🕸️ Let’s think about Quality, for the web

or How I Learned to Stop Worrying and Love the Web

A couple of months ago, I was volunteering for Paris Web, the leading French conference on the subject of Design, Accessibility and Quality, for the Web. Most of the talks are in French, but we also invite a handful of English-speaking speakers every year, so as not to run around in circles in the francophone Web community. (The Call for Papers takes place in February-March, follow @ParisWeb / if you’d like to submit a talk for our 2019 edition next October.)

As I was welcoming a speaker from the United States, we discussed about Paris Web, the hows and the whys and our focus on advocating for the users through design, accessibility and quality. She asked me what quality meant for us, as it’s on our website, badges and stickers, and I got a little confused when I tried to explain, so I didn’t do a good job. I this article, I’ll try make you care about quality when you design, make, or just think of the Web. I’ll also talk about Opquast, its approach to quality and its Open Quality Standards.

What is quality?

Let’s look at what Wikipedia says about quality:

In business, engineering, and manufacturing, quality has a pragmatic interpretation as the non-inferiority or superiority of something; it’s also defined as being suitable for its intended purpose (fitness for purpose) while satisfying customer expectations. Quality is a perceptual, conditional, and somewhat subjective attribute and may be understood differently by different people.
Wikipedia: Quality (business)

The main takeaway here is that it is first defined by its antithesis: quality is really about non-quality ; in an industry, it is about all the rules and processes used to make a product not suck, so that the customer’s expectations are satisfied. For the web, is about the non-suckiness of the user’s experience.

Automobile makers or chemical engineers have very strict quality processes. You won’t drive a car that hasn’t passed through strict QA testings following common industry standards. Of course, some car makers spend more on quality and it shows—and in the end you get what you pay for.

Automobile makers have had a century or more to get together and—sometimes through legislation and sometimes through good ole business sense—create quality standards, that would make the use of cars better for their customers, and better for their non-customers. ABS brakes can protect the driver of a car, but also pedestrians.

The Web is a young industry. The oldest among us remember the Wild West era of web making, when we wrangled <marquee> and <blink> tags inside a <frameset> or <table>, when we didn’t care if one form featured a big blue button, while the next had but a small gray default one ; a nightmare for accessibility and UX, assuredly (but we had fun, making bad HTML was the mutant spider byte that got us weaving the WWW). Rest assured, while we now have standards for markup and every other technical aspect of web making is documented, and every other design blog has its own UX or SEO ebook to sell, we still don’t have industry-wide Best Practices.

How to define Quality for the Web?

First off, the end of the quote, Quality is a […] somewhat subjective attribute and may be understood differently by different people, shows the crux of the problem.

If I were to ask you, or a colleague of yours, or my niece, or one of the billions of people who access the Web from a slow mobile phone over slow networks, what quality entails when applied to a website, the answers would all be different. Someone would want everything to be legible, someone would talk about colours or how she can’t decide if the website is trustworthy, or how the videos don’t have transcriptions. They all have different expectations, implicitly or not. These expectations are based on our culture, experiences and digital literacy, so finding universal guidelines to follow is going to be hard.

How can we translate an expectation to a rule or a recommendation? How can we more precise than “a website should be ergonomic and intuitive”? These mentions are not as useful as they seem, because they ask more than they answer, each of us will have a different opinion as to what makes a website or an app ergonomic or intuitive. And the quality of the end result will be very subjective. Generic demands have to be replaced by individual rules ; looking for best practices is mainly about asking what quantifiable expectations lay implicitly behind an inexact demand.

Finding common best practices (in which I talk about Opquast)

A best practice is hard to find.

A best practice should be useful. Enforcing useless practices is a waste of time, ours and the user’s.
It should also be universal, as in not rooted in any culture or law ; RGPD Compliance is a good thing in the EU, but it’s not universal, so it’s not a best practice (caring about collecting as little data as possible is a good idea though, respect your users).
A best practice must be realistic. Everything best practice should be feasible, possible to implement.
A best practice should be consensual among web professionals and users ; I dislike web ads and think it’d be a good idea to not have ads and tracking on every other website, but some people rely on them to make their website worth it financially, so eliminating ads cannot be a best practice (caring about not bludgeoning your user with ads is a good idea though, respect your users).
And last of all, a best practice has to be checkable. How can you do QA if you don’t have clear rules against which to test your website?

The best way to go about discovering best practices would be to gather many web professionals and users, and to ask them what quality means for them as a user, or as a specialist in their field. Such rules would then be debated and edited to fit the five criteria

That’s were I get to tell you about Opquast. Opquast is a project led by a French team, whose goal is to push for better quality in online services, through the use of checklists of best practices (distributed under Creative Commons BY-SA licences), and other QA tools. The work started in 2000, and gave way to 3 versions of their Open Quality Standard checklist with the help of dozens of workshops with Web industry professionals, and thousands of messages on forums and blogs dedicated to Quality and UX.

Opquast V3 gathers 226 best practices which are all useful, universal, realistic, consensual and checkable, as well as 70 recommendations (which can be les consensual or universal). Here are some best practices, as an example:

#39 - A product or service can be purchased without creating an account.
  • To allow buyers to place an order immediately.
  • To remove the barrier of account creation.
  • To increase the conversion rate.
#82 - Sound and video are launched by the user.
  • To give the user control over the visual and auditory interface when visiting a website.
  • To not surprise the user by unexpectedly playing audio content.
  • To not impose the playback of animated content on the user.
#32 - Each page provides a title that enables one to identify the site
  • To allow users to immediately identify the website in tabs, favorites, windows, browser histories, and screen readers.
  • To improve the site’s SEO and its presentation in search engine results.

Their work on Web Quality is being recognised in the French Web industry, and they offer certifications (website compliance or personal expert diplomas) and training. Many web schools include quality in their curriculum and their students can now be Opquast certified when they graduate, and a lot of Web agencies or development studios care about hiring candidates that are Opquast trained or certified. Hey, I even just got a job because I was recently certified (graded at 945/1000, with the level Expert), and my new company was lacking someone who could bring processes and methodology to ameliorate the quality of their online services.

Most of the project is available in English: the checklist, as well as the certification guide, training and exam.

Why should we care?

As web professionals, we make things that can deliver information at the speed of light (or at least at the speed of radio waves or electricity in copper), everywhere around the globe (where there’s connectivity) to the whole world population (if they can read our language). A great man said that great powers come with great responsibilities—and it’s especially true for us Web weavers. Our responsibility to the world’s people is to make their Web experience less sucky. We can do that through a common understanding of quality.

Designing, developing or leading a web project (or an app) have inherent constraints, and adding to them might take time and resources you do not have. On the other hand, avoiding non-quality can bring you much in lost revenue through better position on search engines, better compatibility, speed and accessibility leading to a lesser quit rate, or better word-of-mouth following great after-sale service.

Caring about quality makes you a better Web maker, because it makes you care about your users. Building them a less sucky experience shows respect and creates trust, which is hard to earn but quick to be lost. You should care because you’re human, and you’re making the Web for your fellow humans.

Pushing Quality around the globe

Sadly, I see no such project or organisation pushing for Quality for Online Services on the English-speaking part of the Web.

One of the best things about Opquast is that the process and the result are open. Which means that one can discover the best practices and participate in their elaboration. This is a first step, which can be followed by an Opquast Certification. It might not be well recognised outside of France, but it’ll help you create better online experiences.

The next step is to make your peers and colleagues care about quality. Not in terms of a blurry ideal changing from person to person, but in terms of useful, checkable, consensual, realistic and universal guidelines, that can be activated to achieve a better satisfaction of your user’s implicit and explicit expectations.

Once we as an industry start thinking like that, we’ll be closer than we’ve ever been to a universal WorldWide Web, fit for human use. And that is a the best goal I can think of. Let’s make the Web less sucky, one good practice at a time.


I have just been certified by Opquast as an Expert in Web Quality, I paid for the certification myself as a regular client. Opquast has been a sponsor of Paris Web, where I volunteer, for a number of years. Other than that, there is no monetary link between Opquast and I.

On en discute ?…