A Dartmouth Florist's Own Website Links to a Page That Sells a Competitor's Flowers

Start with what the florists on this coast are unambiguously good at.

Across two audit runs we identified 19 independent retail florists in Fall River, Somerset, Swansea, Westport, Dartmouth, New Bedford and Warren. Between them they hold 1,882 Google reviews, and 17 of the 19 are rated 4.5 stars or better.

You cannot buy that. You cannot automate it. It is what happens when someone builds a sympathy arrangement at 6am for a family they know by name, for thirty years running.

So this is not an article about florists who are bad at their jobs. It is an article about what happens to that reputation on the last twenty feet of the journey, when a customer who has already decided to buy opens the website.

The one question#

Earlier autopsies on this site asked whether a local business can take a transaction online at all. This one assumes yes and asks the question that comes next:

When the customer clicks "Shop," whose store are they standing in?

On Wednesday, 2 September 2026, we fetched the homepage of every florist website in our two florist briefs, plus seven common store paths on each (/shop/, /store/, /order/, /products/, /flowers/, /cart/, /shop-online/). We read the returned HTML for the storefront platform's own fingerprints, for cart and product markup on the business's own domain, and for links pointing off the domain.

Ten sites answered. Five refused our checker or returned an error, and are recorded as not measured rather than counted against anyone; a bot-block is not a missing store. Four more of the nineteen have no website at all.

Six of ten can take an order. That is far ahead of the South Coast average: across every category Avalon has measured, 193 of 253 business websites can take no transaction of any kind. Florists are well ahead of their neighbours here, and they deserve to hear that.

Three of the ten run their storefront on a floral-industry platform rather than a general web platform. Now look at Site D and Site G.

What we found on the Dartmouth site#

Site D is a Dartmouth florist with a genuinely handsome website and a working cart. You can buy an arrangement there without phoning anyone.

The storefront is BloomNation's. The page assets load from the platform's asset host, and in the footer sits a link whose text is the phrase "florist in [their own town]". It points to the platform's own town directory page.

We fetched that directory page. It carries product links for Site D's arrangements and for Site G's, a florist in a neighbouring town, side by side, on a domain neither of them owns.

So the sequence runs like this. A Dartmouth customer searches for a Dartmouth florist. They find Site D. They land on the website Site D pays for. And sitting below the arrangements is a link that carries them to a marketplace where a competitor's bouquets have equal billing with their own.

Site G has the same arrangement in reverse, behind a link reading "visit our Marketplace."

Neither shop is doing anything wrong. Neither, almost certainly, knows the link is there. It came with the template.

Why this is worth an hour of your Monday#

There is an overheated version of this argument, and I want to avoid it. A rented storefront is not a scam. BloomNation, Floranext, Teleflora and the rest do real work: product photography, payment processing, delivery zones, wire orders from out of state. For a two-person shop, renting that is frequently the correct decision. Site J rents from Floranext, runs its own checkout on its own domain, and carries an A grade in our audit.

The problem is narrower and far more fixable than "stop renting." It is that a rented storefront is optimised for the platform's marketplace, not for your shop, and the default configuration quietly treats your website as an entrance to their directory rather than the other way round. You bought the reputation. They hold the option on it.

The same instinct shows up in a plainer form. Site A sells through its own WooCommerce checkout on its own domain, which is exactly right and further than most local businesses ever get. Then you read the filenames of its product photographs: they are wire-service catalogue names. Thirty years of local design work, sold with the same pictures sitting on every florist website in America.

This is the thing Greg Gifford has been saying about local websites for a decade:

He was describing pages with no content. The rented-catalogue version is the same failure in a better suit: a page full of content, none of it yours, none of it a customer could not get from six other shops running the same platform.

The free fix, and it takes about forty minutes#

If you run a florist on the South Coast, do this on Monday. It costs nothing.

1. Find the outbound link. Open your own website on your phone. Scroll to the very bottom. Read every link in the footer, and every link in the small print under your product grid. If one takes you to a marketplace, a directory, or a "flower delivery in [your town]" page on somebody else's domain, you have found it. Screenshot it.

2. Ask your platform to remove it. Email your storefront provider's support address and ask for the marketplace and directory links to come out of your site's footer. On most floral platforms this is a setting or a ten-minute support ticket, not a rebuild. Some plans require the link. If yours does, you now know that, which is more than you knew on Sunday, and you can price it into your renewal.

3. Photograph six of your own arrangements. A phone camera in window light on a Tuesday morning beats catalogue stock, because it is visibly the flowers a customer will actually receive from a shop on their own street. Replace your six best-selling catalogue images with yours. This is the highest-return free hour available to any florist on this list, because your own work is the only thing on that page a competitor cannot buy a licence to.

What we did not measure#

• Whether any of these checkouts actually completes. We read HTML. We did not place an order, enter a card, or test a delivery. Cart markup is evidence of a cart, not proof of a working one.
• Any load time or speed figure. This run measured storefront structure only. Nothing here is a performance claim.
• Five florist sites. Three returned HTTP 403 and two returned HTTP 404 to our checker. Their sites exist. We could not read them, so they appear nowhere in the counts above.
• Contract terms. We do not know what any of these shops pays, what their platform agreement requires, or whether the marketplace link is mandatory on their plan. We measured what is on the page.
• Fresh Dartmouth discovery. Google's Places API refused our key on the day of this run, so no new Dartmouth-only florist set could be pulled. Every business here comes from florist audits already on file from July and August 2026, which is why a Dartmouth shop appears alongside Fall River, Westport, New Bedford and Somerset rather than in a clean single-town sample. Two Westport, Connecticut businesses that the original search pulled in by name collision have been removed, as have two cannabis dispensaries, a fruit-basket franchise and a wholesale florist supply company that the category search returned as "florists."

The part that actually matters#

Site D has dozens of reviews from people in Dartmouth who trusted them with a funeral, a wedding, an apology. That trust is the only asset in this entire analysis that cannot be bought, rebuilt, or replatformed.

Everything else here is a settings page.

If you want a second pair of eyes on where your website is leaking customers, or you want to know what it would take to own your checkout outright, I am in Fall River and the first conversation is free.

774.559.8992 · Joshua.Amado@AvalonPartner.com

Method: homepages and seven common store paths fetched over HTTPS on 2 September 2026 with a standard desktop Chrome user agent, one request per URL, no JavaScript execution. Platform identification is by first-party asset host and plugin path in the returned HTML, which anyone can repeat with view-source on their own site. Review counts, star ratings and the 193-of-253 figure come from Avalon's July and August 2026 category audits; the speed and accessibility numbers in those audits are reproducible by pasting a business URL into pagespeed.web.dev, but the review counts come from the Google Places listing, not from PageSpeed.