Three departments, one bench.

A page is a shopfront — and the shopfront is judged before the bell above the door has finished ringing.

The research desk is blunt on the figures. A reader forms a first impression of a web page in roughly fifty milliseconds — Google's own research times the judgement that fast — and the patience that follows is thin. As a page's load time stretches from one second to three, Google found the likelihood of a visitor abandoning it climbs by about a third; past three seconds, more than half of mobile readers simply leave.

A good landing page, then, does three honest things: it loads before the reader's patience does; it says plainly what the business offers and what to do next; and it earns enough trust in those first seconds to keep the reader reading. The workshop builds to that standard — quick to load, clear in its single purpose, and made to convert the enquiry rather than merely decorate the address bar.

A handsome page that cannot be found, though, is a handbill left in a drawer. So every TinWorx site is built to be discovered: sound structure and the markup that search engines actually read, so the business surfaces when a customer goes looking. We build the SEO in — and we build for the newer readers too: the AI answer-engines now summarising the web. Following Google's guidance for AI-crawler readiness, we give a client's pages a fair chance of appearing not only in the old blue links but in the AI overviews that increasingly sit above them. Improving a business's findability, in both, is part of the commission — not an afterthought sold separately.

Figures: Google / SOASTA Research, 2017; first-impression timing per Google research.

The software trade is changing under everyone's feet — and it is changing in a direction that makes the custom platform more valuable, not less.

The tools have arrived that let almost anyone assemble almost anything without a day's formal training, and the serious software houses have noticed. The clever ones no longer try to be everything; they specialise in one service done supremely well and let others reach it through an interface. That, increasingly, is the shape of modern software: not one general machine, but many specialised ones, joined together.

Which is precisely why a platform cut to your own business is worth more now. To compete, a concern must specialise — and a general, off-the-shelf platform carries a growing business only so far. Sooner or later a competitor arrives wielding a platform stitched from specialised services that quicken its output and sharpen its return, and the generalist is left in the dust.

Consider the maker of hand-crafted ornaments, selling on a crowded marketplace among hundreds of rivals. To be seen she must market; the only marketing that truly tells is the honest word of past buyers; and to earn that she must turn out fine work every time. Yet every hour spent posting listings, chasing deliveries and tallying the books is an hour stolen from the very craft that earns the praise. A general platform will shoulder some of it — at the price of bending her business to suit the platform. A platform cut to her own pattern does the opposite: it scales as she scales and grows as she grows, leaving her hands free for the work only she can do.

The honest caveat, since the Herald does not deal in overselling: not every business needs this, and the workshop will say so plainly. That is what the Business Review is for — it works out whether a custom platform earns its keep, or whether a sharp website is all you require, and you keep the written answer either way.

Under the bonnet, platforms are built on Google Cloud services and Google Cloud Storage, with relational or non-relational databases chosen to fit the work rather than the fashion.

Some readers arrive not with a business in need of tools, but with an idea in need of building. The workshop takes those commissions too.

The trade calls it an MVP; the Herald calls it a first edition — the smallest version of the product that can be put before real readers to prove the idea has a public. The workshop takes a founder from the Business Review (what must this edition prove, and to whom?) through a working first printing, to launch — on the same Google Cloud presses as everything else leaving this bench, built to grow into the full run if the public says yes.

The terms a founder cares about, stated plainly: the code is yours, entirely — no licences back to the workshop, no hostage clauses. Your idea is kept in confidence, and the workshop will sign to that effect before a word of it is discussed. And you deal with the maker directly — the person across the table is the person at the bench. The honest counsel that governs the rest of the shop governs here too: the first edition is kept deliberately small, because the cheapest lesson a founder can buy is finding out early what the market actually wants.

A mobile application is, more often than not, simply a further department built atop the platform — the same engine, reached from the customer's coat pocket. Compiled from a single Flutter codebase and dispatched to both Android and iOS, it spares the business two separate engineering bills. The workshop ships Android first, with Apple to follow.

The Herald publishes how it charges, as an honest paper should. Every commission runs the same way: a paid Business Review first, then a build priced on the Blueprint it produces.