I’m impressed that the Prof pronounces your first name, if it’s one of the names he knows – it makes his talk feel a lot more personal. And he even knows my name, “Rory”, which is fairly uncommon. How many names did you record? How did you choose those names?
On the home page, the “How it works” three-page widget scrolls too slowly when I click one of the three page circles. That speed is fine for the automatic page-turning every 8 seconds, but if I click a circle, I want to see that page right away – the animation should be about 5 times faster. I was slightly annoyed by the slowness when I tried to go back to the first page, which you don’t want.
It also might help to stop that widget from scrolling until the user has scrolled far down enough that you think they have started reading it, so they can read the first step first, and aren’t forced to click on the first-page circle to go there themselves. Perhaps you don’t care which page the user reads first, but the chronological ordering of the pages contradicts that idea.
When I read https://www.sleepio.com/sciencebehindsleepio, I was impressed that you said you had done experimental trials, and I couldn’t see any sign of obvious bias in the experimental design, but I was still a bit skeptical and worried that the experiments were biased. I just realized that an addition that would have convinced me a lot more strongly would have been a LaTeX-formatted scientific paper about the experiment, typeset in Computer Modern. The paper would not promote Sleepio, but just describe the experimental method and results. I don’t know if LaTeX supports exporting to a web page with the Computer Modern font embedded – if not, you’d have to link to a PDF, which you’d have to work hard to convince me to open.
Thanks for taking the time to look at it and give such thorough feedback.
The name thing is the one thing that always seems to delight people; it's indicative of the care/attention we pay (hopefully) to the user throughout the programme. To arrive at names we started with census records of top baby names going back 50 years, and now periodically add to them with names people enter that we don't have. It degrades gracefully though if we don't know your name!
Noted re: speed of transition. We're adding paddles today too. We decided to keep all animations in synch, but that's a good call on waiting til viewed. We'll look at it.
If we can get permission from the journal we supplying the original paper would be a great idea, although a more accessible format might be better. You can read abstract on pubmed here: http://www.ncbi.nlm.nih.gov/pubmed/22654196
Thanks. Yes we initially had a 7 day trial but are now trying it without. The challenge is how to manage expectations from that trial - you need to stick with the programme to see a positive effect (like a weightloss programme), and despite our explanations mot people expect to see some instant results in that first week. We were hoping the 'Try it now' might give people enough of a taste of what the user experience is...but we'll give it some thought afresh.
Starting with "At what time do you normally get in to bed?" to this Non-24 sufferer living in Germany is sort of like opening with "Enter your ten digit phone number" or "What is your ZIP code?"
The thing we've wrestled most with (and haven't yet cracked) is how to accept time entry in a way that is unambiguous to even the least confident users, and also encompasses the range of quanta we need to manage (from mins to many hours). We've tried a few different solutions but 24hr dropdowns result in the least comprehension errors!!
We've taken what are the clinical standard measures - and some indeed feature counterintuitive questions. And you're right, one of the main issues with 'insomnia' is an unreliable sleep pattern.
We don't want to know how you slept last night, since this may be unrepresentative; but as you say an 'average' may also not capture it/be a natural way to report it. 3 x sample points ('last 3 nights') may be longwinded. Will consider...any ideas on a better question?
Start with "do you go to bed at roughly the same time (within 1-2 hours) most nights?" and only ask the "what time" question if people say yes to the first one.
How about a visual analog round clock where people can drag the hour hand to a certain point, and choose to create a range by dragging and so extending the time slice with another hour hand.
Thanks for the suggestion - it's something we've considered; the issue with our tests to date has been getting the minute-to-minute fidelity with a circular interface (plus it's actually quite hard to drag something accurately in a circular motion). But we should revisit this all with fresh eyes - and the idea much appreciated!
I'm currently working on a similar service for training workers in the heavy industries (oil, forestry, mining, etc). Flash based first, HTML5 next. A site design that feels more like an environment than a web page, led by interactions with a virtual character.
It is hugely inspiring to see that you've just nailed this sort of seamless and beautiful experience. I'm going to sign up for a course just so I can see how you've handled the various design challenges inherent to such a service. We've got our own exciting plans, but I'd love to learn from your example.
Also, I've got one nagging question I'd love your answer to. We've got a bunch of artists in-house, but don't have much in the way of great tools for animation. What tools did you use to do the animations on the Prof and the backgrounds? (Love the moustache mountain!) I'd be surprised if that was done with Flash keyframes, as I find them to be painfully slow to work with.
Thanks for releasing this here on HN. I hope you have a great launch. I'll be sharing your service with my team, and my friends, many of whom are frustrated night owls looking for better morning time.
Thanks so much! Sounds fascinating - what's your service called? Would love to see how you've approached that challenge - I know for us it took a lot of iterations!
Yes that's the vision! That when you have a problem, you just 'visit The Prof', and he gives you a blended programme to help you. Sleep is a good starting point, since it is a) socially acceptable to talk about, and b) is highly co-morbid with other problems - a good way in to health in general.
A limitation of our current media-rich approach however is the cost of adapting content - we took a short term cheap, long term expensive approach... :)
To me, the obvious follow-up would be social anxiety because it is intrinsically hard to talk to other people about, yet discussing it with a computer doesn't pose any difficulty. (I'm currently reading a 'use CBT to cure social anxiety' book which has been rather eye-opening.)
Regardless, I like the site, though it sometimes gets a bit choppy on my old macbook pro. I think £50 might be a bit steep as an up-front cost - it almost put me off.
Also, have you considered doing an iphone app for the diary? I use some supposedly motion sensing app to wake me up with music. It doesn't work very well, but for your purposes I could easily see just thumping the phone when I notice I'm awake instead of having to look at a clock and remember a time.
Yes, we'll need an umbrella name. But right now it's more important for it to be immediately obvious what the problem is that we can help with, since we rely on self-diagnosis, (the user is actively looking for a sleep solution either online or in Boots store).
It took A-G-E-S. Since we have bootstrapped for the last 3 years (only just took some seed money) we don't have the cash to pay our way in like most suppliers. Back then I had already got Prof Espie on board (he and I are the co-Founders) after his techniques had cured my insomnia. Since he is a world-leading expert the Boots Innovation Centre attended one of his seminars, and that prompted the first conversation, whilst we were still at planning stage. But since then it's been a very long road, talking to literally hundreds of people and outlasting team after team of buyers and marketing folk...the only approach with someone like Boots is to build a partnership, but that takes time if you don't have money.
After a lot of iterative development and rigorous clinical testing, we yesterday released Sleepio - an automated, online Cognitive Behavioural Therapy (CBT) course for poor sleep. Would much appreciate the wise views of HN, positive and negative!
Good point. We actually have support for subtitles - but our first goal was to mimic a more human interaction and see how it helped adherence rates. So far since users make time to sit down and follow the programme each week this doesn't seem to have been a major problem, but I agree for trying it out the requirement for sound is limiting.
Awesome. I have to say, this is very impressive. Is there a free trial or anything available? I would love to try it out but, alas, I'm a skint student.
Especially since it's the start of the academic year I can imagine this going down well with my student friends - maybe some kind of discount would make it more enticing.
Either way, this looks awesome and clearly the result of a lot of hard work. The Boots partnership makes it all the more exciting. Congratulations guys :)
Thanks so much! We did have a free trial, but we're trying it without...and since we're so close to launch we need to coordinate with what Boots is offering. Out of interest what sort of price do you think you (and your friends) would expect to pay for this, based on what you've seen?
Thanks. I'm a big believer in using physical retail to explain (and confer value on) digital products. See above - the Boots deal took years of time and effort.
This site is the perfect example of why HTML 5 won't replace flash anytime soon. Because, this site pushes the perks of using Flash to its fullest. Beautiful illustration, excellent animations and simply awesome...Keep it up guys.
We built the main sessions in Flash first because it was the quickest way to discover what it was that we were building! So the Flash version has been around longest and is the most stable (and currently the slickest). Once we were confident of the functional requirements we built the HTML5 version, but that's still in beta so we display the Flash version by default if you have it installed/enabled. Quite a few technical challenges we had to overcome to chain videos seamlessly in HTML5 on iPad...we'll definitely be writing about it once the launch dust has settled...
I've watched this startup progress from the early customer/problem discovery stage to MVP, validation and beyond. The execution and commitment involved has been an inspiration. Congrats guys. Very impressive.
Hmmm because our primary target (middle aged women) are not that web-savvy i guess, and the .com is the default TLD. We bagged sleep.io when it first went on sale but haven't considered using it as the default.
To ask the opposite (and possibly stupid) question - why would you make sleep.io the default?
Somebody was in the news the other day because their clever country-code TLD was taken away by the registry. All of the country code TLDs have weird restrictions with which you can easily fail to comply. Recourse for smaller registries can be harder to obtain.
In this case I'm not sure why they went with 'sleepio' if they weren't going to use '.io' other than perhaps it's trendy right now, but I think they probably made the right move getting the .com.
I just noticed it redirects from sleep.io is all. It seems like while .com's are great, having something like sleep. whatever is pretty awesome. I was just curious as to why you chose one redirect over the other.
We decided that while we wanted to capture the .io domain for completeness' sake, for the reasons listed above (SEO implications, not "correct" usage of TLD), sleepio.com made more sense as the default / redirect target.
Answered a similar question the other day. Webmaster tools doesn't let you target .io domains globally, as it's intended for the Indian Ocean countries. Therefore it won't rank as well as a .com in the Google serps, which is a big disadvantage.
Well, the techniques are out there in the literature, but it's how they are tailored to your own situation - and more importantly how you actually deliver positive behaviour change, rather than just info - that's the real challenge that we're trying to address at a price less than the £600 you'd need to pay for a course of face to face therapy.
We're going to try a lower price point ongoing subs model to avoid that initial barrier; the assumption initially was that we need to front-load the value, since we're curing people!!
I'd gladly pay £100 or more.. If I actually believed the product would work on me. However, there is so much crap and scams on the internet that it is hard to believe a web site programme would be effective. All the thousands of "get abs in six weeks!" ads has made me cynical. Maybe if someone I trusted had taken the programme and could recommend it to me it would put be over the trust threshold.
So - we've tried to be strict on limiting our messages, and we've developed these based on iterative testing - with users on the site, but also on the street with strangers (shoving the pack under their noses and asking what they think it is!).
So it's always been a candidate, but other messages seemed more important, largely because it seems that most people think poor sleep -> sleeping pills, and do not think -> CBT / therapy. As in, it's the immediate frame of reference. That's been confirmed in training/talking to Boots staff, but they have found that reference point useful in making the value case to customers.
Ha! Top one is the best - not sure how effective that would be in getting anyone to sleep though...
It's a tricky question - there's no solid data on the awareness of Cognitive Behavioural Therapy but it lends credibility to those who do know it. In our (quick and dirty) research it seems that a majority of our main initial target (middle aged women) have heard about it but don't know what it is...
So that you provide as little friction as possible to getting people to sign up. For example, with your current button I was reluctant to press it because I would assume there would be a big form on the other side. However, I'm more likely to sign up for more info if it's only an email required.
You've inadvertantly highlighted a UX failing of ours - that's not actually a sign up form!
We ask for your first name so you can meet The Prof; another thing we're wrestling with, how to introduce that (non-standard) experience in a clear way...
"All we need is your name to show you a preview", or "You won't need to enter anything else unless you purchase". Emphasise that the name is all you need, I think that would solve the issue.
Thanks, but I actually think we need to be even clearer. It should be obvious what to do without reading anything. And if high-functioning HN users are making that error, your average Boots customer almost certainly is.
Unfortunately there isn't an obvious affordance for this sort of interaction (that I've yet come across) - it looks like an email signup, when in fact it's more like 'play a game'...an interesting challenge...
I'm just thinking out loud here. One approach is to replace the green Try It Now button with an image of the Prof, under which you say "Hi. I'm the Prof. What's your name?", following by text field and green button. Then experiment with what the green button says.
It would be interesting to see if a vague 'what's your name' approach, rather than a specific direction with a stated outcome, wins in a split-test.
Nice idea - we would need to connect the headline explaining the programme with the Prof more clearly, but that's pretty easy. Will work it up, thanks!
On the home page, the “How it works” three-page widget scrolls too slowly when I click one of the three page circles. That speed is fine for the automatic page-turning every 8 seconds, but if I click a circle, I want to see that page right away – the animation should be about 5 times faster. I was slightly annoyed by the slowness when I tried to go back to the first page, which you don’t want.
It also might help to stop that widget from scrolling until the user has scrolled far down enough that you think they have started reading it, so they can read the first step first, and aren’t forced to click on the first-page circle to go there themselves. Perhaps you don’t care which page the user reads first, but the chronological ordering of the pages contradicts that idea.
When I read https://www.sleepio.com/sciencebehindsleepio, I was impressed that you said you had done experimental trials, and I couldn’t see any sign of obvious bias in the experimental design, but I was still a bit skeptical and worried that the experiments were biased. I just realized that an addition that would have convinced me a lot more strongly would have been a LaTeX-formatted scientific paper about the experiment, typeset in Computer Modern. The paper would not promote Sleepio, but just describe the experimental method and results. I don’t know if LaTeX supports exporting to a web page with the Computer Modern font embedded – if not, you’d have to link to a PDF, which you’d have to work hard to convince me to open.