Let's be blunt about the incentive here. A rebuild costs more than a focused fix, so you should be skeptical of anyone who recommends one before they've looked closely. That includes us.
So here's the diagnostic we actually run, and what it usually turns up.
Start with where people are leaving
Before you look at the design, look at the behavior. Pull up your analytics, find the pages getting the most traffic, and for each one find what share of visitors leave without doing anything. Then find the pages that actually bring in work.
You're looking for the shape of the problem. There are three common shapes, and they call for completely different responses.
Traffic shows up and leaves immediately. People land and bounce within seconds. That's almost always load time, or a mismatch between what brought them there and what they found. Neither one needs a redesign.
Traffic stays, reads, and doesn't convert. People engage with the content and then don't take the next step. That's a messaging, proof, or call-to-action problem. Sometimes design plays a part, but usually further down the list than people assume.
Traffic never shows up at all. The site is fine and nobody sees it. That's a search and demand problem, and rebuilding won't fix it. This is the case where a redesign gets sold most often and is warranted least often.
The three fixes that solve most cases
In most audits we run, the highest-return work isn't a rebuild. It's some mix of these three.
Make it load fast on a phone. Compress and size the images properly, stop loading fonts that block rendering, rip out the four analytics and chat scripts nobody uses. More than half of mobile visitors bail on a page that takes over three seconds, so this often beats any amount of visual redesign. It's also the least glamorous work there is, which is part of why it gets skipped.
Rewrite the first screen. Not the whole site. Just the headline, the subheading, and the main call to action on your three busiest pages. Say exactly what you do and who you do it for, and make the next step obvious. It costs a day and moves conversion rate more than anything else on this list.
Fix the form. Most contact forms ask for more than they need, and every extra field costs you completions for nothing. If you're asking for a phone number, a company size, and how they heard about you before they've even decided to talk to you, you're trading away leads for data you never look at.
All three together usually cost a small fraction of a rebuild and take a couple of weeks.
When a rebuild genuinely is the answer
There are four situations where we'll recommend rebuilding, and we try to be careful about it.
You can't edit it without breaking it. If every content change puts the layout at risk, or it's sitting on a platform or plugin stack nobody maintains anymore, you're paying a tax every month. That compounds, and eventually paying it off costs less than living with it.
The structure can't hold what you sell now. If you have six service lines and the site has one generic services page, no amount of copy editing fixes that. You need pages that don't exist yet, and if the foundation can't hold them, that's a rebuild.
Speed can't be salvaged. Some sites are slow because of a few fixable mistakes. Others are slow because of decisions baked in at the foundation: a page builder shipping enormous amounts of unused code, or a theme carrying ten years of cruft. You find out which by trying the cheap fixes first and measuring.
Your positioning genuinely changed. If who you sell to or what you sell has moved, the site is describing a company that doesn't exist anymore. That's a real reason, and it's the one that best justifies starting over.
Notice that "it looks dated" isn't on the list. Looking dated does cost you something, but you can usually fix it without touching the foundation. It's also the reason given most often for rebuilds that never paid for themselves.
How to get an honest answer
Ask whoever is quoting you to show the data behind the recommendation. Specifically: which pages are underperforming, what mobile load time looks like right now, and what they expect to change. Someone who actually looked will answer in specifics. Someone who didn't will talk about modern design, best practices, and how the current site doesn't reflect your brand.
Then ask what the cheapest version of a fix would be. Anyone who won't describe a smaller option isn't giving you a recommendation, they're giving you a quote.
How we approach it
This is a big part of why we sell audits. One costs a fraction of a build, comes back as a written recommendation with the evidence attached, and gets credited against the build if you move forward. A real share of them end with us saying a smaller piece of work will do the job.
That's an awkward business model, and it's the one we're comfortable standing behind. The alternative is recommending a rebuild to everybody and hoping nobody checks.