Popular Misconceptions about Designing for the Web

I’ve been working on web projects for a few years now. Most of the things I’ve contributed to have been based, at least in part, around the principle of ‘one-URL-per-thing’. This, as should be clear from pretty much everything else on my blog, is something I consider to be a good thing. With a URL for every ‘thing’ that a user might be interested in (within reason, and within the scope of your content and data), you greatly increase the findability, sociability and utility of your content. But, time and again, the same criticism comes back – “ah, but not everyone is a geeky wanderer of websites, not everyone will want to use this; this won’t appeal to the mainstream, passive audience; it’s too much like a reference work; a boring user experience; certainly not delightful.” I’m afraid I have to disagree with pretty much each and every one of those criticisms – and now seems as good a time to do so as any. So here goes.

The false dichotomy

Firstly, and most importantly, there’s a huge misunderstanding going on here. It is not a straight choice between building something based around ‘one-URL-per-thing’ and building a ‘delightful experience for the majority of (apparently passive) users’. The two aims are not, and should not be presented as being, in competition with another. In fact, I’d argue that the ‘one-URL-per-thing’ approach is a massively helpful thing to build as a stepping stone to creating the ‘delightful experience’.

Sure, it will depend on the amount of resource and content you have – sometimes, it just won’t be practical to build a Web presence for every single piece of content (and yes, you can easily tie yourself up in knots trying to work out how granular you want to be with addressability) – but I’m sure it’s much easier, and much less expensive, than you’d think. Remember, in general, the more things you do make addressable, the more discoverable your content is, so it’s quite possible that the investment will be worth it.

It’s also a mistake to think that you have to make every single piece of content, at your ideal granularity, available at launch. No, you only need the fundamental shape of your content to be addressable, and in the future you can go in and make things more addressable and more granular. I’d recommend that on the back-end, you perhaps make things a little more granular than you do in your front-end, so that it’s easier to add new features and content quickly, but that’s a hunch rather than a time-honoured, proven strategy.

Now, whilst the ‘delightful experience’ can be built on top of the one-URL-per-thing, it’s much, much harder to do it the other way around. Indeed, most of the pain of innovation that I’ve found is in trying to make websites more granular after-the-fact, when none of the systems or user experience have been designed to accomodate this, even though it would give tremendous benefit to the user, oh, and to the business, too.

Silos are senseless

Because here’s the second major flaw in the criticism. If there’s one thing we learned from the first decade of the twenty-first century in Web development and product management, it’s that it’s expensive to build good experiences. Now, of course the tools may have got better since the early years, but the point still stands, especially for any company that intends to be alive on the Web for longer than the initial buzz around a release of content. Building flashy silos of content, with exceptional user experiences just doesn’t work as a mid-to-long term business strategy.

Never mind the fact that your ‘exceptional’ user experiences are almost certainly tied to a short lived technology, are designed based on some, let’s face it, educated guesswork about your possible users, and probably don’t work for any user not on the latest equipment or with any kind of visual or motor impairment. It just doesn’t make sense as a business model, unless you truly are a one-shot kind of company – which is fine, but, really, is that true of most companies? I’m not sure. This is because the content that you’ve tied up into your flashy experience – you’re almost certainly going to want to redesign, or incorporate into a different experience, in the future. Not just in the future – what happens if there’s a piece of content that you want to design multiple experiences around?

Simply put, it makes business sense to try and achieve economies of scale and of scope. This is done by abstracting common content wherever you can. Again, abstraction can certainly go too far, but to not try at all is a big mistake. Give the common, reusable bits of your content their own handles on the Web, then string together experiences on top of them. This way, you don’t have to re-invent the content every single time, and you almost certainly save yourself a lot of time and effort in doing so. This is a careful process, admittedly – you need to think hard about which things are likely to stay the same over time, and where the change will happen, but again, the return on investment will be worth it. Once you’ve done the hard work this time, the next time, you can concentrate much more on the ‘delightful experience’ – and people who are still engrossed in the previous experience can much more easily be upsold to the newer one.

Good architecture does not impede creativity or experience

Finally, it’s important to note that the design of ‘one-URL-per-thing’ does not have to restrict the experience in any way. Yes, it can produce a ‘boring, reference-work’ style experience – if that’s all you let it be. Indeed, the work of producing the basic site is not meant to represent the ideal experience anyway. It’s certainly not meant to be at odds, or in opposition to, your preferred user experience. Instead, it’s laying the groundwork for your future masterpieces. It’s about the long-term effectiveness of your core content, about ensuring that users can find the content as much as possible, and that they can share it and so on.

You can still go on to design your experience that you think is suitable for ‘Mabel, the woman who’s worked in a dry cleaners all her life, and just wants a passive experience’ (a statement which, while on the surface, and in our weaker moments, might be true, but suggests a slightly disdain for your audience – even a passive experience needs to engage the brain in some kind of inquisitive mode, in order to be successful). I certainly don’t want to restrict your creativity, and force you to only produce experiences that geeky information-hunter-gatherers would enjoy. I want us to produce things that as many people as possible will want to invest in, and enjoy, in as many ways as possible, not just for a one-time experience, but something that can be returned to, built upon, and perhaps even turned from a ‘passive’ into an ‘active’ experience.

But if we ignore the ‘one-URL-per-thing’ approach, not only do you miss out on the business and audience benefits I’ve described, I’d argue you’re misunderstanding the medium of the Web. The Internet is the two way channel, that brings us great things, and great experiences. The Web, as I’ve said previously, opens us up to building longer term, layered, generative experiences that are both delightful and truly creative, in ways we haven’t even imagined yet. Far from being a boring user experience, I think that’s actually a much more exciting future to aim for.

9 thoughts on “Popular Misconceptions about Designing for the Web

  1. Sandra Davis

    Ah, so maybe Channel 4′s “scrap booking” approach is the delightful experience built on top of a good solid base of one address per presenter *and* subject *and* program…

    I have mostly worked on smaller sites to date, which have not had quite so much content to organize, but I will bring this kind of thinking with me in future.

  2. Paul Post author

    Indeed, as I say, may not be always appropriate, but I find it’s a very good general approach – more often than you’d think ;-)

  3. Pingback: Bruce Lawson’s personal site  : Reading List

  4. Chris Hunt

    “If there‚Äôs one thing we learned from the first decade of the twentieth century in Web development and product management…”

    …It’s that it’s very hard to deliver a good user experience over telegraph wires, through the post, or by using Mr Marconi’s new-fangled radio.

    We learnt a lot more by the first decade of the Twenty-First century though.

    Good article, silly typos aside.

  5. Dave Rogers (@daverog)

    Great post, and I think I agree with the essence of it: build a great UX around a thing, at one-URL, where the experience can evolve over time and benefit from its permanent, canonical location (is that a correct summary?).

    Now, I don’t actually have a good solution to this problem, but what about ‘contextualised’ things? Clearly if it’s personal context (geography, device etc) then you’ll want to take responsive approaches, but what about more nuanced contexts that depend on what the user is interested in at a particular time.

    I am thinking along the lines of two contexts like these:
    - I am interested in Arnold Schwarzenegger, the politician
    - I am interested in Arnold Schwarzenegger, the movie star

    Would you consider each one a composite ‘thing’?

    If yes, what is this thing? (and would this be the start of a slippery slope?)

    If no, would each user want to experience something different depending on their interest?
    e.g.
    /things/arnold-s = generic page
    /politics/arnold-s = a politics ‘view’ over this person
    /entertainment/arnold-s = an entertainment ‘view’ over this person

    Dave

  6. Paul Post author

    Thanks for the comment, Dave.

    Context is hard, I agree. I’m not sure of an 100% solution, but I think it’s something that needs to be explored more – at the moment, I feel we shy away from it a little because it is hard, and/or design messily around it, rather than attempting a standard approach, and then iterating on that.

    My instinct is to go for a canonical, public facing URL for each thing, and then append the context to the URL, but also, in the public facing experience, I’d go a little less abstract than /things – so:

    /people/arnold-s
    /people/arnold-s/politics
    /people/arnold-s/entertainment

    This appeals to me mainly because of the hackability – that you could potentially see different aspects of the same person more easily, and because the information is similar across people. It wouldn’t stop you having a /politics or /entertainment aggregation of people, though.

    I can see that your suggested way will also have many advantages – and honestly, whilst I lean towards the above, I’d advocate the importance of trying small prototypes of both approaches and learning from that as a key next step, rather than making the decision up front. But that might not always be possible, of course…

Leave a Reply