Original research · v1.0 · 2026-11-13
What care plans cost, and what they quietly do not promise.
We read the published pricing pages of 169 website and WordPress care plan providers in 12 countries, and recorded every tier and every stated feature. The prices are unremarkable. What the plans decline to promise is the interesting part.
The sample
- Providers checked
- 169
- Publish a price
- 121 (72%)
- Priced tiers recorded
- 369
- Median price a month
- £69
What the data shows
Six findings
Each of these reports what the published pages say. A provider may do more than it advertises; a buyer comparing plans has no way to know that, which is the point.
Only 16% of plans say updates are tested before they go live
This is the single most common cause of a site breaking, and it is the thing care plans are least willing to promise. Most pages describe the opposite order: apply the updates, then look at the site, then restore a backup if something went wrong. The industry's safety story is that it can undo the damage, not that it checked first.
Paying more does not reliably buy it
Testing is more common at the top of the market, but even above £200 a month two thirds of plans still say nothing about it. Where it does appear it is frequently reserved for a provider's most expensive tier, which means the buyer who most needs protection, the one on the cheapest plan, is the least likely to have it.
Accessibility is effectively unsold, at 2% of tiers
A year after the European Accessibility Act began to be enforced, almost no maintenance plan mentions accessibility at all, and the handful that do offer automated detection rather than fixing anything. Whatever else these plans are keeping an eye on, it is not whether the site can be used.
Everyone sells backups. Almost nobody sells a tested restore
Backup frequency is the most advertised feature in the sample and the most common way to tell tiers apart. A restore that has actually been performed is mentioned by a handful of providers in the entire set. An untested backup is a belief, not a control.
Only 32% of tiers state a response time
Plenty promise priority support, rapid response or 24/7 access without ever naming a number. If a plan does not say how quickly someone will start work, the honest reading is that nothing has been promised.
Headline prices are not comparable
Roughly a quarter of tiers bundle hosting and the rest either exclude it or stay silent. VAT is handled differently in every market, and some providers quote per site while others quote per plan. Two plans at the same price can differ by a factor of several in what they actually cover.
What plans actually say
Share of 372 tiers that mention each thing
Monitoring and scanning are nearly standard. The things that determine whether an incident happens, and how fast it ends, are not.
Does paying more help?
A little, and less than you would hope
If price bought safety, the right-hand column would climb steeply. It climbs, but even in the top band two thirds of plans still say nothing about testing updates before they reach your live site.
Testing is also the feature most often reserved for a provider's most expensive tier. The customer on the entry plan, who is least likely to have a developer of their own, is the least likely to be protected.
60 of 372 tiers say updates are tested before going live, and 97 of the 121 providers that publish prices never mention it on any tier.
| Price band | Tiers | Tested first | Response time |
|---|---|---|---|
| Under £50 | 134 | 8% | 22% |
| £50 to £99 | 98 | 11% | 35% |
| £100 to £199 | 73 | 21% | 33% |
| £200 and above | 64 | 34% | 50% |
The prices
What people actually charge
Converted to monthly pounds at indicative rates on the collection date. Annual and quarterly prices are divided down. No adjustment is made for bundled hosting or VAT convention, both of which vary.
By tier position
| Tier | Count | Median |
|---|---|---|
| Cheapest | 123 | £35 |
| Tier 2 | 105 | £69 |
| Tier 3 | 90 | £126 |
| Tier 4 | 33 | £158 |
Ranked within each provider, so “cheapest” is that company's entry plan rather than the cheapest in the market.
By region
| Region | Entry plan | All tiers |
|---|---|---|
| Continental Europe and hosts | £30 | £59 |
| United Kingdom and Ireland | £35 | £59 |
| United States and Canada | £78 | £142 |
North American plans run roughly twice the European and British medians. The gap is wider than the difference in what the plans promise.
If you are buying one
Six questions worth asking
Drawn from what the sample leaves out. None of them is hostile, and a good provider will have ready answers to all six.
If you already have a care plan, these work just as well as a review. The answers tend to be more revealing than the invoice.
Do you test updates before they reach my live site, and where?
The most common cause of an outage, and the promise the market is least willing to put in writing. A good answer names an environment.
When did you last restore one of my backups, and how long did it take?
Backup frequency is marketing. A dated restore is a control. If the answer is never, you have a backup policy and no recovery capability.
How quickly will someone start work when the site is down, and does that cover nights and weekends?
Priority support is not a response time. Ask for the number and the hours it applies to.
What is explicitly not included, and what does it cost?
Malware cleanup, custom development and content work are the usual carve-outs. Ask for the hourly rate now rather than during an incident.
Does the price include hosting, and is it per site?
The two things that make published prices incomparable. Settle them before comparing anyone.
What happens to my site if I cancel?
Some plans are only available on the provider's own hosting, which turns a maintenance decision into a migration.
Method
How this was collected
Public pages only
Every figure was read from a provider's own published pricing page on the collection date. Nothing was taken from roundup articles, directories or comparison posts, because those recycle each other and go stale. Where a secondary source disagreed with a live page, the live page won.
Nothing estimated
If a page did not state something, it is recorded as unstated rather than guessed. That is why several figures below are shares of tiers that say a thing, not shares of tiers that do it. A plan may well test updates without saying so; the point is that the buyer cannot tell.
Converted, not adjusted
Prices are converted to monthly pounds at indicative rates on the collection date, which are recorded with every row. Annual and quarterly prices are divided down. No adjustment is made for purchasing power, VAT convention or bundled hosting, all of which vary and are noted separately.
A sample, not a census
This is 169 providers across 12 countries, not the whole market. It is large enough to show the shape of the industry and too small to call any single figure definitive. The sample grows at each refresh.
Known limitations
- Providers who publish prices are not a random sample of providers. The 48 in the set who publish none may behave differently.
- A handful of pages contradict themselves between tier tables and body copy. Those features are recorded as unclear rather than resolved in either direction.
- Exchange rates move. Comparisons across regions are indicative, and the rate used is stored with every row.
- A few providers block automated access. They are excluded rather than counted as publishing no price.
About the index
What people ask before using these figures in a decision.
How much should a website care plan cost?
Across 369 priced tiers the median is about £69 a month, and a provider's cheapest tier has a median of about £35. But the spread is enormous and price predicts very little: plans that say they test updates have a median of £130 against £64 for those that do not, which is a real difference and still leaves plenty of expensive plans promising nothing. Decide what you need covered first, then find the price for that.
Why do you report what plans say rather than what they do?
Because the published page is what a buyer has to decide on, and it is the only thing we can verify from the outside. A provider may test every update carefully and simply not mention it. That is a marketing problem for them and a buying problem for you, and either way it is the fact worth reporting.
Is a more expensive plan a better plan?
Not reliably. Higher prices do correlate with stated response times and with testing, but the correlation is weak enough that you cannot use price as a shortcut. Several plans in the top price band promise less in writing than plans at a quarter of the price.
Which providers are in the sample?
169 providers across 12 countries: independent agencies, dedicated maintenance companies, freelancers and managed hosts that sell a care product separately from hosting. 48 of them publish no price at all, which is itself one of the findings.
Can we cite this?
Yes, and we would rather you did than repeat a figure from a roundup article. Cite the version and the collection date, and link back so readers reach the current numbers rather than a snapshot.
Hold your provider to it, including us
We publish this partly because our own plan has to stand up to it. If you want the six questions answered about your current provider, or about us, ask.