We Measured 21 Dartmouth Restaurants Going Online. Four of Them Have No Website at All.
Dartmouth restaurants have the hard part figured out. The kitchen works. We measured what happens when a hungry person on a phone tries to find them: median mobile LCP 2.8 seconds, 10 of 17 with no contact form, 7 with no way to order or book without calling. Here is the data, the method, and the fix you can do yourself tonight.
A restaurant that has survived thirty years in Dartmouth has already won the hard argument.
It survived the recession. It survived the pandemic. It survived the year the rent went up and the year the fryer died in August. Somebody's mother built the recipe, somebody's son runs the register now, and on a Friday in Padanaram the dining room is full of people who have been coming since before the kids were born. That is not luck. That is thirty years of getting the food right on a margin nobody outside the business would accept.
None of that is the problem. The problem is much smaller and much stupider than that, and it happens entirely in the ninety seconds before a stranger decides where to eat.
What we measured, and what we could not#
On 18 September 2026 we audited 21 independent restaurants in Dartmouth, Massachusetts, across North Dartmouth, South Dartmouth and Padanaram. National chains excluded. For each one we checked whether an official website exists, then ran the site through Google PageSpeed Insights on mobile and crawled it for the things a customer actually needs: a phone number you can tap, a way to reach the place when it is closed, a way to order or book, and the structured data that tells Google your hours.
Two honest disclosures before any number.
First, the Google Places API refused our key on the day of the run. So this sample was assembled by hand from public directory listings rather than from Google's own result order, and we measured no star ratings, no review counts, no search ranks and no listed phone numbers. Any article telling you what these restaurants' ratings look like is not this one. We did not measure it, so we will not print it.
Second, of the 17 live sites, two would not return a speed measurement at all. They are excluded from every speed figure below rather than counted against them. A site we could not read is not a slow site. It is an unmeasured one.
Everything else here, you can re-run yourself in about a minute per site at pagespeed.web.dev.
The numbers#
LCP is Largest Contentful Paint: the moment the biggest piece of content on the page actually renders. It is the closest single number to "when can a person read this," which is why Google uses it.
For 10 of the 15 sites we could measure, that figure is field data from Google's Chrome User Experience Report. It is not a simulation. It is what real people on real phones experienced over the trailing 28 days.
Nine of those 15 fail Google's "good" threshold. Four are rated "poor," which Google sets at over four seconds. The slowest figure we report is twenty seconds.
One caveat on that last number, because it cuts against us: where Google has field data we report the field data, which is almost always the kinder number. Two sites in this sample post lab estimates above 24 seconds and are reported here at their field values of 2.4s and 2.1s instead. We are showing the generous figure, not the dramatic one.
Then the things that have nothing to do with speed:
| Measured on the 17 live sites | Count | What it costs |
|---|---|---|
| No contact form of any kind | 10 of 17 | Nobody can reach you after close |
| No LocalBusiness schema | 8 of 17 | Google guesses your hours and address |
| No way to order or book online | 7 of 17 | Every order is a phone call, during service |
| No analytics installed | 6 of 17 | No way to know any of this is happening |
No tappable tel: link | 5 of 17 | On three, no phone number on the page at all |
| No meta description | 4 of 17 | Google writes your search snippet for you |
Median performance score: 55 out of 100. Median accessibility: 90 out of 100, which is the one genuinely good number in this audit and worth saying out loud.
The line that matters#
Site G renders in 1.3 seconds on real visitors' phones. Excellent. It also has no contact form, no phone number we could find on the page, no schema and no analytics. It is a beautiful, fast, silent building with the door locked.
At the other end, Site L comes in at 20.0 seconds, Site I at 12.5 and Site O at 14.4. All three of those are lab estimates, not field data, because Google does not yet have enough real traffic on those sites to report on them, so treat them as a simulator's verdict rather than a stopwatch. Even discounted, they are not close. And here is the part that surprised us: it is not the photographs. Of the 15 sites we could measure, not one was flagged by Lighthouse for unoptimised or oversized images. Fourteen of the 15 were flagged for unused JavaScript. The median page in this sample weighs 2,716 KiB. The heaviest weighs 6,119 KiB. That weight is overwhelmingly code the page downloads and never runs: plugins, trackers, builder scripts, a booking widget loaded on every page including the ones with nothing to book.
Which is worth saying plainly, because it inverts the usual advice. Dartmouth restaurants are not slow because somebody uploaded a big photo of the scallops. They are slow because a website builder, or a decade of accumulated plugins, is shipping a few megabytes of JavaScript to a person standing on a sidewalk deciding where to eat.
And four restaurants in this sample have no official website we could find. What they use instead, we did not measure and will not guess at. What we can say is what the absence costs: with no domain of your own, whatever platform you are on owns the customer, owns the pricing, and owns the phone number, and you are renting a storefront you can be evicted from by a policy change you will read about after it happens.
Going without a site is not a moral failing. It is the rational choice when a social page is free and a website quote is four figures. It is just a choice with a landlord attached.
A website is often the first entry point for a business with its customers, and you want it to be a great experience for them.
That is the whole argument. Thirty years of getting the food right builds a reputation that lives in people's mouths. The website is where a stranger meets that reputation for the first time, and right now for most of Dartmouth that first meeting is a spinner.
The fix you can do tonight, for nothing#
Here is the free one. It takes about four minutes and it is the highest-return single change on this entire list.
Make your phone number tappable. Five of the 17 sites have no tel: link anywhere. On two of those, the number is sitting right there on the page as plain text, and on a phone plain text is not a link: the customer has to memorise it, leave your site, open the dialer and type it in, and a meaningful share of them simply do not. On the other three we could not find a phone number on the page at all, which is the same problem with the volume turned up.
Whoever edits your site needs to change this:
774-555-0100
into this:
<a href="tel:+17745550100">774-555-0100</a>
That is it. Squarespace, Wix, WordPress and Toast all let you do this from the page editor by selecting the number and inserting a link with tel: in front of the digits. Put it in the header, not just the footer. Use your real number, with the +1 and no dashes inside the href.
If your number is already tappable, do this one instead: open your own site on your own phone, on cellular data, not the wifi in the dining room. Count the seconds until you can read the menu. That number is what your next new customer experiences, and it is almost never the number the owner expects.
What we would do next, cheapest first#
Free, tonight. Tappable phone number. Claim and complete your Google Business Profile with correct hours, because Google's own documentation is explicit that complete and accurate information makes a business more likely to appear in local results. Four of these restaurants' entire fix is "have a website that exists" and that is a bigger conversation, but completing the Google profile is free and it is the same afternoon.
An afternoon, under $100. Cut the code you are not using. Fourteen of the 15 measurable sites ship unused JavaScript; on one of them Lighthouse measured 897 KiB of it, costing an estimated 4.5 seconds. On WordPress that usually means deactivating plugins nobody has opened in three years. On Wix or Squarespace it means removing the apps and embedded widgets you trialled once. Then add a LocalBusiness schema block so Google stops guessing your hours: eight of the 17 sites have none.
A week, real money. A contact form and an online ordering or reservation path that you own, on your own domain, rather than a delivery app's. Ten of 17 have no form. Seven of 17 have no booking path at all, which means every single order and reservation has to arrive by phone, during service, when the person who can answer it is expediting.
The one nobody wants to hear. Six of 17 have no analytics installed. Those six have no way of knowing any of this is happening to them. That is usually the reason it still is.
Why anonymous#
Every restaurant in this audit is called Site A through Site U, and it will stay that way. None of them asked to be measured and none of them consented to be publicly graded, and publishing a named list of local businesses' failings to sell them a service is exactly the behaviour that has made every restaurant owner in Bristol County delete "FREE SEO AUDIT" emails without opening them.
The data is real, the method is above, and any owner who wants to know which letter is theirs can call and ask. We will tell them, for free, with no meeting.
Joshua Amado, Avalon Partner. Fall River, MA. 774.559.8992 · Joshua.Amado@AvalonPartner.com
If you run a restaurant in Dartmouth, Fall River, Westport or anywhere on the South Coast and you want to know what your own site actually measures, send me the URL. I will run it and send you the numbers whether or not you ever hire me.
Want help putting this into practice?
Avalon Partner helps Fall River and South Coast businesses fix the gaps that cost them leads. Call 774.559.8992 or email Joshua.Amado@AvalonPartner.com.
