In short
- The order is speed, mobile, copy, phone number, forms, photos, structure, tracking. Each fix makes the next one measurable.
- Most of the list can be done on any platform, including a builder. Two or three items sometimes cannot.
- Starting over throws away rankings, addresses and years of small lessons. Do it only when the platform is the ceiling.
- A week of focused fixes usually beats a season of redesign for a site that still works.
- Know the ceiling before you spend: if the platform cannot get fast, the money is better spent on the move.
The case for not starting over
There is a moment in the life of every old website when the owner is ready to burn it down. The site is slow, the photos are from 2017, the form goes to an inbox nobody checks. Starting again feels like the only honest option.
Pause on that. The old site has things a new one will not have on launch day. Google has trusted its addresses for years. Other sites link to it. Its pages rank for searches that took time to earn. A redesign that begins from a blank page gives all of that up and then spends months buying it back, and the redesign pillar lists how often that goes wrong.
The alternative is to treat the old site as a house with good bones and a bad kitchen. Fix what is broken, in order, on the platform it lives on, and stop when the numbers say the site is working. Usually that gets you there for a fraction of a rebuild. Sometimes it hits a ceiling, and this guide says plainly where.
The fix list, in the order that pays
The order is not arbitrary. Each item makes the next one visible. There is no point rewriting the copy for visitors who leave before it loads, and no point fixing the form until you know how many people reach it.
- 1
Speed.
Compress every image; most old sites carry photos ten times larger than the screen needs. Remove the scripts nobody remembers adding: unstaffed chat widgets, pixels for campaigns that ended, fonts loaded from three places. The Core Web Vitals guide walks through the scores; speed work as a service is a fixed price per site.
- 2
Mobile.
Open every main page on a phone. Fix text you cannot read at arm's length, buttons you cannot tap, layouts that overflow sideways, menus that hide the phone number. On most platforms this is a settings and stylesheet job; mobile fixes cover the cases that are not.
- 3
Copy.
Rewrite the first line of every page to say what the page is for and where. Delete “welcome to our website.” Replace the paragraph about the company's journey with a sentence about the customer's problem. Update every service, price and town that has changed. The content guide has the structure.
- 4
Phone number.
In the header, on every page, as a link that dials. This is a ten-minute fix on any platform and it is missing on a remarkable share of old sites.
- 5
Forms.
Cut to three or four fields. Test that it sends, that the message reaches a person, and that the sender gets a confirmation. Add a spam filter that does not punish real customers. Form fixes are a common one-day job.
- 6
Photos.
Replace stock images with photos of your actual work and people. A phone in good light is enough. This is the fix that changes how the site feels most for the least money, and it needs no developer at all.
- 7
Structure.
One clear title per page. Headings that describe the sections. A page for each service you want to be found for, linked from the home page and from each other. Fix or redirect every broken link; broken link repair is quick, and the SEO checklist covers the rest.
- 8
Tracking.
Install conversion tracking that counts calls and forms by page and source. It is last only because it measures everything above. If you can do only one thing this month, do this one, so that next month you know what to do. Tracking setup is a one-time job.
An old site fixed in the right order is a new site that kept its rankings.
What can be done on your platform, and what cannot
Here is the honest map. "Yes" means a competent person can do it without leaving the platform. "Partly" means you can get most of the way. "No" means the platform decides.
| Fix | Drag-and-drop builder | WordPress | Custom coded |
|---|---|---|---|
| Speed | Partly. Images and scripts, yes. The builder's own weight, no. | Yes, with care. Theme and plugin bloat is the usual limit. | Yes |
| Mobile layout | Partly, within the template's rules. | Yes, if the theme is sound. Some themes cannot be saved. | Yes |
| Copy, phone number, photos | Yes | Yes | Yes |
| Forms | Yes, within the builder's form tool | Yes | Yes |
| Structure and redirects | Partly. Titles yes; redirects and structured data often limited. | Yes | Yes |
| Tracking | Partly. Basic analytics yes; call and form events can be awkward. | Yes | Yes |
| New features (booking, quoting, portals) | Only what the app store offers | Through plugins, with the weight they bring | Built to fit; see adding features |
The [WordPress versus custom guide](/guides/diy-vs-professional/wordpress-vs-custom-website) and the [Wix versus custom guide](/guides/diy-vs-professional/wix-vs-custom-website) go deeper on each column.
The three fixes with the largest effect on whether visitors stay, speed, mobile and structure, are the three a builder limits. The fixes anyone can do anywhere, copy and photos and the phone number, are the ones most old sites have also skipped. Do those first, whatever the platform. They are free.
When the platform is the ceiling
You will know you have hit it when a fix from the list cannot be done, or can only be done by making something else worse. The signs are consistent.
The speed test stays orange or red after the images are compressed and the scripts are removed, because what is left is the platform itself. The mobile layout can be improved only by deleting content, because the template has no other way to make room. The feature the business needs next, booking or a calculating quote form or a customer login, exists only as a slow third-party widget that looks like someone else's site. The theme has not been updated in two years and every plugin update is a gamble, the pattern WordPress plugin repair exists for.
At that point the money spent on fixes buys a year, not a solution, and the honest advice is to stop spending it. A move to a platform without the ceiling, done with the redirects and content carried over, costs more once and then stops costing. The WordPress to Next.js page describes how DataCram makes that move without losing rankings, and replacing WordPress covers the same for sites that have outgrown it.
What the fixes cost and what they return
Fixes are cheap next to a rebuild, and the return arrives faster. A typical custom rebuild runs about $6,000 to $20,000 and takes weeks. The items above are mostly days of work each, priced by the job, and three of them cost nothing but your time.
The return follows the order. Speed fixes tend to show up in the analytics within a week as longer sessions and more calls from the same traffic. Mobile and copy show up within a month. Structure shows up in search over two or three months. Tracking returns nothing on its own and makes every other return visible, which is why skipping it is how most improvement projects end with nobody knowing whether they worked.
Fixing broken websites is DataCram's fixed-price version of this list: you say what is wrong, we say what it costs, and the site stays on the platform it lives on. The website cost guide has the ranges for the rebuild if the fixes turn out not to be enough.
8
Fixes on the list, in the order that makes each one measurable
3
Of them cost only your time: copy, phone number, photos
Days
What most of the developer items take, not weeks
A one-week plan
If you want to run the list yourself before calling anyone, here is a week.
- Monday: run the free audit and a speed test on your phone. Write down the scores. Compress every image over 200 KB.
- Tuesday: open every page on a phone. List what you cannot read or tap. Fix what the platform lets you; note what it does not.
- Wednesday: rewrite the first line of every page. Fix the phone number in the header. Update every price, service and town.
- Thursday: shorten the form. Send a test. Confirm a person receives it. Take twenty photos of real work.
- Friday: install or check conversion tracking. Fix broken links. Search your main service and town and note where you appear; the free visibility check does this for you.
- Next month: compare the numbers. If the site improved, keep going. If the speed score did not move, re-read the ceiling section and get a fixed-price quote for the developer items.
Get a number before you compare anyone else.
Tell us what you have and what you want more of. A real person looks before we quote, and the quote is a fixed price you can hold us to.
Fix it or replace it? Find out before you spend.
Tell us what your site is doing wrong and we will tell you, in plain words, whether it needs a repair or a rebuild. Free, and the answer is yours either way.
Which of these sounds like you? (pick any)
Questions people ask
Straight answers
In most cases, yes. Speed, mobile layout, copy, the phone number, forms, photos, page structure and tracking can all be fixed on the platform the site already lives on, and doing so keeps the rankings and addresses the old site has earned. The exceptions are a builder that cannot be made fast no matter what is removed, a theme that cannot work on a phone, or a feature the platform simply cannot take.
Speed, because every other fix is invisible to visitors who leave before the page loads. Compress the images, remove scripts that no longer do anything, and check the host. Then the mobile layout, then the first line of copy on each page, then the phone number in the header. Each fix makes the next one measurable, which is why the order matters more than most owners expect.
Replace the stock photos with real ones, remove the slideshow, enlarge the body text, rewrite the first line of each page to talk about the customer instead of the company, update the copyright year, and make it fast. Those six changes do more for how old a site feels than a new layout would, and none of them requires leaving the platform. A new design on top of slow loading and stock photos still reads as old.
It depends on the theme and the plugins. A WordPress site on a well-maintained theme with a reasonable plugin count can usually be made fast, mobile-friendly and well structured where it is. One on an abandoned theme with a heavy page builder and dozens of plugins often cannot, and every fix becomes a fight with the next update. If the speed score stays red after the images and scripts are cleaned up, the platform is the limit.
Start with images: most old sites serve photos far larger than any screen needs, and compressing them is the biggest single gain. Then remove scripts that no longer earn their place, such as unused chat widgets and old tracking pixels. Then check the host and any caching. If the score is still poor after that, what remains is the platform's own weight, and only a lighter platform fixes it.
Far less than rebuilding it. Copy, photos and the phone number cost your time. Speed, mobile and form fixes are typically a few days of developer work each, quoted at a fixed price by the job. Structure and tracking are a day or two each. A full custom rebuild, by comparison, runs roughly $6,000 to $20,000, so a set of fixes usually comes to a fraction of that and shows results within weeks.
Where to go from here