Houston's Courier has been moving freight and time-critical deliveries around the Gulf Coast for decades. Their website had not kept up. This is what we rebuilt, what we surfaced, what we added, and what is still on the bench.
Starting with the part customers see
The old site was nearly a decade old. It had a mobile layout — it just did not work properly, with errors throughout that got worse the smaller the screen got. Underneath that, the site was thin where it counted: very little SEO foundation and effectively no structured data, so search engines had to guess at what the business did, where it operated, and what services it offered.
The service list had also fallen behind. Not because the company changed direction — Houston's Courier does what it has always done — but because it kept adding capabilities and the website never caught up. Work they were actively selling simply was not on the site.
The new site is a ground-up replacement: same-day delivery leads, the catalog reflects everything the company actually offers in 2026, structured data describes the business properly, and the client login sits on the homepage instead of behind a hostname nobody could remember.
One service the site was hiding, and one that is coming
A courier already sends a bonded, uniformed, background-checked person to a customer's door with a package. That position turns out to be worth more than a single delivery.
- Mobile notary — already offered, never advertised. This is not a new service line; Houston's Courier has been doing it. It was nowhere on the old website. Documents that need a notary are frequently the same documents that need a courier, so one visit covers both — and the company was earning none of the search traffic for it, because as far as Google was concerned the service did not exist.
- Electronic ID verification for age-restricted deliveries (18+ and 21+) — built, not yet launched. Alcohol, tobacco, vape, and certain pharmacy items carry a legal obligation to verify age at handoff. This is a genuinely new capability rather than an unadvertised one, and it ships with the new dispatch system described below — it is not orderable today.
Those two are worth separating clearly, because they are different kinds of work. One was a marketing failure with an easy fix — the service existed, the website just never said so. The other required building something that did not exist, and it launches when the new system does.
The driver portal, where the compliance actually happens
"We verify ID" is a claim. The driver portal is the evidence. Drivers work from a portal built for a phone in one hand and a package in the other:
- Barcode scanning off the air bill. The driver scans rather than types. That kills transposed tracking numbers, which are the single most common cause of a package that is physically fine and administratively lost.
- Photo delivery confirmation. A timestamped photo at the drop point, attached to the stop. "It was left at the side door" stops being a memory and becomes a record.
- Electronic ID verification at handoff, over a secure network. Launching with the new system: the age check is performed and captured electronically at the point of delivery and transmitted over an encrypted connection — not written on a signature sheet, not photographed and emailed, and not reconstructed from memory weeks later when someone asks for proof.
The through-line is that each of these replaces something a driver used to remember with something the system holds, securely. That matters most on the day a client disputes a delivery: the answer is a record retrieved in seconds, rather than a phone call to a driver who has made two hundred stops since.
Dispatch: replacing the system, not skinning it
The largest piece of this engagement is roughly two-thirds built, and it is worth being precise about what it replaces.
Dispatch, driver management, and the customer portal all lived inside one integrated legacy application on manage.houstonscourier.com. That coupling is the problem. Every change to how dispatchers work risked the customer-facing side, and every change a customer noticed had to be regression-tested against dispatch. Nothing could be improved in isolation.
The replacement is a standalone system with three separate surfaces:
- Admin and dispatcher — order intake, assignment, and the live board dispatchers work from all day.
- Driver — the portal above: scanning, photo confirmation, electronic ID capture.
- Client — and this one goes into the new website rather than beside it. Customers place and track orders on the same domain they landed on, not on a subdomain that looks and behaves like different software, because it was.
Three things ship with it:
- A suite of new security controls across all three surfaces. A system holding delivery addresses, signed proof-of-delivery, and age-verification records is holding sensitive data, and the legacy application was not designed with that in mind.
- A far easier experience end to end — from submitting an order to paying the invoice. The old flow worked if you already knew it. The new one does not require that.
- Continuous route optimization using the current Google Maps platform APIs, re-optimizing as the day changes rather than planning once in the morning and hoping. Traffic in Houston does not hold still, and neither should the route.
Splitting one application into three surfaces on a shared core costs more up front than restyling the old one. It buys the ability to change a dispatcher's workflow on a Tuesday without touching anything a customer can see.
No AI in the workflow. That is a decision, not an omission.
It is worth stating plainly, because in 2026 the default assumption runs the other way: no AI touches the day-to-day operation of Houston's Courier. Every delivery ticket, every email, every phone call, and every delivery is handled by a real person.
The software's job here is to remove clerical work and to remember things accurately — scan instead of type, capture instead of recall, re-optimize a route as conditions change. The judgment calls stay with the dispatchers and drivers who know the accounts, the buildings, and which receiving dock actually opens at 7am.
For a same-day courier that is not nostalgia, it is the product. When a delivery is time-critical and something goes wrong, the customer wants a person who can fix it — not a queue.
What the search numbers did
A rebuild is a genuine risk to organic search. New URLs, new structure, and a decade of accumulated signals to carry across — sites lose traffic doing this. Six months on, compared against the six months before:
- Impressions: 153K → 273K, up 78%
- Clicks: 3.05K → 3.84K, up 26%
Worth reading those two together rather than separately. Impressions grew faster than clicks, which means the site now surfaces for a far wider set of queries — including broader, higher-funnel searches that were previously out of reach entirely and that convert at a lower rate by nature. Average click-through moved from about 2.0% to about 1.4% as a result. That is the arithmetic of a site appearing in a great many more places, not a site getting worse at earning the click, and the number that pays the bills — actual visits from search — is up by roughly 790 a month.
A good share of that is the unglamorous work: real structured data, a service catalog that lists the services, and pages for offerings like mobile notary that were previously invisible to search entirely.
Still to come
The dispatch and portal build is about two-thirds done. The remaining third is the part that has to be right rather than fast: cutting live operations off the legacy system without dropping an order mid-flight, and moving the client portal into the new site so customers stop seeing two different pieces of software.
Age-restricted delivery goes live at the same time, with its own service page — there is no point advertising a capability before the system that performs it is carrying real traffic.
We will write that up when it ships, including whatever goes wrong on the way.
Houston's Courier is a Tuxxin LLC client. If you are running a business on software a decade past its service life — and you can usually tell, because everyone has a workaround for it — get in touch.