WEB DESIGN & BRANDING · September 2026 · ~10 min read
Hosting decisions for a small business site
Pick a host on three criteria: it stays up, it is fast for visitors near you, and you can log into it yourself. Everything else in a hosting comparison is noise for a business your size. Keep your email somewhere separate from your website, and never let the hosting account live in someone else's name.
On this page
- 01What actually matters in a host?
- 02How much is uptime actually worth to me?
- 03Managed hosting or cheap shared hosting?
- 04What should hosting cost, and who says so?
- 05What hosting mistakes cause real damage?
- 06Is my hosting the reason my site is slow?
- 07How do I know if my current hosting is a problem?
- 08What to do this week
- 09When you do not need this
- 10Sources
- 11Related reading
- 12Questions about your hosting?
Hosting is the least interesting decision in a web project and one of the few that can quietly cost you for years. It gets made in about ninety seconds, usually by whoever is building the site, using whatever they always use.
That is often fine. It is worth thirty minutes of your attention anyway, because the failure modes are slow, invisible, and annoying to reverse.
01What actually matters in a host?
Three things, in order.
Uptime. If the site is down, nothing else matters. Reputable managed hosts are all broadly fine here. The cheapest shared plans are where you find the trouble, because they pack many sites onto one machine and your neighbors' traffic becomes your problem.
Speed for your actual visitors. Not a score in a testing tool, the experience of a person in your service area on a phone. A server physically close to your customers, plus a content delivery network, covers almost all of it. If your customers are in the Bay Area or in Hawaii, that is worth checking specifically, because latency to the islands is not the same as latency to San Jose.
Access. You hold the account. Your developer gets a login under it. This is the criterion people skip and later regret.
After those three, most differences between reputable hosts are marketing. Storage limits and bandwidth caps almost never bind a local business site.
02How much is uptime actually worth to me?
Worked example
Less than the sales page implies, and you can settle it with arithmetic in five minutes. This matters because uptime is the number hosts compete on, and it is usually the smallest of your three criteria.
Step one, the units. A 30 day month is 43,200 minutes. A year is 8,760 hours.
Step two, translate the percentages. A 99.9% guarantee allows 43.2 minutes of downtime a month, or 8.76 hours a year. A 99% guarantee allows 7.2 hours a month, or 87.6 hours a year, which is three and a half days. That gap is real. Adding one more nine past 99.9% only buys back about eight hours a year, which is why the tiers above it are sold rather than needed.
Step three, convert to inquiries. Say your site produces 40 inquiries a month, or 480 a year, and assume they arrive evenly. At 99.9% availability you lose 0.1% of the year, which is about half an inquiry (derived). At 99% you lose about five (derived).
Step four, read it. Half an inquiry a year is not a reason to move hosts, buy an upgrade tier, or accept a longer contract. Five a year might be. The decision is whether your host is in the 99% class or the 99.9% class, not which decimal it advertises past that.
Step five, the part the arithmetic hides. Downtime does not arrive evenly. Seven hours spread across a year is trivial. Seven hours on the Friday before a holiday weekend is not, and no availability percentage tells you when the outage lands. That is the honest argument for a host with real support hours, and it is a different argument from the one on the pricing page.
Managed hosting or cheap shared hosting?
For most small businesses, managed. The price difference is small and the difference in what you have to think about is large.
Managed hosting means updates, caching, backups, and security are handled by people who do it all day. Cheap shared hosting means you own those problems, and you will discover that the month the site goes down.
There is a middle option worth knowing about. If your site is simple and does not need a database, static hosting is fast, extremely cheap, and has almost nothing to break. It is not right for everyone, but for a business with a dozen pages that change occasionally, it is a genuinely good answer that nobody offers because there is no recurring revenue in it.
The best host for you is the boring one your developer already knows well. A host somebody is learning on your project is a risk you are paying to take.
04What should hosting cost, and who says so?
Nobody credible, and that is the answer.
Every "hosting should run this much a month" and "maintenance should run this much a year" figure you will be shown traces back to agency pricing pages quoting other agency pricing pages. There is no survey with a disclosed sample and method behind any of them. The same applies to the build-cost ranges that circulate: a widely repeated figure claiming most small business buyers spend under a certain amount on a website is attributed to a 2026 Clutch survey through third-party pages, and the underlying report and its sample have not been produced.
This matters in a specific way. A benchmark you cannot trace cannot tell you whether you are overpaying, but it can be used to make an unusual quote look normal. When a vendor says their price is in line with industry averages, the correct question is which survey, how many respondents, and when. A vendor quoting their own pricing as their own pricing is being straight with you. A vendor quoting it as an industry standard is not.
What you can compare is like for like. Ask two hosts for the same specification, the same backup frequency, the same support response window, and compare those. And separate the hosting line from the labour line, because they get bundled precisely so that neither can be checked.
Where hosting shows up in your budget is the ongoing line rather than the build line, which is one of the pieces routinely left out when working out what a small business website actually costs.
05What hosting mistakes cause real damage?
Four, and they are all recoverable but tedious.
Email on the same account as the website. When you move hosts, your email goes down with it. Keep mail with a dedicated provider. This is a fifteen minute decision that prevents an entire category of bad days.
Nobody owns the renewal. Sites go dark because a card expired on an account belonging to a developer who moved on. Put the billing on your card and the renewal in your calendar.
No staging environment. Editing a live site is how a Tuesday afternoon becomes an outage. Any decent managed host gives you a copy to break instead.
Migrating without a plan. Moving hosts is not the same as moving domains, and it is usually safe, but it goes wrong when the redirects and the certificate are afterthoughts. If a redesign and a host change happen together, the risk compounds, and migrating a website without losing your search rankings is worth reading before you schedule either.
06Is my hosting the reason my site is slow?
Usually not. Usually it is what you put on the page, and this is where a hosting upgrade gets sold to fix a problem it cannot reach.
Google Research sampled 11 million mobile web domains over a representative 4G connection in January 2018 and found that 79% of pages were over 1MB, 53% were over 2MB, and 23% were over 4MB. It also found that 25% of pages could save more than 250KB by compressing images and text alone. A faster server delivers a 4MB page slightly sooner. It does not make it a 1MB page.
Grade the outcome against Google's published thresholds rather than a score out of 100. Largest Contentful Paint is rated good at 2.5 seconds or below, Interaction to Next Paint at 200 milliseconds or below, and Cumulative Layout Shift at 0.1 or below, each measured at the 75th percentile of visits, and all three must pass. The 75th percentile is the point: you are grading the experience of a slow quarter of your visitors, not the fast one you tested from your office.
The size of the prize is worth knowing before you spend on either. Google, Deloitte Digital and Fifty-Five studied 37 brands across roughly 30 million sessions over four weeks in late 2019 and reported that a 0.1 second improvement in mobile load time went with retail conversions 8.4% higher and lead generation information page bounce rate 8.3% better. Quote the baseline with it every time: one tenth of a second, measured against each site's own starting point, on 2019 data.
07How do I know if my current hosting is a problem?
Measure rather than guess, because hosting gets blamed for a lot of things it did not do.
Load your site on your phone, on cellular, away from your office. Then ask somebody across town to do the same. Slow in both places points at the site itself, usually large images. Slow in one place points at network or server location.
Check whether your site has ever been down without you knowing. Free uptime monitors will tell you going forward. If you have no monitoring, you have no idea, and most owners discover outages from a customer.
Then be honest about whether speed is your actual constraint. A faster server does not create demand. If your problem is that too few people arrive, hosting is not the lever, and the conversion math that decides whether ads make sense is a more useful place to spend an hour.
The same discipline applies to how you review performance month to month. Server metrics belong in an operations conversation, not a marketing one, and what a monthly local SEO report should actually contain is a short list that does not include uptime graphs.
08What to do this week
Find out where your site is hosted and who pays for it. If you cannot answer both, that is the week's task.
Put the billing on the business card and set a renewal reminder in your own calendar, along with the domain expiration date.
Turn on a free uptime monitor. It takes five minutes and it means you find out before your customers do.
Confirm your email is not dependent on your web host. If it is, plan to separate them, ideally before your next site change rather than during it.
Then test the site on cellular, away from the office, and note the result. Keep that number. It is part of the baseline you will want when the site changes, and the 90 days after a website launches go much better when you can compare against something real.
Be honest with yourself
When you do not need this
If you are on a hosted builder, you do not have hosting decisions to make. The platform is the host. Confirm the account is in your name and move on.
If your site works, loads quickly, and you can log into the account, do not switch hosts. Migration carries risk and delivers nothing visible to customers. Faster hosting pitches sell to businesses whose real problem is a page full of enormous images.
And if your website produces no leads at all, hosting is not your project. A faster empty room is still an empty room, and the money is better spent on the reason nobody is arriving.
Sources
- Google, "... new industry benchmarks for mobile page speed," Daniel An. Includes the Google Research figure sampling 11 million mobile web domains on a representative 4G connection, January 2018: 79% over 1MB, 53% over 2MB, 23% over 4MB, 25% able to save 250KB. Primary, and dated.
- Google, Core Web Vitals thresholds, web.dev. LCP, INP and CLS thresholds, all assessed at the 75th percentile. Primary documentation.
- Google, Deloitte Digital and Fifty-Five, "Milliseconds Make Millions". 37 brands, roughly 30 million sessions, October and November 2019. The 8.4% and 8.3% figures follow a 0.1 second mobile improvement against each site's own baseline.
- Google, PageSpeed Insights. Where to read your own field data against the thresholds above rather than a single score.
Related reading
- Website maintenance: what it covers and what it should cost. The recurring line hosting is usually bundled into, unpicked.
- Owning your domain, your site, and your data. The access criterion above, applied to every other asset in the project.
- Website builders versus a custom build: an honest comparison. If you are on a builder, this is the decision that replaces the hosting decision.
- Page speed and conversion: separating the myth from the measurement. Read this before you buy a faster plan to fix a heavy page.
Questions about your hosting?
Email me at eric@seod.com with the name of your host and roughly what you pay a month. I will tell you whether that is a reasonable arrangement for your size, whether you are paying an agency markup on a plan you could hold yourself, and whether anything about it would make a future move painful. Short answer, no meeting.
More on the practical side of projects in the web design and branding library.