As many of my readers know (that is, 5 out of 5 who read this blog), I’m actively involved in the startup space – our startup is Pressyo, and we have a couple of projects sitting on the launchpad, some with ignition already. My startup team is a very good team. We’ve went through multiple projects, failing most of the time (and the occasional success we’ve had rapidly degenerated into failures), and undoubtedly, we will fail many times to come in the future. For every failure we make, we dissect the failure, and find pain points, and fix them with tools. We argue a lot over why we failed, but the important thing is that we learn. This article is >2300 words long. If you want, you may jump to the TL;DR of Polish, Passivity, Privacy, Pressyo instead.
Polish, or the Cool Cam
My father, ever the gadget lover, used a smartphone before the phrase “smartphone” existed. Remember the O2 XDA? He had one of those. He went on from smartphone to smartphone, from O2 XDA to a HTC Touch to a HTC Diamond 2, which was last stolen in Dubai. Then he switched to an iPhone. Despite the lack of features, he had still found it a better phone, simply because the iPhone does what it is supposed to do very well. The user experience for the iPhone surpasses anything my dad had ever used.
I too, had been using smartphones from the time they were called PDA phones. My first was a HP h2210 hacked together with a CF-based GSM receiver. I then moved on to other phones, and finally settling on a HTC Desire. If I ever upgrade any time soon (damn you Telstra and your restrictive contracts), I would probably upgrade to a similar Android device. Because to me, Android phones do one thing really well: flexibility. Heck, I once pulled data off a csv off my work email and wrote a regression analysis while en route on a flight from Sydney to Brisbane on my phone.
The difference between the iPhone 3Gs and my HTC Desire? They cater to different people. The chrome of the polish shows differently to different people. I like to tinker  with stuff. I am tolerant of terrible user experiences – the worst of my smartphone experiences comes from dropped calls and terrible hacking of my HP h2210 (I literally took a program apart, figured out what was wrong with it and tried to recompile it to no success) – even my Symbian experiences with Nokia, whilst traumatic, wasn’t as bad. My HTC Desire has been a far cry from all that. Were I to present the modern day Android phone to someone of my father’s caliber, he would, I guess instantly like it, but probably not as much as the iPhone. Likewise, while I like the iPhone, I find it frustrating at times given that I can’t hack around to make it do things the way I want it to.
So you see, polish is not a single facet (as my prose may have led you to think). Androids are polished on their flexibility end. iPhones on the other hand, are polished on the UX end. If I were mean, I’d say the UX of the iPhone is the sizzle that sells the slightly-overcooked steak; while the Android is a perfectly cooked steak but because it was cooked sous-vide, has no sizzle.
Despite this, I agreed thoroughly with the counter-argument. The crux of the argument is that polish is necessary, and whilst I don’t agree that the polish on the UX is as necessary as the polish and chrome on the actual features, I have come to learn that UX can more often than not, act as the cool cam. I don’t think I will ever place as much of an importance of the external polish factor as Steve Jobs put, but I’ve thought through this myself long and hard, and concluded that UX (and other external polish like cool interface etc) are required, even at bare minimum. I think of it exactly like the situation as described in the Daily WTF, except, instead of board members, executives or investors that you’re thowing the product at, the people who judge you are your users of your product. Give them a cool cam, and they will shut up for a bit while they learn the ropes of your system.
Passivity, or I’m Too Lazy to Create an Account
Another thing we talked long and hard about was the concept of passivity, or ease-of-use. I would like to clear this upfront: I’m not talking about usability or accesibility, rather I am talking about a form of laziness that humans seem to have. I work as an analyst, and one of my many job descriptions is to use statistics to improve conversion rates. One thing is clear and obvious: the more intermediate steps there are to a point of user conversion, the more mid-funnel drops there are. To optimize for conversion then is to remove as many of the steps as possible without impeding required information.
Now, people who design web[based] apps/sites/services know this. People in eCommerce have yet to catch on (save Book Depository) – at least the sites I visit. You see, when I visit an eCommerce site, whether I want to buy books, or a new keyboard, or computer parts, I don’t need a friggin’ account! I just want to enter my credit card details, give you money, and get my stuff! Book Depository does a good job understanding that. The various computer stores I now have accounts with (14 of them, my last count, 8 of them don’t have my current and accurate address) don’t. Having to sign up for an account in order to get the cheapest Core i7 on the market is labourious, but I’m a cheapskate Asian, so I’d do it.
Imagine how many purchases I had cancelled because of that stupid fact that I had to create an account and do all sorts of complicated stuff. Thankfully, I do keep track. It’s roughly 50 transactions that would have been if I hadn’t been required to sign up for an account.
So what do people who design web apps, sites and services do? Enter OAuth and the like (ahem, Facebook Connect). I’m not going into the nitty gritty details of OAuth, but suffice to say, I’ve been following its developments since its inception. Essentially, if you’ve got an identity lodged with Identity Providers (Twitter, Google, Facebook, blabla), you can use that to sign in instead of having to create an account. An account will most likely still be created for you, but particulars are filled in from OAuth, if authorized by the user. This saves the user from having to fill out particulars again and again, removing potentially up to 3 steps from a conversion funnel.
This is a good idea. Sign in once, sign in everywhere. Brilliant idea. So why did we have a long talk/argument over it? Because it’s not a panacea to all the problems. I’m not referring only to the single-sign-on or OAuth or Facebook Connect. No, I’m talking about the entire concept of making things simple for the end user. It is a good idea, when used correctly. I have about 50 would-be transactions, but yet in the past month, I have bought myself a computer, with parts from 4 different online computer shops. Why is that? Because I needed it – my normal computer had died for over two months while I agonized over the decision on whether to spend money to buy a computer. I needed it, and so I signed up new accounts to 4 of the cheapest online computer shops, bought the parts and built the computer myself.
This is what we economists call signalling. Here are some very interesting facts on what I do at work: there are many conversion processes and they range from simple , to moderate, to tough; The simpler the conversion process is, the lower the customer retention rate is. While it baffled me at first, I gradually realized: simply put, a conversion from a low barrier of entry signals lower interest, hence higher churn rates.
Is SSO a good thing? Yes and no. Yes because it increases conversion rates (and from experience: by quite a bit ~ 20% range). No, because with increased conversion rates, comes increased churn rates (from experience: a fair bit). There is an old marketing adage, that is to say, it’s cheaper to retain a client than to acquire new ones. Whilst we’re planning to do both SSOs and manual account creation processes for our projects, we have to be very careful in balancing the act of acquiring and retaining new customers. Furthermore, we have to be aware that SSOs do pose a privacy problem for the end users, thereby increasing reluctance to convert, which brings me on to my next part: –
Privacy or You Did WHAT with My Email Contacts?
Over the last week, I had signed up to two new social networking sites. Not that I needed more friends or needed more ways to connect with friends. Quite the opposite actually These were LinkedIn-style professional networking sites. Not that I needed career advancement – I’m quite content writing machine learning algorithms to determine what sites are brand-safe for advertisers. One of them was pretty much a LinkedIn clone with emphasis on skills called Skillpages, the other was Reference.me. I’ll tell you upfront, that I was rather alarmed by the latter.
If you visit the site, right off the bat, you notice that it either uses Facebook Connect or Google OAuth to sign you in. This is what happens when you try to sign in with Google:
So, it wants to manage my contacts, eh? Normally I’d say NO WAY IN A FROZEN HELL AND A BURNING ASGARD and hightail it out of there, but I wanted to see what Mixtent could do with such a nice domain name. So I allowed access anyway. I was greeted with a page full of polish. Everything was smooth, everything went well. But something bugged the hell outta me. So like any good curious cat, I followed on, played around with it.
Overall, I was pleased with the layout, feel and look of reference.me. It felt good, graphs were functional, but I’d never use it. Ever. First, let me tell you how I got an account – I had received an email, saying “[REDACTED] has started following you on reference.me!” The rest of the email simply said: “[REDACTED] wants to follow you on reference.me! Wondering who else wants to follow you? click here”. The person in [REDACTED] was an acquaintence I knew from Electronics Frontiers Australia, but this is the best part: when I signed into reference.me, there was no indication that he ever invited me. There was no “you have a link pending”.
Me being me, and having worked with companies that would do seedy things to increase conversion counts, I immediately suspected that by requesting my permission to gain access to manage my contacts, the site had already sent out, on my behalf, invites to people. If I were a seedy company, that’s exactly what I’d do. And I’d only send the email to 10-20 people in a contact list – the 10-20 people of which contact had been the oldest and most infrequent.
I cannot tell if that email had been sent out, but this wanting to manage my contact list and the non-existent invite certainly did put me on high alert. As you can see from above, I actually did invite three other people (i.e. my closest friends and coincidentally people who are involved with me on the same startup) but none of them responded. With alarm bells ringing, I decided to write this blog about my experiences so far. I’d like to make it clear that I’m not in anyway insinuating that reference.me had sent me a viral email, or had sent any on my behalf. It is most probably terrible design and design decisions that led to alarm bells going off in my head.
Reference.me had checked two of the boxes above for me – it had Polish, and it put Passivity to great use, but privacy? That is a big Question Mark.
I have learned from my experiences, and privacy had been a issue that we talked a lot too. That is why most of the actionable public activity on our projects will be anonymous and private. Transactions happen anonymously too. We put a great deal into privacy simply because of stuff like that. To us privacy is a big design issue, and it demands a lot of attention – attention that many web app/site/service designers sadly not give.
Do notice the lack of the current buzzwords like lean startup, or pivot, or gamification. I’ll just leave them here as keywords for SEO purposes ;P, but I’ll actually come back and talk about it. One day.
- Polish means doing something really really well. Not just the UX
- UX is a selling point, not THE selling point
- More often than not, having a sizzle sells. Your steak may be cooked in the best possible method, but if there is no sizzle, you don’t sell. UX is that sizzle
- UX is necessary. It is the Cool Cam of a product
- Conversion rate depends on conversion funnel – the more complex a path to conversion, the lower the conversion rate
- Single-sign-on is a good thing because it reduces a step and simplifies the process to convert a passerby into a user/customer
- The higher the conversion rate, the lower the customer retention rate
- Must balance between increasing conversion rates and keeping a low retention rate
- It’s cheaper to retain customers than to acquire new ones
- Privacy is king on the Internet
- Don’t do dodgy stuff that may reduce the appearance of privacy or security
- All Pressyo projects are committed to privacy – publicly actionable activities are anonymous
Tell me what you think?
- Google will of course later be hyper obsessive-compulsive about everything↩
- and while doing that I sacrifice artifacts↩
- PocketPC and Windows mobile programs were called “programs” not “apps”↩
- which for the record, in my opinion is the best way to cook steak↩
-  a simple process would be like email or zip/postcode submit↩
-  like signing up for a user account↩
- like purchasing something and having to enter credit card and other more personal information↩
- by whatever metric: continuity continuation rate, activity level, repeat purchase level, response rates, etc↩
- depending on the product, but across all products, they’re statistically significant, and they’re a fair percentage, usually in the 5-30% range↩
- I’m a terrible loner with very few people whom I can call friends, and I most certainly do not need to find new ways to connect with them.↩