Insights
Launch is not the finish line: why a website needs a support plan and somewhere to test
A website is not a thing you buy once. It is a thing that runs — and like anything that runs, it needs looking after by someone who knows what they are doing, with somewhere safe to try things first.
By Jad Khater, Owner & founder · · 16 min read
Does a website really need a support plan after launch?
Yes, on any platform. A live site depends on software that keeps changing: plugins and apps on their own release schedules, a theme you customised, a language version that goes end of life, and integrations whose APIs move under it. A support plan is someone keeping that stack current, with somewhere safe to test every change first.
Most websites are sold as a project with an end date. The site goes live, the invoice is settled, everyone moves on — and the thing quietly starts rotting, because software does that whether or not anyone is watching it.
This is not an argument that you should buy a support plan from us. It is an argument about what actually happens to a live website in its second year, on either platform, and what it takes to stop the worst of it.
WordPress vs Shopify: what you still have to maintain
| WordPress / WooCommerce | Shopify | |
|---|---|---|
| Platform and server patching | Yours, or your host’s | Shopify’s |
| Plugins or apps | Twenty to thirty plugins, each on its own schedule | Apps, each third-party, several usually touching checkout |
| Theme customisations | Yours | Yours — and a theme update can undo them |
| Your data | Yours to back up and restore | Yours to export; Shopify will not roll back a bad import |
| Somewhere to test | A full one-click copy on most managed hosts | Theme level only — no products, collections or pages |
| PCI scope | Yours | Shopify’s, for checkout |
| Typical way it breaks | Broken into through a plugin | Broken by an app update |
| What a support plan mostly buys | Updates, testing, backups, monitoring | Attention, app discipline, data exports, monitoring |
How risky is an unmaintained WordPress site?
Patchstack’s State of WordPress Security in 2026, which reports on 2025, found 11,334 new vulnerabilities in the WordPress ecosystem — a 42% increase on the year before. Of those, 4,124 were serious enough to need protection rules and 1,966 carried a high severity score.
Where they were matters as much as how many. 91% were in plugins and 9% in themes. WordPress core itself had six, all low priority. So the risk is not “WordPress is insecure”; the risk is the twenty or thirty plugins your site runs on, each maintained by someone you have never met, on their own schedule.
Two findings in that report should change how you think about updates:
- The median time to mass exploitation of a heavily exploited vulnerability is five hours. Not five days. A plan that says “we update plugins monthly” is not a security measure — it is a hope that nothing interesting happens in the other twenty-nine days.
- In a large-scale penetration test of popular hosting companies, only 26% of vulnerability attacks were blocked. Good hosting is necessary and it is not sufficient. “Our host handles security” is the single most common and most expensive misunderstanding we meet.
Then there is the part nobody thinks about until it breaks: the language the site runs on. PHP’s own support schedule moves every year — as of late 2026, PHP 8.2 is on security fixes only, and anything older than that is end of life and receives nothing at all. A site left alone long enough ends up on an unsupported version, at which point upgrading is no longer a routine task but a project, usually an urgent one.
If Shopify is managed, what is left for you to maintain?
Shopify genuinely does remove a category of work. It runs the infrastructure, keeps the platform patched, carries the checkout’s PCI burden and absorbs your traffic spikes. If you have ever spent a weekend on a compromised WordPress site, that is worth real money.
What it does not do is look after your store. Still yours:
- Apps. Every app is third-party software with its own release schedule, its own bugs and its own commercial decisions. Apps change, break, get acquired, raise prices and get abandoned — and several of them are usually touching your checkout.
- The theme. Your customisations are yours. A theme update can undo them, and theme-store themes are only licensed to the store that bought them.
- Your data. Products, collections, pages, customers, orders, metafields. Shopify protects its platform; the contents of your store are not something you can ask Shopify to roll back after a bad import or a mistaken bulk edit.
- Integrations. Payment gateway, courier, accounting, ERP, marketing. Each one is an API that can change under you.
- Everything operational. Discounts, inventory rules, shipping zones, tax settings — the settings that quietly decide whether orders work.
The practical consequence: a Shopify store needs less *system* maintenance and about as much *attention* as a WordPress one. The failure mode is simply different. WordPress sites get broken into. Shopify stores get broken by an app update at 11pm on a Thursday.
Why do you need a staging environment?
There is one rule underneath all of this: never test on the thing that is taking money. A staging environment is simply a copy of your site that nobody is buying from, where a change can fail harmlessly.
Without one, every update is a live experiment on your customers. With one, the same update is a five-minute check. The difference in cost is not subtle, and it shows up on exactly the days you can least afford it.
WordPress
This is WordPress at its best. Most managed hosts create a one-click staging copy of the whole site — files, database, plugins, content — and push changes back when they work. A proper routine looks like this: clone to staging, apply updates there, click through the paths that matter (checkout, forms, logins, anything custom), then repeat on production with a fresh backup taken first and a known way back.
And the backup point deserves saying plainly: a backup you have never restored is not a backup, it is a belief. Restoring one occasionally is the only way to find out whether it works, and the day you need it is the wrong day to find out.
Shopify
Here you have to be realistic, because Shopify does not offer staging in the sense above. What it offers is theme-level testing: duplicate the live theme, work on the copy, preview it, publish when it is right. Shopify’s own documentation recommends duplicating a theme before you customise it, and that is the correct habit.
But know its limits, which Shopify states on that same page: theme downloads don’t include your products, collections, menus, pages, blog posts or images, and a store is capped at 20 themes, with deletion permanent. So a duplicated theme lets you test a template. It does not let you test an app update, a checkout change, a bulk price edit or an integration — those happen on the live store, with real customers in it.
Which means the discipline has to come from elsewhere:
- Change one thing at a time, so that when something breaks you know what broke it.
- Never install or update an app during a campaign, a sale, or the week before one.
- Keep a duplicate of the last known-good theme rather than relying on undo.
- Export your data on a schedule. Shopify gives you CSV exports; a backup app automates them. Either way, the responsibility is yours.
- Place a test order after anything that touches checkout, including app updates you did not initiate.
How should updates actually be applied?
The interface is misleading. WordPress offers a tick-all box and an Update button; Shopify shows a list of apps with updates waiting. Both make it look like one action. It is not one action — it is twenty or thirty separate pieces of software changing at once, and if the site breaks afterwards you have twenty or thirty suspects and no way to tell which.
The discipline is unglamorous and it is the whole job: one critical component at a time, and a testing session after each one.
- Take a backup first, and know how long restoring it takes. Not last night’s — one from minutes ago.
- Work on the staging copy, where a failure costs nothing.
- Update one critical component. Just one.
- Test it, with the checklist below, not by glancing at the home page.
- Write down what changed and when. This is how “it broke on Tuesday” becomes “we updated the payment gateway plugin on Tuesday” instead of a guess.
- Then the next one, and repeat.
- Batch the genuinely low-risk ones at the end — and still click through the site afterwards.
- Repeat the whole sequence on production, with a fresh backup and the same tests.
What counts as critical is anything a sale or an enquiry passes through, and anything with custom code attached to it: the payment gateway, cart and checkout, forms, booking or membership, the page builder, the theme, and WooCommerce or WordPress itself on a major version. Those get one update, one test, one line in the log. A caching plugin or an icon pack does not need the same ceremony.
A testing session is a scripted walk — the same one every time, so nothing depends on remembering:
- The home page, a category or landing page, and a product or service page.
- Search, filters, and anything custom that was built for this site specifically.
- Add to cart and place a real test order, through the payment method your customers actually use.
- Every form that matters, including the one that sends you enquiries — and confirm the email arrives.
- Login, account pages, and the admin screens your team uses daily.
- The same journey on a phone, which is where most of your visitors are.
- The browser console and the server error log, because plenty of breakage is silent on screen.
Major versions — WooCommerce, PHP, a theme framework — are not part of a routine cycle at all. They get their own scheduled window, their own rollback plan, and a day when someone is available to deal with the consequences. Never a Friday, never during a campaign, never the week before a sale.
On Shopify the same rule applies with less room to manoeuvre, because you are doing it on the live store: update one app at a time and place a test order after each. An app that quietly stops sending orders to your courier, or stops a discount applying, will not announce itself — you will hear about it from a customer, days later, after it has cost you.
What does a competent support provider actually do?
This is the part that sounds like marketing and is not. The difference between a competent support arrangement and an expensive one is specific and checkable.
- They read changelogs before clicking update, and they know which of your plugins or apps are load-bearing.
- They test on a copy first and have a written way back — a backup from minutes ago, a previous theme, a rollback step.
- They watch rather than wait. Uptime, errors, checkout failures and Core Web Vitals are monitored, so problems are found before a customer reports them.
- They track the boring expiries — domain, SSL, DNS, PHP version, API credentials. A domain lapsing on a Friday has ended more trading weekends than any hacker.
- They restore a backup occasionally, deliberately, to prove it works.
- They keep access tidy. Who has admin, what the old developer still has, what the ex-employee’s account can still do.
- They tell you what they did. A plan with no report is a subscription, not a service.
What you are buying is not hours. It is the judgement to know which update is routine and which one needs a Tuesday morning and a rollback plan — and the fact that someone is paying attention to your site when you are not thinking about it at all.
How do you judge a support plan before signing?
Ask these before signing anything, including with us:
- What is the response time, and is it a response or a resolution? These are different promises and the difference is where disappointment lives. A plan that promises to *fix* anything in two hours is either expensive or not telling the truth.
- What hours does it cover, and what happens outside them?
- What counts as covered, and what gets quoted separately? New features are not maintenance.
- Where do backups live, how long are they kept, and when was one last restored?
- Is there a staging environment, and who pays for it?
- What am I sent, and how often? Updates applied, issues found, performance, uptime.
- Who exactly does the work, and what happens when they are on holiday?
- What do I keep if we part ways? Access, backups, documentation.
Our own answers are published rather than negotiated: Website Care & Digital Support covers updates, technical care and performance, and the response times are published — 1 to 2 hours for an emergency such as a site down or checkout broken, 24 hours for support questions — during 10:00 to 17:00 Beirut time. Response time is how long it takes us to respond to a request. It doesn't include the time needed to solve the problem or complete the work.
What does skipping a support plan cost?
The argument against a care plan is always the monthly fee. The argument for it is everything the fee is quietly preventing:
- A compromised site costs a cleanup, often a rebuild, and sometimes a Google warning on your own domain that takes weeks to clear.
- A broken checkout is lost revenue for every hour it is broken, and you usually find out from a customer rather than a dashboard.
- An abandoned plugin or app turns into an emergency migration at the worst possible moment.
- A site that quietly got slow loses conversions and search visibility for months before anyone connects the two.
- A year of skipped updates turns a routine afternoon into a multi-day project, because each version has to be stepped through.
None of it is dramatic until it is. The failure is always cheaper to prevent than to explain.
Which level of support fits which site?
| Your site | What it needs at a minimum |
|---|---|
| Brochure site, no transactions, few plugins | Monthly updates on a staging copy, a backup restored occasionally to prove it works, expiry monitoring |
| WooCommerce store taking payments | Prompt security updates one component at a time, staging, a test order after every change, uptime and checkout monitoring |
| Shopify store running several apps | App-by-app updates outside campaigns, the last known-good theme kept, scheduled data exports, a test order after anything touching checkout |
| Membership, course or donation platform | All of the above plus a rehearsed restore — here the data is the product |
| Nobody has touched it in over a year | A health check first: the state of it decides whether care or a rebuild is the cheaper answer |
| An agency reselling to clients | A white-label partner with published response times and reports you can forward unedited |
How do you know which support plan is right for your business?
The right plan depends on what your site does, what an hour of downtime costs you, how much custom work it carries and who is available when something breaks. There is no universal answer, and a wrong one shows up late — on the morning the checkout fails, or a year later when a routine afternoon of updates has become a multi-day project.
Five questions decide it:
- What does the site actually do — brochure, store, or a platform people log into?
- What does an hour of it being down or broken cost you, in revenue and in trust?
- How much of it is custom: plugins, apps, theme work, integrations?
- Who looks after it today, and what happens when that person is unavailable?
- When was a backup last restored, and does anyone know whether it works?
RedefinedWeb is a Beirut-based team that handles this end to end: your website, online store, payments, marketing and branding from one partner. We look after WordPress, WooCommerce and Shopify sites for businesses in Lebanon and abroad, including as a white-label partner for agencies, with 20+ years of hands-on work behind it.
What happens on a call:
- We look at what the site runs on, what state it is in and what it would cost you to have it down.
- We recommend a level of care, say plainly where you need less than you expected, and flag anything that should be fixed before a plan starts.
- You receive a written proposal with scope, timeline and cost before anything starts.
If your site is live and nobody is looking after it, start by finding out what state it is in. Our Website & UX Health Check is a fixed-scope review of performance, UX, SEO, security and plugin or app health, delivered as a prioritised list you can act on with us or with anyone else — including your current developer. If you are not sure whether the site needs care or a rebuild, that is exactly what it answers.
Frequently asked questions
- Do I really need a support plan for a Shopify store?
- Shopify maintains the platform, the hosting and the checkout’s PCI compliance, which removes a real category of risk. It does not maintain your apps, your theme customisations, your data or your integrations — and those are what usually break. A Shopify store needs less system maintenance than WordPress and roughly the same amount of attention.
- Isn’t my hosting company handling security?
- Only partly. Patchstack’s 2026 report found that in a large-scale penetration test of popular hosting companies, only 26% of vulnerability attacks were blocked. Hosting is a necessary layer, not a complete defence, because most WordPress vulnerabilities are in plugins rather than in the server.
- How often should a WordPress site be updated?
- More often than monthly. The median time to mass exploitation of a heavily exploited vulnerability is about five hours, so a fixed monthly cycle leaves long windows open. The practical approach is prompt security updates applied one critical component at a time, tested on a staging copy first, with scheduled routine maintenance around them.
- Can I just click “update all”?
- You can, and it is how most sites break. Updating twenty plugins at once means that if something fails you have twenty suspects and no way to isolate the cause. Update one critical component at a time — payment gateway, checkout, forms, page builder, theme — test after each, and record what changed and when. Batch only the genuinely low-risk ones, and still check the site afterwards.
- What should a testing session cover after an update?
- The same scripted walk every time: home, a landing page and a product or service page; search and anything custom; add to cart and a real test order through the payment method customers use; every form, confirming the email arrives; login and admin screens; the same journey on a phone; and the browser console and server error log, since plenty of breakage is silent on screen.
- What is a staging environment?
- A copy of your website that no customer is using, where changes can be made and tested before they go near the live site. On WordPress most managed hosts create one in a click. Shopify has no true equivalent — you can duplicate and preview a theme, but not your data, apps or checkout.
- Can I test changes on Shopify before they go live?
- You can test theme changes by duplicating your theme, customising the copy and previewing it before publishing, which Shopify recommends. You cannot test app updates, checkout changes or data edits in the same way — those happen on the live store, which is why changing one thing at a time and placing a test order afterwards matters.
- What should a website support plan include?
- Updates applied and tested, backups that are actually restored from time to time, security and uptime monitoring, performance and Core Web Vitals checks, tracking of expiries such as domain, SSL and PHP version, a stated response time, and a report telling you what was done. Anything without a report is a subscription rather than a service.
References and links
Terms in this market change, and some providers don’t publish them. These are the primary sources behind the guide — check the current position yourself before you commit to anything.
- Patchstack — State of WordPress Security in 2026
11,334 new vulnerabilities in 2025, 91% in plugins, five-hour median to mass exploitation, 26% of attacks blocked by hosts. Vendor research, published with methodology.
- PHP — supported versions
Which PHP versions still receive security fixes, and when each one stops.
- Shopify — duplicating themes
The recommended way to test theme changes, and the limits: no store content, 20 themes, deletion is permanent.
Related reading
Shopify or WooCommerce for a Lebanese business? The payment gateway decides it
Choosing the platform you will then have to look after.
Read it →How to accept online payments in Lebanon: areeba, MPGS, NetCommerce and Whish compared
The checkout that a bad update takes down, and who to call when it does.
Read it →
Where this fits on the site: E-commerce and website builds · Payment gateways we integrate · Stores we have built
Next step
Nobody looking after your website?
Start with a health check and find out what state it is actually in — performance, security, plugin and app health — as a prioritised list you can act on with us or with anyone else.