Winemag · Fine wine publishing · South Africa

South African wine criticism, rebuilt and run.

Winemag is where South African fine wine gets reviewed, rated and argued about. Its readers had moved to their phones, and the site was still built for a desk - so every slow load was a reader deciding to do something else before the first tasting note.

A mobile rebuild, and the revenue that had to survive it.

The winemag.co.za home page on a laptop: the masthead over a grid of featured wine reviews, each with a photograph, a category label and a standfirst.

Client: Winemag
Sector: Publishing · South African fine wine
Role: Mobile-first rebuild
Platform: WordPress
Engagement: Built, and still run by us
Expertise: Web Development, Interface Design, Cloud and DevOps Engineering

The need

An audience on a phone, and a site built for a desk.

Wine magazine ran in print from 1993 until 2011, and winemag.co.za is where it carried on - reviews, ratings on a 100-point scale, and the arguments that follow a score. It is a publication with genuine authority in its field and none of the budget that usually comes with that, which is a combination that makes every technical decision count twice.

Its readers had done what every publication's readers have done, and moved to the device in their pocket. The site had not gone with them. Pages were slow, a layout drawn for a desktop column had been squeezed rather than rethought, and getting from one review to the next was a small negotiation every time.

For a publication that is not a usability complaint, it is the business. Attention is the entire asset, and everything else - the sign-ups, the shop, the advertising - is downstream of somebody staying long enough to read.

The load
A reader who arrives from a search or a share and waits is a reader who has already been given a reason to leave, before a single word has been read.
The layout
A desktop column squeezed onto a phone does not just look tight. It flattens the hierarchy that tells a reader what a review is, what it scored and where it goes next.
The next one
Publishing lives on the second article. If getting to it is work, the session ends at one - and one article is a visit rather than a reader.

What we did

Rebuilt around the reading, then around what it carries.

The brief was not a mobile skin over the same site. A publication earns from attention and converts it three ways - registrations, commerce and advertising - and each of those either survives the reading experience or quietly erodes it. So the work ran on both halves at once: make the reading better, and make everything it carries less intrusive than what it replaced.

Mobile layouts rebuilt for reading - hierarchy, type and a scroll that does not fight the reader
Registration and gated content streamlined, so the cost of signing up drops
Email templates redesigned to surface more of the archive and carry an offer honestly
Subscription products and e-commerce integrated, so the shop is part of the publication rather than beside it
Advertising moved into defined, labelled slots in the flow - no overlays, no interruptions

The 100-point scale is the product

Winemag rates on the same 100-point convention the international critics use, and the bands carry real meaning - 96 and up is extraordinary, below 80 is a polite instruction to look elsewhere. A redesign that made the writing prettier and the scores harder to find would have improved the wrong thing.

The reading

A review should open, not load.

The article template is where the whole thing is decided, so it got the most attention: the masthead small and out of the way, the category and date where they can be scanned rather than read, the headline given room, and the byline and comment count sitting together because on this site the argument under a review is part of the review.

The end of an article matters as much as the start. Related reviews now arrive as real editorial - category, date and the opening lines - rather than as a row of thumbnails that ask a reader to guess.

The top of a Winemag review on a phone: masthead and menu, a breadcrumb, the headline, and the byline crediting Christian Eedes with the comment count beside it.

A phone showing the end of a Winemag article, where a Related Articles block offers the next two reviews with their category, date and opening lines.

The Winemag rating system page on a desktop, setting out the 100-point scale from 96-100 Extraordinary down to Below 80, with a labelled advertisement in the sidebar.

What it carries

Three revenue lines, none of them in the way.

An independent publication does not get to choose between selling things, taking advertising and asking for support. It does all three, and the reading experience either absorbs them or is damaged by them. Each one got built as part of the site rather than bolted to it.

The shop
Awards entries, reports and glassware, sold on the same site they are written about, with the product pages built to the same measure as the articles.
The email
Transactional mail redesigned as something worth opening - the masthead, a real image, and room to point a reader back at the archive instead of ending at a receipt.
The advertising
Labelled slots in the page flow, sized and placed on purpose. Visible enough to be worth selling, quiet enough that nobody reaches for the close button.

Three phones showing the Winemag shop: the product list sorted by best sellers, a glassware product page, and an awards entry with its add-to-cart button.

A Winemag product page on a phone, with the add-to-cart button above a labelled advertisement sitting in its own slot in the page flow.

A Winemag contribution receipt open on a tablet, the masthead over a wine photograph above the payment details.

Since then

The rebuild was the start of it.

We host and maintain the site, which for a publication this size is the part that actually changes the working week. An editorial team whose asset is judgement about wine should be spending its time on wine, and the surest way to stop that is to make them responsible for uptime, updates and the slow decay every content site drifts into between redesigns.

Speed is the clearest case. A load time is not a thing you fix once and file - it is a thing that gets worse every time somebody adds a plugin, a tracking script or a heavier image, unless someone owns it. Someone does.

An engineering team they do not employ
Hosting, maintenance and the feature work as it comes up, at the scale a title like this actually needs rather than the scale an agency would prefer to sell.
Performance as a standing job
The 32% was a result. Keeping it is a habit, and it belongs to whoever runs the site rather than to whoever last redesigned it.
The editorial team writes
Which is the entire point, and the only one of these three a reader ever notices.

By the numbers

What a faster read is worth.

-0.00%average page load time, which is the change everything else follows from

+0.00%mobile page views, on the surface the rebuild was actually for

+0.00%sessions across the whole site, desktop included

Three figures that are all the same figure read three ways: the site got quicker, so more of it got read, so more people came back. What they do not measure is the shop, the sign-ups or the advertising - those were built in the same pass, and their results are the client's to publish rather than ours to claim.

The word from the founder

“We were very lucky to discover Touchfoundry who offered both branding savvy and technical know-how. Front-end looks great because back-end is so easy to use.”