Every agency pitch in the Okanagan has the same slide in it. A red PageSpeed score, a scary number of seconds, and the line that Google punishes slow websites. We wanted to know two things. How slow are local trades websites, really? And does being fast actually get you into the top three?
So we tested all of them. Not a sample. Every contractor website that showed up when we searched 16 trades in Vernon, Kelowna, West Kelowna and Penticton. The short version: 95 percent of them miss Google's own speed targets on a phone, and the businesses sitting in the top three were no faster than the ones below them. Both of those things are true, and you need both to decide what to spend money on.
One disclosure before the numbers. Osprey builds websites, including in Vernon and Kelowna, so we have a stake in you caring about this. That is exactly why we are going to tell you where speed does not matter as much as the pitch says.
We Tested Every Trades Site That Came Up in Okanagan Local Search
On September 18, 2026 we ran 16 trade searches (roofing, plumbing, electrical, heating and cooling, landscaping, concrete, gutters, tree service and so on) in each of four towns through Google's Places search. That returned 978 result slots. We stripped out suppliers and retail stores, which left 613 local service businesses. 77 of them had no website on their Google profile at all. The rest pointed to 508 unique websites.
We ran every one of those through Lighthouse, which is the engine behind Google's PageSpeed Insights tool, on its standard mobile setting: a mid-range phone on a slow 4G connection. 487 sites returned a full result. 19 did not (mostly certificate errors that stopped Chrome from loading the page), and two were Facebook or Google Sites pages we left out.
Three honest caveats, because they matter:
- These are lab numbers, not your visitors. The lab test is deliberately harsh. Google's ranking systems use field data from real Chrome users, collected over 28 days, and a site only gets field data if it is "sufficiently popular". A lot of small business sites never cross that line. So do not read "8.8 seconds" as "Google thinks your site takes 8.8 seconds".
- We tested the link on the Google profile, because that is what a searcher taps. On 279 of the 487 sites that link redirected first (usually from http to https, or to add www). We kept that redirect in the test on purpose, because your customer sits through it too.
- One run per site. Lighthouse wobbles a few points run to run, so we only report medians and shares, never a single business. We are not naming anyone.
95 Percent Fail Google's Mobile Speed Targets, and It Is Almost Always the Loading
Google publishes three Core Web Vitals: how fast the main content appears (Largest Contentful Paint, "good" is 2.5 seconds or less), how much the page jumps around while loading (Cumulative Layout Shift, 0.1 or less), and how quickly it reacts to a tap (Interaction to Next Paint, under 200 milliseconds). Tap response cannot be measured in a lab, so Lighthouse uses Total Blocking Time as its stand-in. Here is how 487 Okanagan trades sites did.
| What we measured (487 sites, mobile) | Google's "good" line | Okanagan median | Sites that hit it |
|---|---|---|---|
| Largest Contentful Paint (main content visible) | 2.5 seconds or less | 8.8 seconds | 29 of 487 (6.0%) |
| Cumulative Layout Shift (page jumping around) | 0.1 or less | 0.005 | 84.2% |
| Total Blocking Time (lab stand-in for tap response) | 200 ms or less | 36 ms | 89.1% |
| All three at once | 24 of 487 (4.9%) | ||
| Lighthouse performance score | 90 or higher | 64 | 39 of 487 (8.0%) |
Only 24 of 487 sites passed all three lab targets. That is 4.9 percent. But look at where they failed. Layout shift and tap response are fine on most sites. The problem is almost entirely one number: how long it takes for the main thing on the page to show up. The median site needed 8.8 seconds on the test phone. 216 sites, 44.4 percent, took longer than 10 seconds.
That is good news, oddly. Loading problems are the easiest kind to fix, and they usually come from the same three or four causes, which we get to below.
If you read another agency's page that says most template sites score under 50, that is not what we found here. Only 42 of 487 sites (8.6 percent) scored under 50. The typical Okanagan trades site scores in the 60s. It is not a disaster. It is just slow.
The Top Three Were No Faster Than Everyone Else
Here is the part the agency slide leaves out. For each search, we compared the sites in the top three against the sites ranked below them in the same list, same trade, same town. Of the 52 searches with enough scored sites on both sides to compare, the top three were faster in 24 and slower in 27, with one tie. That is a coin flip.
- The number one result was the fastest site in its own list in 4 of 53 searches. With about 14 sites per list, pure chance would give you about 4.
- Of 152 top-three slots with a scored website, 133 belonged to a site that took over 4 seconds to show its main content. Only 6 were under Google's 2.5 second line.
- The rank correlation between position and load time across all sites was -0.06. Effectively zero.
This is one snapshot of one set of searches, and we did not control for distance, reviews or how well each Google profile matches the search. So it does not prove speed never matters. What it shows is that in the Okanagan right now, speed is not what separates the top three from everyone else. That fits what Google says itself on its page experience documentation: "Google Search always seeks to show the most relevant content, even if the page experience is sub-par."
If you want to rank in the map, the levers are your Google profile, your reviews and your location pages. We measured the review side of that in how many Google reviews you actually need in Vernon or Kelowna, and our local SEO work starts there, not with a speed score.
Want to Know What Is Actually Holding Your Ranking Back?
Book a free strategy call. We will pull your trade's top three in your town, compare your Google profile, reviews and site against them, and tell you which one is the real gap. If it is not your site speed, we will say so.
→ Book Free Strategy CallSpeed Is a Conversion Problem: Fix It for the Homeowner on a Phone
So why fix it at all? Because ranking gets someone to tap your link, and then your site has to do the rest. The person tapping is often standing in a flooded basement or looking up at a dripping soffit, on a phone, on whatever signal they have. Eight or nine seconds of white screen is a long time to wait when the next result is one thumb away.
We do not have a contractor-specific number for how many of those people leave, and we are not going to invent one. The most careful study we know of is Milliseconds Make Millions, run by Deloitte and 55 and commissioned by Google, across 37 large brands and over 30 million mobile sessions. For lead generation sites, a 0.1 second improvement across four speed measures went with 21.6 percent more visitors reaching the form submission page. Those were big national brands, not local trades, and Google paid for the study. Treat it as a direction, not a promise. But the direction is clear.
Where you build matters less than how, but it shows up in the medians:
| Platform | Sites | Median score | Median main-content load | Under 2.5 s | Median page size |
|---|---|---|---|---|---|
| GoDaddy Website Builder | 34 | 80 | 4.7 s | 5.9% | 1.01 MB |
| Custom or unidentified | 83 | 69 | 6.3 s | 9.6% | 2.06 MB |
| Wix | 87 | 71 | 8.6 s | 9.2% | 2.28 MB |
| WordPress | 235 | 60 | 10.4 s | 4.3% | 2.91 MB |
| Squarespace | 23 | 55 | 11.9 s | 0.0% | 2.35 MB |
Platforms with fewer than 10 sites in our set are not shown. The GoDaddy builder pages are fast mostly because they are small, about a megabyte. WordPress is the widest spread of all: some of the fastest sites in the valley and a lot of the slowest, depending on the theme and plugins. Towns barely differed. Vernon's median load was 8.4 seconds, Kelowna and West Kelowna 9.2, Penticton 9.3.
The Four Fixes Worth Paying For, in Order
Lighthouse also tells you why a page is slow. For 296 of the sites (everything we tested after we switched on its detailed breakdown), we pulled those reasons. Four causes cover most of the gap.
1. Fix the link on your Google profile (free, five minutes)
57 percent of sites (279 of 487) redirected before loading, and 227 of those were because the Google profile still links to an old http:// address. In the lab each redirect cost a median of 835 milliseconds. Redirected sites had a median load of 9.2 seconds against 8.2 for the rest. Open your site in a browser, copy the address it ends up on, and paste that exact address into the website field of your Google Business Profile. This is the one fix on this list that costs nothing.
2. Shrink the hero photo and let the browser find it early
On 242 of the 296 sites (81.8 percent) the slowest thing on the page was an image, usually the big photo at the top. Of those, 89 (36.8 percent) were hidden from the browser until a script or stylesheet loaded, often because they sat in a slider or a CSS background, and 172 (71.1 percent) were not marked as a priority. On about a quarter of all sites, Lighthouse estimated that better image delivery alone would save a second or more. One properly sized, compressed photo in a plain image tag beats a five-slide carousel every time.
3. Cut the scripts and plugins that block the first paint
The median site had about 850 milliseconds of stylesheets and scripts that stop anything from appearing until they finish. On WordPress this is usually a page builder, a slider plugin, a chat widget and three font families. Remove what you do not use and defer the rest.
4. Check your hosting
Most sites answered quickly, with a median server response of 113 milliseconds. But 17.5 percent took longer than 600 milliseconds just to start sending the page. If that is you, no amount of image work will save it. Cheap shared hosting is the usual cause.
What not to pay for: chasing a 100. Only 39 sites in the whole valley scored 90 or better, and they were not the ones on top. Getting your main content up in two to three seconds on a phone is the goal. Past that, spend the money on your Google profile and your reviews. If your site gets visitors but no calls, speed is only one of the reasons, and we covered the others in why your contractor website gets traffic but no leads.
Run the Same Test on Your Own Site in Five Minutes
- Go to PageSpeed Insights and paste the exact link from your Google Business Profile, not the address you type from memory.
- Leave it on Mobile and run it.
- Look at the top section first. If it shows real-user data ("Discover what your real users are experiencing"), that is closest to what Google uses. If it says there is not enough data, your site does not have field data, and the lab result below is all you have to go on.
- In the lab section, ignore the big score and read Largest Contentful Paint. Under 2.5 seconds is good. Under 4 puts you ahead of about 85 percent of the trades sites we tested.
- Scroll to the diagnostics and look for the four causes above: redirects, the hero image, render-blocking resources and server response time.
If you want someone to look at it with you, our free SEO audit covers speed alongside the things that actually move local rankings. We work with trades in Vernon and Kelowna, and last month we also checked whether Okanagan trades sites even have a tap-to-call button. One in five did not, which is a bigger problem than any speed score.