Stack audit · Bookkeeping Sites
Web hosting for bookkeepers — Three checks, ten minutes, and a list of things not to buy
A small practice site rarely has a performance problem worth a project — it has an out-of-date runtime, an unverified mail path and one plugin doing more than you realised.
The short answer
On a five-page practice site the highest-value work is an audit, not an optimisation. Three checks cover almost all of it: confirm the runtime is PHP 8.x with OPcache active, confirm a cached page returns its first byte well inside Google's 800 ms good boundary, and confirm that mail you send from the site actually arrives and authenticates. Each takes a few minutes and each finds a real fault more often than a caching plugin would.
For bookkeeping in particular the third matters most, because your site's job is to start a correspondence. The infrastructure under that does not vary by tier: NVMe storage, LiteSpeed caching in the server, free SSL, free migration, mailboxes on your own domain and humans on support at any hour, from $2.42/mo with the renewal rate matching the first term.
By the HostingFast team · Reviewed 24 August 2026
99.9%
Availability target
24/7
People on shift
Free
Certificates, all plans
NVMe
Disk class, entry included
There is an industry of advice aimed at sites a hundred times the size of a bookkeeping practice's, and applying it produces effort with no measurable result. A site of five pages and a contact form does not have a caching architecture problem. It usually has one of three specific faults, and they are all cheap to find.
The other reason to audit rather than optimise is that the site's commercial job is narrow: name the software you work in, state your monthly packages, and let an accountant or a client start an email thread. Delivery of that email is closer to the revenue than any millisecond on the page.
What follows is the audit in order, plus an explicit list of the things it is not worth spending money on at this size — because that list saves more than the tuning does.
Check one: what is actually executing
Open your application's own health screen and read the PHP version it reports, then compare it against the current stable releases. Sites carried across from an old host frequently arrive still pinned to a runtime several years old, which costs real speed on every uncached request and eventually costs plugin compatibility as well.
While you are there, confirm OPcache is active. It keeps compiled PHP in memory so the interpreter is not re-parsing your plugin files on every request, and its absence is a quiet tax on the exact routes — form submissions, admin work — that were never going to be cached. Neither check requires a terminal, and both are the sort of thing that stays wrong for years because nobody looks.
Check two: one number for the server
Run `curl -s -o /dev/null -w '%{time_starttransfer}' https://yourdomain.example` five times and keep the median. Against Google's 800 ms good boundary, a cached page off LiteSpeed and NVMe — a UK connection to our London origin, 5–15 ms of round trip — should come back with a great deal of room to spare.
If it does not, check the cache status header before anything else. On a small site the usual culprit is a single plugin that disables caching sitewide, and finding it is worth more than every other item on this page. If the number is good, you have just eliminated hosting as a suspect, which is the most useful thing an audit can do.
Check three: does the mail arrive, and does it authenticate
Send a real test from the site's contact form to an address you control, then read the received headers. You are looking for two things: that it arrived at all, and that SPF and DKIM pass. A form that reports success while its mail lands in spam is the most expensive failure a bookkeeping site can have, and nothing on the page will hint at it.
Mailboxes on your own domain come with every plan and outbound mail goes through MailChannels, which is the part we can do for you. Publishing the right DNS records and re-testing after any DNS change is the part that stays yours — and it is worth repeating the test after every such change rather than assuming it survived.
The list of things not to buy
Not a second caching plugin: the cache lives in the web server here, and two things setting the same headers is a fight rather than a feature. Not a CDN, unless you have measured a real asset-weight problem — it handles images, CSS, fonts and cached HTML, and does nothing for form submissions or database queries. Not a bigger plan, until you can name the uncached route that is queueing.
What is worth doing is unglamorous: one font family, images at display dimensions, no plugin installed for a feature you are not using. That list, plus the three checks above, is the whole performance strategy for a site this size, and it is honest about being short.
The parts that are already handled
Current cPanel with a supported PHP runtime, NVMe storage on every tier, LiteSpeed caching inside the web server, free SSL that renews itself and daily backups restorable from the panel.
Mailboxes on your own domain with spam and virus screening as standard, and MailChannels handling outbound delivery — so the correspondence your site exists to start actually lands.
- LiteSpeed caching handled below the application layer
- NVMe storage on the tier you are actually paying for
- Domain mailboxes with screening included
- MailChannels handling outbound delivery
Why HostingFast
Standard on every plan
A runtime that is current
The latest cPanel with a supported PHP version, so the audit finds a modern stack rather than a migration's leftovers.
Caching without configuration
Server-level LiteSpeed means there is nothing to install and nothing to tune — and nothing for a second plugin to argue with.
Mail with a delivery path
Domain mailboxes plus MailChannels on outbound. The test still has to be run, but the infrastructure behind it is not the weak link.
Screening on every mailbox
Spam and virus filtering as standard, so inbound client correspondence is not being lost in the other direction either.
SSL that never lapses
Free certificates issuing and renewing themselves on every plan. One fewer thing that can quietly break between audits.
Support that will say it is fine
Humans at any hour who will tell you the site does not need a bigger plan, because selling you one you do not need is a poor long-term trade.
Quick Start
From order to online
- 1
Read the runtime, not the marketing
Application health screen: PHP version and OPcache status. Compare against current stable releases. Two minutes, and it is wrong more often than you would expect.
- 2
Take one server number
Five curl runs, median TTFB, plus the cache status header. Either hosting is eliminated as a suspect or you have found a caching plugin to remove.
- 3
Prove the mail path end to end
Real test from the live form to an address you control, then read the headers for SPF and DKIM. Repeat after every DNS change rather than hoping.
Built In
Loaded onto every plan
- PHP version read off the application health screen and compared to current
- OPcache confirmed active, so plugin files are not re-parsed per request
- Median of five curl timings kept, not the best run, and held to the 800 ms boundary
- Cache status header checked on a logged-out request
- A real form submission delivered to an address you control
- SPF and DKIM confirmed as passing in the received headers
- Mail path re-tested after any DNS or nameserver change
- No second caching plugin layered on the server-level cache
- Plugins removed for features the site is not actually using
- Daily backups restored once so the recovery path is proven
Frequently Asked
What people ask us most often
How do I check which PHP version my site is running?
Open your application's own health or system-info screen and read the version it reports, rather than trusting what the account was set to at some point in the past. Compare it against the current stable releases. Sites migrated from an older host commonly arrive pinned to something years old, which costs measurable time on every uncached request and eventually breaks plugin updates.
Does OPcache matter on a site this small?
It matters on precisely the requests a page cache cannot help: form submissions, logins, admin work. OPcache keeps compiled PHP in memory so the interpreter is not re-reading and re-parsing your plugin files every time. On a small site the effect is invisible on cached pages and clearly noticeable in the dashboard, which is a good way to confirm it is doing something.
My contact form says it sent. How do I know it arrived?
By sending a real one to an address you control and reading the received headers, checking that SPF and DKIM pass. Delivery and submission are separate things, and a form can report success while its message is being filtered as spam at the far end. Do this test again after any DNS change, because that is exactly when authentication quietly breaks.
Is there anything at this size where a bigger plan genuinely helps?
Only if you can name the uncached route that is queueing — and on a five-page practice site you almost never can. What does help is a current PHP runtime, one font family, images at display dimensions, and fewer plugins. If you outgrow the tier later, upgrades happen in place from the client area with no migration, so there is nothing to lose by starting small.
Keep reading
Beauty Spas — Hosting Guide
Gallery weight on a treatment-led site, measured on a throttled mobile profile.
Churches & Faith Groups — Hosting Guide
Volunteer-run sites, small budgets, and the checks that find real faults.
Best Hosting With Free SSL
Free certificates on every plan, and how to confirm a renewal actually happened.
Mail Filtering — Built for Bookkeepers
Inbound filtering on every mailbox, and what it does to a client correspondence.
WordPress Hosting
Managed WordPress with Redis object caching for the routes that skip the cache.
Changing hosts? Run through our checklist first.
A straightforward sequence for a switch your visitors never feel: which files move first, how to shift email across without losing a single message, the right moment to repoint DNS, and the two mistakes behind almost all the downtime we get asked to rescue.
Audit it before you optimise it.
Current cPanel and PHP, NVMe under every tier, LiteSpeed in the server, and domain mailboxes with MailChannels on outbound.
View Mail Filtering — Built for Bookkeepers plans