'Private only to me' seems clumsy

Is there a better way to phrase this? I had to pause and figure out what it meant. What other types of privacy does Google recognise?

Perhaps Google Docs should reserve the Private label for unshared documents and describe the extent to which other files have been shared. For example:

• this document has been shared with 20 people
• anybody can view this document

Two things, mainly: The search itself and the way WordPress presents the results. Let’s say you’re searching for a post I published a while back about The Times recent redesign. You might (logically enough) tap ‘new Times design’ into the search box. Here’s what you’d get:

Search results for 'new times design' don't return the article about The Times redesign

The article isn’t returned at all because of the over–literal way WordPress searches through posts.

How WordPress searches posts

WordPress returns a post if it finds every word in your search phrase as a string within the title or content. If every word isn’t contained in the post then it won’t be returned. Note that it looks for strings rather than words, so a search that contains the word ‘new’ will find words such as ‘newspaper’, ‘news–stand’ and ‘Newcastle’.

It ignores proximity, so it doesn’t matter if the words ‘new times design’ occur one after another or are scattered throughout the post (or scattered throughout the post as parts of possibly unrelated words). Even more annoyingly, it doesn’t search excerpts, so even if the exact phrase occurs in the excerpt, the search won’t find it:

The article we were looking for even contains the exact search phrase in the excerpt

All strings are equal to WordPress, so using ‘the’ in your search term will generate a search ‘hit’ of equal value to a word that is more relevant, even if the post only contains strings of which ‘the’ is a substring; ‘another’, for example.

How WordPress displays search results

Having ‘found’ a bunch of posts in the database WordPress will display them using the search.php template and a loop. Loops are good for displaying sequential information, but inappropriate for displaying information that requires fuzzier sorting (such as by relevancy. Of course, WordPress has no concept of relevancy anyway).

This problem is exacerbated by the fact that WordPress is limited by what it can display within a loop. As there is no search_snippet() function theme authors can only display, at best, the post’s excerpt, and hope that this bears some relation to the search term. Some themes will display the whole post, which makes scanning for relevant results next to impossible.

How to improve WordPress search

WordPress needs to implement the following:

• whole post search, including the excerpt
• optional tag and category search
• search for whole phrases, as well as individual words
• search for complete words, surrounded by spaces, not just strings
• some form of weighting and relevancy: attach more relevance to non–common words, phrase matches and word frequency
• allow search loops to sort posts by a relevancy score
• implement something like a search_snippet(); function which returns text from a post around a search match
• output HTML that can be styled (a mark tag would be useful here, but a span would be fine)

There are a few search plugins that can enhance WordPress search, but it’s such a basic feature of a website that I think something like Sphider or Lucene should be incorporated within the core files.

Or maybe blogs don’t need advanced search? Perhaps users rarely visit blogs to find a particular piece of information. What do you reckon?

Ignoring the implementation problems, I think part of the charm of Twitter is its un–advertising. The brevity, immediacy and capacity to misspell or misphrase something are all part of this charm. True, I’m sure vast marketing departments prepare keyword–laden tweets months in advance, but the general tone is still pretty informal.

But they are our missives, so we should retain the right to delete them if we want to. Without a delete button we’d be just a tiny bit more wary when we tweet.

What do these symbols represent?

Now test using the website. Swap positions (and costumes) with your scientific mate. Get them to perform the following tasks: find the programme, find out about the venue, find out who the sponsors are (I guess this is of interest occasionally). Ask them to speak about the task and note how easy they find the experience.

I’m willing to bet that this isn’t the easiest set of tasks in the world and that the experience isn’t particularly enjoyable. Probably because the icons suggest the following:

What do the icons suggest to you?

The HD Live site has used something called mystery meat navigation. Although it was rampant a few years back, you still see it occasionally. I guess designers prefer the look of images over text, or just want to try something different occasionally. It is a gorgeous looking site, by the way:

A beautiful website to look at

And the answer to the question? It’d be easy to say never but that would be wrong: sometimes we might want to experiment, sometimes it’s appropriate to sacrifice usability for prettiness (on “creative”, artistic sites, for example), or sometimes it might just work. But textual lables are really the simplest way to do navigation menus as they take no time to interpret; why not use them?

My employer’s new website design — by which I mean a complete structural and visual overhaul — went live in mid–July. One of the targets set around the redevelopment of the website was to generate more enquiries from individuals and organisations about membership.

Long, buried forms equal no enquiries

Previously, we’d expect a handful enquiries a month (sometimes none). The reasons for the previous lack of interest were quite obvious:

• basic contact information wasn’t particularly visible
• links to contact forms were placed below long swathes of copy
• contact forms asked visitors to complete lots of fields and, even worse, make decisions as to what type of membership would be appropriate for them (rather than leave that decision to us)
• these links were placed quite deep within the site’s structure
• content detailed our processes and used our language to label products and ideas

The web developers came up with some simple ideas to make it easy to contact us:

• a join today call to action was placed in the sidebar of every page; this linked to a form
• a contact us link was placed in the top right corner of every page

The Join today link is available in the sidebar on every page

And I rewrote lots of copy:

• products were labelled with descriptive, audience–friendly terms rather than our own product names; for example, we labelled our customer satisfaction measurement product customer satisfaction measurement rather than ServiceMark
• products were described in terms of how they could benefit customers; we avoided long, procedural descriptions (and where detail was required it was placed further into the site structure)
• each page was given a clear call to action which was placed near the top of the text (note: not in an effort to keep it “above the fold” but in an effort to make it easily scannable)
• the number of form fields was kept to a minimum

One of the site's submenus. Note the absence of product names and the use of plain language labels.

The results

It wasn’t really difficult to improve on the old site’s performance: Since launch on July 15 the new site has generated a total of 140 membership enquiries.

What is interesting is where these enquiries came from. The vast majority came from the join today link placed in the sidebar of each page. I think there are several possible explanations for this:

• visitors are essentially lazy. Or, rather, they’ll reach for the easiest means of contacting someone. A sidebar call to action that exists everywhere requires less effort to find and hit
• similarly, visitors don’t bother reading copy in great detail; instead, they’ll scan and then look for the easiest way to respond to what they’ve just read
• visitors react to all sorts of content. So, instead of relying on them to find the section of your website about membership and complete the contact us about membership form (which has generated just 5 enquiries), it’s better to offer them the opportunity to join as a reaction to reading a good blog post, resource or anything they’ve found interesting.
• direct language that promises something in return (i.e. membership) attracts clicks
• the graphic brings attention to the call to action

This isn’t (of course!) to denigrate my own writing efforts: You obviously need to offer a good product and communicate its features clearly in order to generate the desire to click the join now button. And frictionless forms surely help. But it’s interesting to see how design decisions work in the real, measurable world, and how you can infer visitor behaviour from its results.

The taxonomy is exhaustive: it consists of 52 terms grouped into seven broader categories. It is — I think — an impressive piece of work.

But I’m not sure it’s actually very useful.

Scientific versus subjective taxonomies

Taxonomy is an attractive latinate term that confers a certain objectivity. There are, I think, two broad types of taxonomy, namely scientific taxonomies — the sort that categorise objective, physical entities, such as butterflies — and subjective taxonomies, which attempt to categorise abstract concepts, such as customer service.

Some subjective taxonomies err towards objectivity more than others, normally as a result of attaining a certain respectability and through agreement on underlying concepts over time. Ontology, epistomology and philosophy of the mind are established areas of philosophy, for example. Within customer service we could maybe state that customer experience is an established sub–term of customer service, but terms such as coaching, IVR and performance management are either too faddish, too vague or not connected closely enough to customer service alone to be included within the taxonomy.

Without agreement over concepts users — even experts — will find it difficult to relate content to taxonomical terms. What exactly are performance management and diversity and language?

New subjective taxonomies should be general

This is not to argue that newer areas of study shouldn’t have taxonomies. By using broad, general terms and sub–terms that have gained traction within a subject area publishers will be able to guide users through their content. Over time more thorough taxonomies will develop. Who knows; perhaps one day there’ll be university departments dedicated to areas such as customer service, something my employer would no doubt love.

• Download the 133 theme for WordPress (version 2, ZIP, 16K)
• See the live demo

Flicking and reading

The thinking behind 133 was to provide a theme that gave the reader nothing to do apart from read and flick through posts. It was built on Sandbox and took a grand total of 133 minutes to write (geddit?)

It’s been my second most popular theme (surprising, considering it wasn’t listed by Smashing Magazine), perhaps providing evidence that there’s a demand for design that promotes reading and little else (see Readability and Today’s Guardian).

Screenshot of the 133 theme.

To get an idea of what it does it’s best to list what it doesn’t:

1. no search
2. no archives
3. no tags or categories
4. no navbar
5. no menu
6. no comments
7. no Twitter integration (there was in version one, but it was pointless)

Usage

Excerpts

The theme makes use of excerpts by displaying them on the front page and above posts on single pages. If a post doesn’t have an excerpt, the front page will display the whole post instead, and the single page won’t display an excerpt (obviously), nor an abbreviated version of the full post. In short, feel free to use excerpts but don’t worry if you don’t.

Images

The content column is 560 pixels wide. If you add the class pull to an image it’ll place it in the empty left hand column. Pulled images must be 265 pixels wide; if they’re not, they’ll be resized. (Note: you can apply the pull class to any element). Centring images and floating left and right work in the normal way.

You can also incorporate full page images (the theme is 855 pixels wide in all) by adding a wide class to images.

Asides

I’m a fan of (proper) conversational asides in blog posts. Wrap anything in aside tags and 133 will pull them into the left hand column. (Warning: WordPress can be somewhat moody with asides; check your HTML after making any amendments).

Additional information

The theme will attempt to display the following info from behind the scenes:

Users → Your profile → About yourself → Biographical information is displayed in the footer.

Settings → General → Description serves as a tagline and is displayed in the header, aligned right.

Behind the scenes

The theme has been re–written from the ground up. It’s now marked up in shiny HTML5 and I’ve used the Google Font API to sprinkle some IM Fell DW Pixa around.

(A by–product of all this is to reduce the size of the zipped theme to 15.7K, including a screenshot. The stylesheet weighs in at 291 lines, including comments.)

It looks different

The structure is essentially the same: three columns, the first left mainly empty, the content placed off centre to the right. Titles are now set in IM Fell DW Pixa and body copy remains 100% (16 pixels). If you’re on Windows you’ll get Georgia. If you’re on a Mac or Linux you’ll get Palatino. Everyone’s happy.

The content column has been widened from 500 pixels to 560 pixels.

Image from Wikimedia

Perhaps you’re fond of these latinate words. What era does the word evoke? Victorian, Edwardian, perhaps; probably the past.

I came across a discussion of the term on the Content Strategy Google Group. It’s interesting to note that the contributor from the offline world dislikes curate; for her, it’s redolent of a patronising approach to presenting knowledge:

I think that my problem with the term, coloured by my background, is that it can be perceived as too grandiose, old-fashioned, bureaucratic and somewhat  alienating. As I read your comments and sift through some of the information in the various links, I am definitely having second thoughts about it as this new appropriation of the term certainly brings with it a wider sense of scope and a responsibility to audience that is quite different to that of a formal curator role in a museum. Amy Thibodeau

While the online world just can’t wait to appropriate it:

As someone working for an organization (and not an agency), “curation” is a great way to describe the strategic content work we do. Hillary Marsh.

What does this interest in the past tell us? At a design level it’s playful and textural (when done well) or annoying and kitsch (when not). It also indicates what the online world lacks; when anyone can set up a website and express an opinion, and when the medium is so modern, language, typography and imagery that suggest gravitas, learnedness and history (and even exclusivity) are attractive, especially to people who make their living from web design and writing. It confers legitimacy and professionalism.

Can you think of any other words that have been appropriated? The architecture of information architecture, perhaps?