Answers / Old business web apps

Should you refresh, rewrite, or replace an old PHP web app?

Matt Turley, Continuum. Updated September 29, 2026.

Your business runs on a web app someone built years ago. PHP, MySQL, HTML tables. It works, but it looks dated, it is slow in places, and you worry about the day it breaks. You want to freshen it up, and someone has told you it needs a full rebuild.

The short answer

Should you refresh, rewrite, or replace an old PHP web app?

If the app works and your business runs on it, refresh it: update the look, fix the slow and broken parts, and move it to a supported PHP version in stages while it keeps running. Rewrite only when the code can no longer be changed safely, and switch to off-the-shelf software when your process is standard and an existing tool already does the job.

Refresh vs rewrite vs off-the-shelf

RefreshRewriteOff-the-shelf SaaS
What it meansKeep the app. New look, fixed weak spots, supported PHP, in stages.Build a new app from scratch, then move your data and users over.Stop running your own app. Move to a subscription tool.
CostAt Continuum, a Ship Fix is $4,000 to $15,000, or ongoing work on a retainer.At Continuum, from $15,000 for a first version. Clutch: most projects $10,000 to $49,999.A monthly fee per plan or per user. Jobber, for example: $49 to $499/mo.
TimelineFirst visible changes in weeks. The app stays live while it changes.Six to twelve weeks for a first version, longer to match everything the old app does.Days to weeks, mostly spent moving data and retraining staff.
Who it fitsThe app works and holds rules your business depends on.The code cannot be changed without breaking something else.Your process is standard for your industry.
Main riskSome old code limits what you can change quickly.It runs long, and the old app still needs care the whole time.You bend your process to the tool, and you do not own it.

Sources: Clutch software development pricing (September 2026), Jobber pricing (month-to-month plans, September 2026). Continuum prices are from the pricing page.

Why a refresh is usually the right first move

An old app that still runs your business holds years of decisions: pricing rules, edge cases, the fix for the customer who always books twice. None of that is written down anywhere except the code. A refresh keeps all of it and changes what people see and what keeps breaking.

A refresh also ships in pieces. The booking screen gets a new look first. Then the slow report gets fixed. Then the app moves to a supported PHP version. Each step goes live on its own, so you see progress in weeks and never have a big switch-over day.

I have done this. TourPass came to me as somebody else's codebase. I redesigned and refactored it, and I have kept it shipping since early 2025.

When a rewrite is worth it

  • Every small change breaks something unrelated, and nobody can say why.
  • The app is built on something that is no longer maintained and cannot be upgraded in place.
  • What the business needs now is so different that little of the old app would survive anyway.
  • There is no working test or staging copy, and one cannot be built from what exists.

Go in with open eyes. The new app must match everything the old one does before anyone can switch, and that list is always longer than it looks. Joel Spolsky called rewriting from scratch "the single worst strategic mistake" a software company can make, after Netscape went almost three years between major releases.

When off-the-shelf software is the better answer

If your app mostly does what a standard tool does, like scheduling, invoicing, or bookings that work the same way everyone else's do, a subscription may cost less than keeping your own app alive. Check that you can export your data first, and price it over three years, not one. See off-the-shelf vs custom software for how to make that call.

The deadline you may not know about

According to php.net, every PHP version before 8.2 is past end of life and gets no security fixes. PHP 8.2 loses security support on December 31, 2026. If your app runs on PHP 7 or 5, the upgrade comes before any redesign. A new look on an unpatched server still leaves the door open.

When not to hire me

  • An off-the-shelf tool already does what your app does. Switch, and save the build money.
  • You want the old app copied line for line onto a new framework and nothing else. I will push back on that, so hire someone who will just do it.
  • The app has many developers working on it every day and needs a team lead, not a builder.

Not sure if yours needs a refresh or a rebuild?

Tell me what the app does and what bothers you about it. I will tell you which of the three it needs, including when the answer is a subscription tool. If you want it in writing first, a $299 Ship Check is a read of your codebase that says whether it is fine, fixable, or a rewrite.

Book a call →See all services and prices →
FAQ

Straight answers

Can you modernize an old PHP app without rewriting it?
Usually, yes. Most of the work people call a rebuild is really three smaller jobs: a new look on the screens customers see, fixes to the slow or fragile parts, and an upgrade to a supported PHP version. Each one can ship on its own while the app stays live. You keep your data, your URLs and the rules your business already runs on.
Is my PHP version still supported?
Check it against php.net. As of September 2026, every PHP version before 8.2 is past end of life and gets no security fixes. PHP 8.2 gets security fixes only until December 31, 2026. If your app runs on PHP 7 or 5, upgrading it is the first job, before any redesign.
Refactor vs rewrite: what is the difference?
A refactor changes how the code is organized without changing what the app does, in small steps, while it keeps running. A rewrite throws the old code away and builds the app again from scratch. Refactoring keeps years of fixes and edge cases your business depends on. A rewrite has to rediscover all of them.
How long does a rewrite take?
Longer than anyone plans for, because the new app has to do everything the old one does before you can switch. Clutch puts the average reviewed software project at about 13 months. The classic warning is Netscape, which rewrote its browser from scratch and went almost three years between major releases.
Should I just switch to an off-the-shelf tool instead?
Sometimes. If your business works the way most businesses in your field do, a subscription tool may cover it for less. Field-service software like Jobber, for example, lists plans from $49 to $499 a month without a contract. Check two things first: whether you can export your data from the old app, and whether the tool forces you to change how you work in ways that cost you.
Keep reading
Ship CheckI read your codebase and tell you what it needs. $299.Ship FixMake a shaky app stable and shippable. $4,000 to $15,000.Who should do the workAgency, freelancer, or one senior developer who stays.TourPassAn inherited booking platform, redesigned, refactored and kept running.