WeDispatch 1.7.0: the blog declares both its languages, and two new pieces
Every paired article now points to its counterpart in the other language, so search engines serve the right version to the right reader. Plus two new articles: evaluating a neural voice, and troubleshooting text-to-speech.
Our blog exists in French and in English, and the two versions of any given subject are two distinct pages. This release links them explicitly, and adds two articles where readers were searching and finding nothing.
The two versions of an article now recognise each other
A search engine has no way of guessing that two pages cover the same subject in two languages: it has to be told. Every paired article now carries the address of its counterpart, and the pages emit the expected tags from it, fifty pairs, a hundred articles.
The effect is straightforward: an English-speaking reader who lands on one of our subjects arrives on the English version, and a French-speaking reader on the French one. A link at the top of each article also switches between the two in one click.
A few articles deliberately have no counterpart. A piece on the French accessibility framework, or on US hyperlocal news, has no true equivalent in the other language, and declaring an approximate match would be less useful than declaring none. Four automated checks verify at every update that each reference exists, that it is reciprocal, and that no article is claimed twice.
The same logic now covers the pricing, comparison, documentation, blog and what's-new pages, which declare their counterpart in both directions.
Two new articles
How to evaluate a neural voice before you commit. Every voice sounds good for thirty seconds, on text chosen to flatter it. The piece sets out six tests to run on your own content instead (your proper nouns, your figures, ten minutes of listening rather than thirty seconds) and the three questions that decide actual use: cost per article rather than per million characters, turnaround time, and what becomes of your archive if a voice is withdrawn from a catalogue.
Text-to-speech not working on your site: the causes, in order. Three very different faults hide behind that one sentence, the missing player, the audio that will not start, the output that is wrong. The article separates them, ranks their causes by frequency, and gives the three-check method that rules out half the cases in two minutes.
Both are also available in French.
The rest
The automated suite goes from 86 to 87 offline checks. No functional change in the product: this release is about search visibility and content.
Give your articles a voice with WeDispatch
This blog is itself voiced by WeDispatch. Curious how it sounds on your content?
Book a demo