Last year I had an influx of mobile projects and needed to find a way to carry out usability testing on a mobile device. I’ve been meaning to share my solution for a while, but it’s taken me until now to get round to it.

When it comes to carrying out mobile usability testing there’s a variety of well documented solutions, for example Harry Brignull’s usability testing sled made for a fiver, Nick Bowmast’s variation on a theme, and Lokion Interactives pimped sled beautifully monogramed and made by Ponoko. There’s also a great slide-deck from this years IA Summit which summarises the different approaches that can be taken, including the pros and cons of each.

What would MacGyver do?

At first I considered copying an existing solution, especially as there’s so many good ones already floating around. However, I had several requirements that I didn’t feel previous sleds had answered completely enough. I wanted to create a testing sled that was:

• Unobtrusive for the person using it – this isn’t easy by any means but I wanted to try and stay out of the way as much as possible, meaning the sled and camera had to be small, lightweight (light enough to hold in one hand) and have minimal impact on the participants field of vision.
• Of a professional standard – as I work in a commercial context the sled had to represent my client and agency in a professional way. I’ve seen some solutions that rely on Blu Tack or sticky-tape to hold them together, which is a perfectly fine solution but something I personally wanted to avoid.
• Adjustable and interchangeable – I wanted the ability to alter the camera position (to allow for lefties and righties) as well as accommodate multiple devices without too much fuss.
• Easily duplicated and disassembled - heading up a rapidly growing UX team at the time the rig had to be remade easily and consistently within a short time frame. To accommodate different devices and testing in multiple locations it had to be in kit form.
• Made from widely available parts – to allow for future duplication I wanted the rig to contain no expensive, limited edition or bespoke parts.

The ingredients

I shopped around for affordable parts that were readily available so I could make more in the future if everything went to plan, and if something went horribly wrong I could source replacement parts easily. I managed to get everything, including the camera, for a little under £42 (including postage). I used:

• Hue HD USB webcam
• A cheap flexible iPhone case
• 1.8m USB extension cable
• Black cable ties
• Adhesive Velcro patches*

* Initially I planned to use superglue, but instead opted for small adhesive Velcro patches so that it could be disassembled if necessary

The only tools used were a scalpel and a Sharpie (to mark where to cut), MacGyver would have been proud.

Making the sled

Firstly I had to attach the webcam to the case, in a way that wasn’t permanent but was secure and stable. The Hue HD webcam comes with a USB stand, but it wasn’t necessary as the USB on the camera could be plugged directly into the extension lead.

Four small cuts were made in the back of the iPhone case, the space between them equal to the width of the USB extension lead. Two cable ties were passed through, horizontally to the case, and left untied.

Then the USB lead was placed between the cable ties, which were tightened to secure the lead in place. A small square of adhesive Velcro was placed just above the USB port with the other half stuck to the webcam’s USB plug. This meant that once the webcam was attached to the USB port the Velcro held it in place and stopped it from moving around, or detaching under it’s own weight. Initially I was worried that the Velcro wouldn’t hold the weight of the camera but it actually worked well and was pretty solid. Finally the iPhone was clipped into the case.

The end result

The camera was attached so that it curved up from the bottom of the phone and therefore didn’t obstruct the users view too much, it also meant that the cameras built-in mic was close to their mouth. Once the camera was attached and the phone was in the holder the USB lead could be attached to a laptop, which in this instance was equipped with Morae testing software. By using Morae, we could position a second webcam (we used the laptops built-in webcam) to capture the participants facial expressions and body language.

It took a little bit of tweaking to get the camera positioned correctly so that it was in focus, and the webcam did add weight to the phone and unbalance things a little, but without hands on experience of other testing sleds I can’t say whether this was better or worse than other solutions. We also found that, if we didn’t get the position perfect, occasionally the camera was susceptible to wilting to the right or left but only very slightly and not to the extent that it was noticeable to the participant or detrimental to the recording.

That said for only £42 (not including the recording software license) and only taking 30 minutes to build from scratch I was really pleased with the end result. It was straight forward to adapt for other devices (e.g. iPad and Android devices) and very convenient to transport. I’d recommend it as a solution, and it definitely worked for me, but without trying out alternatives I couldn’t say how it compares.

If you have a go at recreating this sled I’d be really interested to hear about it, whether the experience is good or bad.

Last week I commented on Michael Wilson’s post about ‘sort by default‘ as an option when customising search results or product listings. I shared my personal experience with a recent client and thought it was worth sharing here too.

Intelligent defaults

In short, the point Michael made was that sites providing the ability to sort content without setting a relevant default are missing a trick.

An example of how ASOS don't set a default 'sort' state

Michael used the example of ASOS. Although they include a sort feature with the options ‘what’s new?’, ‘Price High to Low‘ and vice-verse they don’t explicitly set a default state. This raises the question of how the products are currently sorted, is it editor-defined, chronological, alphabetical or something else entirely?

Although this is a valid issue and something, as User Experience professionals, we need to be aware of it’s a relatively quick fix. The complexity is in understanding what the sort says about the website in the first place.

What does a default sort say about a brand?

Recently I had a conversation with a travel destination client about this exact issue. During a workshop intended to cover off the finer points of a prototype a heated debate started around what the default ‘sort’ state should be for holiday search results. Should it be an ‘editors choice’ or be sorted by more neutral means;  location, accommodation type (lodge, chalet, etc), availability, or price.

After much discussion everyone agreed that, based on our knowledge of the customer, price was the best option. But then came the question; should we sort high to low, or low to high?

By preselecting ‘high to low‘ you communicate that you are a higher-end brand and that quality, rather than cost, is a priority for your customers.

Conversely, by presenting items ‘low to high‘ you align your proposition with affordability, value, and competitive/budget pricing rather than the sense of exclusivity or luxury.

Sorted

For the travel brand it came down to making a fundamental decision about themselves that, surprisingly, they hadn’t openly discussed or defined before; are we a value/budget brand (such as Butlins or Easyjet) or is money not an issue for our customers and therefore focused on quality, closer aligned to brands such as Mr & Mrs Smith or Kuoni? Once we posed this question to them it was an easy decision to make and helped drive other decisions across the site.

Its safe to say, with hindsight, that this was an issue we should have had clearly defined at the start of the project as part of a wider strategy. In actual fact it was, but with so many stakeholders in the room it became apparent that it was not a shared view and certainly not something that had been openly discussed.

The default state of a sort isn’t going to dramatically change people’s perceptions but it’s this kind of little detail in my opinion that really matters as it can help to provide a cohesive and consist experience.

With a clearly defined experience strategy these sorts of decisions should be straight forward and not open for debate (e.g. “we’re a value brand appealing to families therefore the only logical answer is to provide our customers with the cheapest options first.”), without this the experience can end up feeling disjointed and can lead to conflict.

In short, do sweat the small stuff, but be clear on your strategy and proposition so that you keep the sweat to a minimum.

It’s been a while (6 months to be precise) since I posted anything on here and although I’d rather my first post back was one of the many half started blog posts I have stored away, I felt this was an ideal time for me to share some news and go some way to explaining why I’ve been so quiet of late.

• Launched a new portfolio – After about a year of procrastination, and with a little help from Matt Budd‘s coding skills, I finally relaunched my personal portfolio website. A lot of  work has gone into it and I’ll hopefully blog more about it soon.
• Exciting projects in the pipeline – The first half of 2011 has been an exciting time for Redweb and the UX team has been kept busy with some very interesting, and challenging, projects. There’s a lot of things I’ve learnt while working on these projects that I’d like to share but I’ll wait until the website’s have gone live before I do.
• A baby on the way – We’re expecting our first baby later this year, which I’m understandably stoked about. I can’t wait to become a Dad, and although I know it’ll have a profound effect on me I’ve made the decision not to write about my personal life from now on so I’ll be trying my hardest not to mention it again.
• Made a big decision – after 6 years at Redweb I felt it was time I set myself some new challenges and move on. It’s been great to witness Redweb grow and evolve over the years and I feel privileged to have played a part in it. But, as of this week, I’m joining SapientNitro as an Information Architecture Manager.  I’m really looking forward to the challenges this will bring and hopefully I’ll share them on here (if I get the chance).

This means there’s now a UX-shaped hole at Redweb. Although, initially at least, they’re not directly replacing me they are looking to recruit a new User Experience Consultant to help bolster the team.

Before leaving I worked closely with the team to make sure the quality of the UX offering and the way it permeates through the agency is not lost or negatively impacted. I have the utmost confidence in the team I leave behind and I’ll be watching them with interest over the coming months. The vacancy is a great opportunity for anyone wanting to work on the south coast as part of a professional, well established and award winning team with some interesting and varied clients.

Sound interesting? Take a look at the job description for more details. Alternatively feel free to get in touch with me if you have any questions you’d like to ask.