Lee Duddell is the Founder of WhatUsersDo, a UK based company that offers online user testing to customers including O2, Dixons, ASOS and Schuh.
I've been asking Lee about the challenges of starting the company, the common user experience problems unearthed by testing, and how he sees the UX market developing in the next few years.
What gave you the idea for whatusersdo?
Like many startups the idea came from having been a client of the status quo approach to the industry that I ended up disrupting. I'd felt for a long time that the lab-based approach to usability testing was cumbersome. Cumbersome because it took weeks, involved consultants, numerous meetings, fat reports that nobody read, did not have the geographic reach and involved people leaving their natural browsing environment to come to a lab.
As a customer of the legacy approach I wanted something that was online, fast and that got me close to the users. I was so convinced that if I wanted this, so would others, that I did no market research and started building an online service that would meet my needs.
Underpinning it all was my central belief that the answers to the conversion and optimisation challenges that website owners face are already there - in the insights that users provide. All I had to do was find a way to get website owners closer to their users.
What were the major challenges in getting the company off the ground? Did you take any funding?
Getting the company off the ground was the easy part. I took seed funding to build the technology and launched a minimal viable product (and I really do mean minimal) within 12 months. Since then we've raised more private money and our initial investors (North Star Equity Investors) have recently re-invested to accelerate our growth.
The major challenge has been one of building awareness and educating the market that WhatUsersDo is not only an alternative to lab based testing, but actually better in many many ways. We're still on that educating mission and improving how we communicate the benefits of what we do.
I'm constantly filling my diary with speaking engagements spanning every conceivable sector to help spread the word.
When you're disrupting, positioning is always a challenge. Initially, we majored on price as being our main differentiator, but soon discovered that it was speed, efficiency and depth of insight that were resonating best with clients and were what they actually needed! The fact that we're 5% of the cost of lab testing is great, but not the main reason people use our service.
Another challenge has been to avoid becoming yet another UX consultancy. Don't get me wrong, I've nothing against UX consultancies and I'm pleased to say many of them are our clients. But, as a business our focus is to get our customers as close to their users' experience just as efficiently as we can.
Our experience has shown that if we do this right there's simply no need for us to become full blown consultants. We don't make UX recommendations to our clients, we provide the insights so they can either make their own design decisions or work with UX consultancies who'll make design recommendations.
What we have had to develop is the right level of analysis, support, training, advice and guidance for our corporate customers so that they can get the most from our platform and truly embed usability testing in their organisations.
I think we've cracked this in a way that means we can scale without the overhead of hiring consultants. It's always been a challenge and turning away consultancy money in the early days (when cash was only flowing out) was certainly hard!
What can other would-be entrepreneurs learn from your experience?
I'd say getting to market quickly with a minimal viable product and then listening to customers is what I'd recommend to any would-be entrepreneur. Then iterating on what you have and experimenting with every aspect of your business from pricing to product to positioning is the best way to grow something that moves from a good idea to something that people truly value.
Even as we grow we've maintained this agile culture, by remaining close to our clients and letting them dictate the shape of our product and supporting services.
How is business at the moment? Can you give me an idea of user numbers / revenue?
What recession? We're experiencing terrific growth not only in pure revenue terms (300% over the previous year) but also in the rate of new customer acquisition both through our SME SaaS service and through our corporate plans.
In the past year we've seen the likes of O2, Virgin Atlantic, Dixons and Sage really embedding WhatUsersDo usability testing into their design and development processes.
We're also starting to see an increase in demand for BRIC countries as e-retailers look to growth markets, such as Russia and Brazil, as viable sources of income.
Our recently launched mobile testing solution (covering SmartPhones and Tablets) is really taking off too as website owners start thinking seriously about how to provide the best User Experience in an increasingly mobile world.
Technically, we've seen a 1200% increase in mobile testing in the past two months (but as it's only just launched that figure is a little dramatic).
How do you select testers? How many do you have signed up?
The quality of our panel members is key to our business and we've many measure in place to manage this. We recruit users to our panel from social networks, direct advertising and job sites - there's no silver bullet and we're always adapting how we recruit.
Recently, word of mouth and referrals are becoming increasingly important sources of testers to our panel.
Panelists apply online through an audition test - each test is reviewed before anyone can join our panel. This checks not only their tech setup (e.g. mic levels) but also their ability to speak aloud while using a website.
We provide feedback to those that don't make it and reject about 20% of applicants. Panelists are continually rated as they perform tests and these ratings feedback into our QA process.
We not only offer a core panel (of about 100k from UK, USA, France, Germany, Italy, Netherlands and Spain) but also recruit on demand for specific projects - sometimes this involves using a third party panel company or recruiting direct through social networks (Facebook is a great place as people provide all the profiling information you could ever need).
Recently we launched our Private Panel service which allows clients to invite their own customers to participate in usability tests and is proving very popular.
What are the most common UX problems unearthed by the tests?
As you'd imagine the range is vast, from the pure usability issues that you'd expect such as forms not behaving as users expect to the back button losing search results, but also insight into the PET (Persuasion, Emotion and Trust) aspects of a website.
From last month's tests some of the most common (and interesting) issues were:
- Unclear returns policy (written by a legal team, not a web content editor).
- No guest checkout (users are increasingly expecting this).
- Name entry form not supporting apostrophes (we ran a test with Irish users).
- Irritation that out of stock items were added to the basket (speaks for itself).
- Price sliders hard to use (on an iPad test).
- And my personal favourite, one user described the Scottish Widows home page as "looking a bit like Dracula"!
How do companies use whatusersdo? Is it one of a range of tools for improve UX? Do they run regular tests with you?
We're about unearthing the WHY of user behaviour, so that web teams can look beyond their analytical reports and really understand their users.
WhatUsersDo is used in conjunction with content testing tools (e.g. Test and Target, Google Optimiser) and other sources of user feedback (e.g. Kissmetrics) to improve UX.
What I've personally found really interesting is the way that our clients use WhatUsersDo to not only test their own live website, but also to get user insight into:
- Competitors' websites.
- Pre-release websites and prototypes (e.g. Axure).
- Observe people naturally searching for products, starting at Google.
- Running card-sorting and tree tests.
- Content testing of landing pages, email marketing and other comms.
Our corporate clients now run tests at every stage of the development/design lifecycle, even before starting a new project they research how users respond to competing offerings and build with that UX insight in mind.
I know I shouldn't have a favourite client, but I love how the folks at DRL Limited (www.appliancesonline.co.uk) use WhatUsersDo.
It's not just the 9.5% increase in online sales that this activity lead to that's interesting, but the way the entire organisation became focussed on user needs rather than their own hunches and opinions.
We enabled a cultural change to a UCD approach, something that we're seeing at many of our clients - especially the larger ones.
How many tests are needed to get a true impression of the user experience on a site?
We get asked this a lot. It used to be widely accepted that five user tests is all you need to find 80% of usability issues and that it's diminishing returns after that, but that figure is primarily relevant to lab based testing.
Our experience has shown that the answer comes down to several variables:
- The number of journeys that you need to test.
- The range of devices (mobile users will uncover different issues to fixed web users).
- The different audiences that your site serves.
- The nature of what your testing.
Most websites are not one-dimensional, they increasingly serve multiple customer types with many different needs and goals.
For example, even a relatively simple clothes retail site should be tested for gifting journeys, self-purchase journeys all with different starting points - home page, product page (post search) as well as being benchmarked against competing sites.
Bring in an international audience, new customers vs existing and multiple devices and you'll soon see why five wont cut it.
So, our recommendation for those people starting out is to test with 10 people from at least two segments on the main buying journey. Then, by looking at your analytics, identify the next round of testing. And keep testing!
How do you see UX developing over the next few years?
We'll see more sophisticated ways for companies to measure UX as it is increasingly making it into the boardroom and thus needs to be reported on and measured. NPS (Net Promoter Score) is one measure, though perhaps too simplistic, that's making its way into company boardrooms across Europe now.
We're working with our clients to develop UX measures that align to their KPIs, in order to give a board level view of UX.
It goes without saying that UX is becoming increasingly challenging, not least because the mobile web means new behavioural patterns are emerging that brands need to understand and serve.
Next year the TV will become the battleground for Apple and Google - owning the living room, with all of its ad revenue (remember, Google is an ad business, not a search engine) will bring its own UX challenges.