
Cumulative Layout Shift (CLS). How much the page jumps around while loading. Target: under 0.1.
In the sites we've audited for prospective clients, INP is where most businesses fail without knowing it. A site built in 2021 passed FID cleanly. The same site, tested against INP in 2026, reveals every slow JavaScript handler the old metric hid. If you haven't retested since March 2024, you're reading stale numbers.
Is your site actually mobile-first, or just responsive?
Responsive means your site doesn't break on a small screen. Mobile-first means your site was designed for that screen and adapted up. These are not the same thing, and the gap between them is where mid-market websites quietly lose revenue.
Google judges visual appeal in 50 milliseconds and credibility follows (Lindgaard et al., Carleton University). Stanford's Web Credibility research found that 75% of users form their judgement about a company's credibility based on website design (Stanford). On a phone, held at arm's length, that judgement is even less forgiving.
What to audit on mobile specifically:
Tap targets under 44 x 44 pixels. If a thumb can't hit it cleanly, it's a drop-off.
Forms that reveal more than three fields above the fold.
Hover states that exist only on desktop and quietly kill the mobile version of the interaction.
Pop-ups or cookie banners that cover the primary call-to-action on iPhone SE width (375 px). This is the one most teams never check because nobody on the team uses an iPhone SE.
Responsive is a 2012 standard. Mobile-first is a 2026 baseline. If your designer still starts in Figma at 1440 px and shrinks down, you're building the site for the person who signs the invoice, not the person who's actually reading it.
Can 70 million people use your site?
Most likely, no. The 2025 WebAIM Million scan of the top 1,000,000 home pages found 94.8% had detectable WCAG 2 failures, averaging 51 distinct errors per page (WebAIM). Low-contrast text, missing alt text, form fields without labels, links that don't announce themselves to a screen reader. Most of it is invisible to the people building the site, and blocking to the people who need it.
28.7% of US adults, over 70 million people, report having a disability. That's cognitive, vision, hearing, and motor disability combined. Add temporary impairments (a broken wrist, a bright outdoor screen, a noisy train) and the share of users with access barriers is larger still.
There's a legal tail to this. UsableNet tracked 4,000+ ADA digital-accessibility lawsuits in 2024, and 25% of those cases named companies that already had an accessibility widget installed. The widget isn't the fix. The widget is the tell that nobody fixed the underlying code.
Trade-off honesty: automated tools catch roughly a third of real accessibility issues. A WAVE scan is the floor, not the ceiling. Manual keyboard-only testing and a screen-reader pass on the three highest-traffic templates will find more than any plugin.
Does your website do the selling, or wait to be sold to?
Your website is your hardest-working salesperson. On most mid-market sites it's also the least supervised.
61% of B2B buyers now prefer a rep-free buying experience, and 75% of B2B buyers in HubSpot's 2024 survey said they prefer to research independently before talking to a vendor (HubSpot). Combined with the Gartner finding that buyers spend just 17% of their time with any one supplier, the implication is simple. If your site can't answer the buyer's real questions, compare honestly against alternatives, and produce an on-ramp to a conversation, it isn't doing half the job.
The benchmarks worth measuring yourself against. Median B2B landing-page conversion rate: 3.8%. All-industry median: 6.6% (Unbounce Conversion Benchmark Report, 57 million conversions across 44,000 pages, 2025). Global e-commerce conversion rate: 1.89% all-industry average, 1.7% on the IRP global 2025 rolling average (IRP Commerce).
If your site is meaningfully below these for your category, the question to ask is not "how do we drive more traffic?" It's "what happens to the traffic we already have?"
Content without strategy is just expensive noise. The same logic applies to a landing page. If the call-to-action, the proof, and the pricing can't line up in the reader's head in under 30 seconds, they won't.
Is your tech stack a liability?
Sometimes it is. The question is whether you know, and whether the cost of not knowing is about to find you.
Patchstack's State of WordPress Security found 7,966 new WordPress-ecosystem vulnerabilities disclosed in 2024, a 34% year-on-year increase, with 96% of them in plugins (Patchstack). Over 1,000 of those were in plugins installed on more than 100,000 sites. Verizon's 2024 Data Breach Investigations Report found that exploitation of vulnerabilities as an initial breach vector nearly tripled, a 180% increase over the prior year (Verizon DBIR).
WordPress works. Until it doesn't. The decision isn't ideological, it's operational. A small marketing site with a handful of plugins, maintained by a team that patches weekly, is fine. A ten-year-old enterprise install with 40 plugins, three inherited from an agency that no longer exists, and a staging environment nobody dares touch: that's not a website, that's a risk register.
The minimum checks: plugin count, last-updated date on every active plugin, PHP version, SSL certificate expiry, last security audit, last penetration test. If you don't know any of these numbers, the stack is already a liability.
What do you do if the audit says yes?
You fix the foundation before you turn on the tap.
The mistake most businesses make after an audit is ordering a redesign. A redesign treats the symptom (it looks dated) and not the cause (the CTA, the load time, the mobile flow, the missing structured data). Six months and a quarter of a million euros later, the new site looks beautiful and converts the same, because nobody fixed the reason the old site converted poorly in the first place.
Three rough bands of intervention, in order of cost:
Optimisation. Same site, surgical fixes. Plugin audit, image compression, copy rewrite on the top three pages, CTA restructure, analytics cleanup. Weeks, not months. This is where most sites should start.
Re-platform. Same content, new foundation. Move off a CMS that's fighting you, onto a stack that isn't. Preserve URL structure and content. Months, not quarters.
Rebuild. New site, new strategy, new positioning. Justified when the business has moved and the website hasn't, or when the gap between brand and site has become visible to buyers. This is the option most sold and least often needed.
We're not going to pretend there isn't a trade-off. Optimisation is cheapest and fastest but hits a ceiling. Rebuild is the most expensive and the biggest lift for the team. The right answer depends on where the real bottleneck is, which is exactly what the seven checks above are meant to reveal.
Frequently Asked Questions
How long should a self-audit of my website take?
A basic audit covering the seven checks above takes a focused afternoon. Run your top three pages through PageSpeed Insights and the CrUX report, do a keyboard-only pass on your main conversion flow, pull your Core Web Vitals from Search Console, and benchmark your landing-page conversion rate against the 3.8% B2B median from the 2025 Unbounce Conversion Benchmark Report. That's enough to tell you whether you have a fixable problem or a structural one.
How often should I re-audit my website?
Twice a year at minimum. Google moves metrics (INP replaced FID in March 2024), browsers ship updates, plugins break, and content decays. A quarterly mini-audit on Core Web Vitals and conversion benchmarks is more useful than an annual deep-dive that nobody acts on. Write the checks into a recurring calendar invite. Most mid-market sites we see haven't been audited since the last redesign, which means years.
Do I need to rebuild my website if it fails some of these checks?
No. Most failures are fixable without a rebuild. A 2.5-second mobile LCP can often be solved with image optimisation and script deferral. An accessibility pass on the top ten templates can clear the majority of WCAG flags. Reserve a rebuild for when the strategy behind the site has shifted (new audience, new product, new positioning) and the current site can't carry it. Redesign to solve presentation problems, re-platform to solve performance problems, optimise to solve almost everything else.
What's the minimum viable fix if a full rebuild is out of budget?
Start with the top three pages by traffic. Homepage, primary services page, and your highest-traffic blog or landing page. Fix load time below 2.5 seconds, pass Core Web Vitals, clear your top ten accessibility flags, and rewrite the calls-to-action for clarity. Those four moves, on three pages, typically recover the majority of the revenue leak without touching the rest of the site. The 80/20 of website problems almost always lives on the pages buyers actually land on.
What's the difference between a performance issue and a positioning issue?
A performance issue is a site that loads slowly, fails Core Web Vitals, or leaks conversions against benchmark. A positioning issue is a site where everything works but nothing persuades, because the site is talking to the wrong buyer, or about the wrong problem. Performance issues show up in Search Console and GA4. Positioning issues show up in your sales team's mouth, usually as "our leads don't know what we do until the call." Two different fixes. One is engineering, the other is strategy.
So, is it?
If you've read this far, you already have a guess. The seven checks will give you the number behind it.
If that's the problem you're actually trying to solve, let's talk.