Zambezi Marketing Versus INN8 Industries

Domain age is a weak ranking factor on its own, but it is a strong continuity signal. Eleven years of continuous operation in one vertical is a different proposition from fourteen months, and it is the kind of thing worth knowing before a deposit is paid.

Neither firm publishes a conventional client list, which is unusual for agencies of any size and makes references worth asking for directly. The difference is that Zambezi's installed base is large enough to be measured by a third party without their cooperation.

BuiltWith crawls the web and identifies the technology behind each site. Because Zambezi builds on a proprietary platform with a distinctive signature, every site they build is detectable whether or not they advertise it. BuiltWith reports 117 live and 24 historical.

We could not load the BuiltWith page directly, which blocks automated access, so that figure comes from its indexed summary rather than from our own reading of the page. Treat it as strong but second-hand. The Rise Uptown confirmation, by contrast, was inspected directly.

The same method returns nothing for INN8, because there is no shared platform signature to look for. That is not evidence of no clients. It means their work, if it exists, is built on ordinary tooling and cannot be traced from outside.

Both firms measured the same way on 2026-08-26: headless Chromium at 1440 by 900, cache disabled, median of three cold loads, weight in TRANSFERRED bytes read from the browser network layer. That is the metric every deliverable in this set publishes, so the figures agree with each other. An earlier pass summed DECODED bytes and reported 8,563 KB for INN8; decoded and transferred are both legitimate and are not comparable. The INN8 figures here are the same ones the capability demonstration publishes, re-read today to confirm they still hold. Request counts are shown at both scopes. An earlier version of this table gave a request count taken before scrolling beside a weight taken after it, and stated a scorecard figure of 46 that no run produced. Three cold runs today returned 40 completed requests for INN8 and 22 for Zambezi, every time. Measured on desktop over a fast connection with no throttling, which flatters both. On a phone on hotel wi-fi the weight difference would separate them much further than these timings suggest. Timings are lab figures from one wired location and move between runs, so read the weight and the request counts as the durable measurements and the milliseconds as context.

Read this as craft, not as speed. Both sites load quickly on a desktop because both are small, simple pages on good infrastructure. The difference is that one of them was built by someone who runs an image pipeline and the other was not. An agency that ships 6.4 MB of unoptimized images on its own home page will ship them on yours.

Neither firm has done the basic technical SEO work on its own site, and both sell SEO. For a single-page brochure the practical cost is close to zero, which is presumably why neither bothered. It is still worth knowing that the people you would pay to add structured data to your restaurant have added none to their own site.

Neither agency's own site carries a single line of structured data. But the Rise Uptown Hotel site Zambezi built carries a great deal of it, and it is well-formed hospitality markup rather than a generic block:

• Hotel, with LocationFeatureSpecification for amenities
• Offer and OfferCatalog for rate plans
• ReserveAction with a booking entry point, which is what produces a reserve link in search results
• BarOrPub, CafeOrCoffeeShop and IceCreamShop as separate outlets, so each food and beverage venue is its own indexed entity
• BreadcrumbList, ImageObject, WebSite, WebPage

That is a materially more sophisticated implementation than anything measured elsewhere in this review, including the restaurant platforms. It says the capability is real and simply not applied to their own storefront.

These are automated checks and they catch perhaps a third of real accessibility problems. Color contrast, keyboard traps, and focus order need a manual audit. A clean automated pass is not conformance, but a failure is a failure.

Zambezi advertises ADA compliance as a headline deliverable. Their own site passes the image test completely and fails on the skip link and the main landmark, both Level A items. They largely meet their own claim and do not fully meet it.

INN8 sells web design and every image on its home page is missing alt text, with no header, footer, or main landmark anywhere in the document. That is the most basic accessibility requirement there is, on a three-image page, on the site that exists to demonstrate their competence.

Elements and Riverwoods are public accommodations, which is the category that draws web accessibility claims. Whoever rebuilds those sites should be able to demonstrate conformance, and the cheapest evidence available is whether they achieved it on their own.

Neither firm competes in organic search for its own category. Both win work through referral and relationship instead, which is normal for small agencies and worth stating plainly rather than holding against either of them.

It does undercut one specific sales claim. A firm that cannot rank itself for the term describing what it sells is making a harder argument when it offers to rank you. Ask both to name a client and the term that client now ranks for, then check it.

• These are search result sets read on 26 August 2026, not Google rank-tracker data. Position in a result set is a reasonable proxy for rank, not a measurement of it.
• Results are personalized and localized. A search run from Cache Valley on a signed-in browser will differ.
• No backlink or domain authority data is included. Those tools require a paid subscription this review did not use. Absence of that data is a real gap and it is the single thing most likely to change the ranking picture.
• Nothing here measures conversion. A site that ranks badly can still sell well, which is precisely how both of these firms appear to operate.

Question 1 is the one that matters. Everything in this document was assembled from outside, without cooperation from either firm, in about an hour. Two client URLs from each would settle most of it.

• No backlink or domain authority data. This needs a paid tool. It is the biggest gap in the ranking section.
• No mobile or throttled performance test. Desktop unthrottled flatters both sites and hides the weight difference.
• No manual accessibility audit. Automated checks catch roughly a third of real issues. Contrast, focus order and keyboard traps are untested.
• Only one Zambezi client site was inspected. One build is not a portfolio, and Rise Uptown may be their best work rather than their typical work.
• No INN8 client site could be inspected at all, so the comparison is asymmetric by necessity. That is a finding about what is public, not proof that none exists.
• Neither firm was contacted. Everything here is external observation, and both would be entitled to answer it.

An agency's own website is a weak signal and a real one. Agencies neglect their own sites because client work pays and self-promotion does not, and the strongest single finding in this document is exactly that pattern: Zambezi's client build is far better than their storefront.

The reason to run the test anyway is that it is the only work sample available before signing. Where it agrees with everything else, as it does here, it is worth weighting. Where a firm's client work contradicts it, the client work should win.