Most business websites don’t fail all at once. They drift. A homepage that made sense two product launches ago. A contact form nobody’s rewritten since the business had half the staff it has now. A checkout flow that was fine when order volume was a tenth of what it is today. Nobody decides to let it happen – it just accumulates, quarter after quarter, until one day the website is describing a company that no longer exists.
The hard part is that “the site looks a bit dated” and “the site is actively costing you revenue” feel identical from the inside. Below are nine signals that separate the two, each one backed by data rather than a gut feeling about your homepage’s font choices.
1. Your conversion rate has flatlined while everything else kept growing
Traffic is up. Ad spend is up. Headcount is up. And your conversion rate has sat within half a percentage point of the same number for three straight quarters. That’s not a coincidence – it’s a page-level problem, because a healthy site should convert incrementally better as you refine offers and audiences, not stay static while you pour more money into the top of the funnel.
The average website conversion rate across industries sits somewhere between 2.5% and 3.5%, and the gap between the median site and the top 25% of performers is roughly double that. If your number hasn’t moved in nine months despite real investment elsewhere, the site itself – not the offer, not the ad copy – is the bottleneck. I’d go further: a flat conversion rate for a growing business is the single most reliable “outgrown” signal on this list, because it’s the one number that can’t be explained away by seasonality or a slow news month.
2. The bounce-and-abandon math nobody’s watching
Google’s own research found that as mobile load time stretches from one second to ten, the probability of a visitor bouncing rises by 123%. Put more plainly: every extra second is doing measurable damage before a visitor reads a single word of your pitch. Multiple independent analyses converge on roughly a 7% conversion loss for each additional second of load time in the first few seconds, and 53% of mobile visitors abandon a page that hasn’t loaded within three seconds.
If your business runs any kind of online cart or booking flow, the checkout step compounds the problem. Baymard Institute’s checkout usability research, built on more than a decade of testing live production sites, puts the average documented cart abandonment rate at 70.22% – and finds that the average large ecommerce site could lift its conversion rate by roughly 35% purely through better checkout design. That’s not a rounding error; on real order volumes, it’s the difference between a good quarter and a mediocre one.
3. “Can you just email the developer?” is now part of your workflow
This is the sign that shows up in Slack messages, not analytics dashboards. If publishing a blog post, swapping a hero image, or updating a price requires opening a ticket and waiting two days, the CMS has become a tax on every marketing decision the business tries to make. Teams in this position quietly stop publishing altogether, which then shows up months later as a separate, seemingly unrelated problem: declining organic traffic.
I’ve seen this pattern described the same way across dozens of agency case studies: the backlog of “quick” content changes grows faster than anyone can clear it, and eventually the marketing team just stops adding to the list. That’s not a content problem. That’s a platform that has outgrown its usefulness to the people running it day to day.
4. Your homepage still describes the company you were three years ago
Businesses evolve – new services, new audiences, new pricing tiers, sometimes an entirely different core offer than the one the site was built to sell. But websites don’t evolve on their own, and a surprising number of growing businesses are still running a homepage that answers a question customers stopped asking a while back.
Here’s my honest opinion on this one, because it’s the sign every listicle leads with and I think that’s a mistake: an outdated look is real, but it’s the least dangerous item on this list. Stanford’s web credibility research found that about 46% of people judge a site’s credibility on visual design alone, which matters – but a site that looks fine and structurally can’t support the business is a bigger liability than a site that looks a little tired and still converts. Fix positioning and messaging first. The paint job is a distant second.
5. Mobile visitors get the leftovers, not the main course
Mobile now accounts for well over half of global web traffic, and Google’s index has been mobile-first for years. Yet it’s still common to find sites where the mobile experience is visibly an afterthought: desktop images crammed into a smaller viewport, primary calls-to-action buried below a collapsed menu, forms that require zooming to complete.
The 2025 Web Almanac’s performance chapter, built on real Chrome user data, found that only 48% of mobile sites pass all three Core Web Vitals thresholds – Largest Contentful Paint, Interaction to Next Paint, and Cumulative Layout Shift – compared with 56% on desktop. Largest Contentful Paint is the specific metric dragging that number down: just 62% of mobile origins load their main content within Google’s 2.5-second “good” threshold. If you’ve never actually opened your own site on a mid-range Android phone over a 4G connection, that’s worth doing this week before you read the analytics.
6. AI answer engines have never heard of you
This is the newest sign on the list, and the one I’d argue business owners are most likely to shrug off – which is exactly why it belongs here. ChatGPT, Perplexity, Gemini, and Google’s AI Overviews now answer a meaningful share of commercial research queries directly, without the user ever clicking through to a website. These systems favor fast, semantically structured pages with clear heading hierarchy and schema markup; a site built on outdated templates with tangled code and no structured data essentially doesn’t exist to them.
If a prospect asks an AI assistant to compare vendors in your category and your business never comes up, you’ve lost a sale you didn’t even know was in motion. Traditional SEO metrics won’t show this gap, because the visitor never generated a session to measure.
7. The accessibility debt has turned into legal exposure
WebAIM’s 2026 accessibility audit of the top one million homepages found an average of 56.1 detectable accessibility errors per page, and 95.9% of homepages had at least one detected WCAG 2 failure – both figures that actually got worse compared with the year before, reversing a six-year trend of gradual improvement. The most common issues are mundane: low-contrast text, images without alt text, form fields without labels. None of them are hard to fix individually. All of them add up.
This stopped being a purely ethical argument a while ago. Federal ADA digital accessibility lawsuit filings have been climbing year over year, with e-commerce sites the most frequent target, and WCAG 2.2 is now the operating standard auditors measure against. If your site hasn’t had an accessibility pass in the last 18 months, treat that as a compliance gap, not a nice-to-have.
8. A prospect compares you to a competitor, and you’ve already lost
This is the soft signal founders are most reluctant to admit, and the one salespeople notice first. If a prospect open two tabs – yours and a competitor’s – and the competitor’s site simply feels more current: cleaner hierarchy, faster load, more confident copy, they’ve often half-decided before your sales rep even picks up the phone. Competitive presence isn’t vanity. It’s conversion infrastructure that happens upstream of any conversation your team actually has.
9. Patch-and-pray maintenance now costs more than a rebuild would
Track two numbers honestly: how many hours per quarter your team (or your agency) spends patching bugs, updating plugins, and working around limitations on the existing site, versus a realistic estimate for rebuilding on a modern stack. For a lot of businesses that math has quietly flipped, and the current site is now the more expensive option even before counting the revenue it’s leaving on the table.
Security patching adds urgency here too. WordPress powers a huge share of the web, and vulnerability tracking firms have reported the ecosystem’s disclosed vulnerabilities rising sharply year over year, with a majority found in plugins rather than the core software. An old site running a stack of half-maintained plugins isn’t just slow – it’s a growing security liability that someone eventually has to account for.
So how many of these actually matter?
Hitting one or two of these signs probably means a targeted fix: a faster checkout, an updated homepage headline, a plugin audit. Hitting four or more is a different conversation – at that point you’re not patching a website, you’re propping up a structural problem with duct tape, and every month you wait adds to both the lost-revenue side of the ledger and the eventual rebuild cost. Here’s roughly where the numbers land if you want to benchmark your own site against the current data rather than a hunch:
| Metric | Current benchmark | Source |
|---|---|---|
| Mobile Core Web Vitals pass rate | 48% of sites pass all three | HTTP Archive, 2025 Web Almanac |
| Mobile LCP under 2.5 seconds | Only 62% of mobile sites achieve this | HTTP Archive, 2025 Web Almanac |
| Average website conversion rate | 2.5%–3.5% across industries | Aggregated industry benchmarks, 2026 |
| Average cart abandonment rate | 70.22% | Baymard Institute checkout research |
| Homepages with a detectable WCAG failure | 95.9% | WebAIM Million, 2026 |
Read that table for what it says about the field, not just your own site: even the median business is failing at least one of these benchmarks. The question isn’t whether your site has some gap – it’s whether the gap is costing you customers you’d otherwise keep.
The counterargument: not every business needs a full rebuild
It’s worth steel-manning the opposite position, because plenty of agencies make it convincingly: if your site’s structure, brand, and content are still fundamentally sound, targeted repairs can outperform a full rebuild on both cost and speed. A site that’s one to two years old with a couple of these signs is usually a candidate for a focused refresh – new headlines, a faster checkout, an accessibility pass – not a twelve-week rebuild. Spending five figures to rebuild something that needed a two-week fix is its own kind of waste.
Where I land: the threshold is roughly three or more signs, or a site that’s crossed the three-year mark, whichever comes first. Below that, fix the leaks first. Above it, the maintenance and opportunity cost of staying put usually outpaces the cost of doing it properly, and a partial fix on a structurally outdated site tends to just delay the same conversation by a year.
What to actually do next
Start with a one-page brief, not a designer search: what business outcome should the new site drive, who exactly are you designing for, and what would success look like in numbers you can actually track – conversion rate, qualified leads, organic traffic, AI citation share. If your business sells online and cart abandonment or checkout friction shows up on your own numbers anywhere near that 70% industry average, that’s the specific problem to brief a partner on rather than a vague “make it look better” request. For businesses in that position, working with a specialist Ecommerce Website Design Company tends to produce a more targeted result than a generalist agency treating your checkout the same way it treats a static brochure page.
How this article was put together: the statistics above were pulled from primary sources current as of mid-2026 – HTTP Archive’s 2025 Web Almanac performance chapter, WebAIM’s 2026 Million accessibility report, and Baymard Institute’s ongoing checkout usability research – alongside aggregated industry conversion-rate benchmarks cross-checked across multiple 2026 analyses. Figures on ADA lawsuit filings and WordPress vulnerability counts are directional trends reported by legal and security trade publications rather than a single authoritative dataset, so treat those two as indicative rather than exact. Core Web Vitals thresholds and cart abandonment figures should be rechecked in 12 months, since both move meaningfully year over year.
