Jared Spool has an interesting article on top user experience challenges. They are:
- Scalability
- Visual Design
- Comprehension
- Interactivity
- Change Management
The other gap I see very often is Site Speed. In my experience, speed makes a tremendous difference to how much a site gets used. What is your sense on readiness of India’s skill pool to address the above – are we there or getting there to world class levels on these areas? I personally feel more confident of software issues such as change management, than areas like visual design and interactivity. Any thoughts?
Latest posts by Alok Mittal (see all)
- Promoters or Entrepreneurs – A choice for Private Equity players - August 3, 2019
- Startup Marathon Mindset - March 25, 2019
- What’s your Customer Culture? - March 4, 2019
Thanks Alok. I didn’t even know Reliance had their service up etc. – 12 K a month isn’t ultra expensive. Anyone use them here? Speed and service ok?
You’re right – no one “demands” better usability, but it’s always a trade-off between usability and features. Two approaches – the 37 signals approach of having less features and focussing on usability closely, versus the “big fat feature set” approach of building hajaar features and then getting to usability. The trade off is usually one versus the other but there is a middle ground – I could have the most usable web site, but it doesn’t do much to keep people on – you have to transition the firsttime user to the advanced user by ensuring that the guys that stay longer can do things faster.
Like Autocad. You can use the squiggly wiggly toolbars or just type /(command) and you’re as fine (IIRC) – in fact the more you use autocad, the more proficient you can be on it. EMACS on the other hand has huge features, and little usability. Visual Studio and Eclipse figured out an interesting middle ground using cues on source code editors.
Now the middle ground for a web site may be : find the features your users will absolutely love, and forget the rest. And then go and make them as usable as possible. And in today’s world and where a more “responsive” team is considered a better bet than a less responsive but better feature set, you are better off releasing often initially to just get your users to see something and tell you what they think (so you can respond and show how responsive you can be). So basically don’t go nuts about fine-tuned-testing, and don’t be alarmed by people who hate your site and say so. Nothing is better than being talked about, and if you can actually listen, change your site and make them like you, you have some seriously loyal users on your side.
Designing for scale in everything is crazy – you will have very little idea which part of your site will be so popular and which will need to scale like crazy. Imagine putting all your efforts into optimising user search results, when the max usage on your site is users checking group messages. It’s best to react in certain circumstances. Yes I can’t generalise but sometimes we refuse to admit our assumptions were wrong – and the refusal is costly.
Another thing to do is to think out of the “standard” box. I have programmed from the time when memory was HUGE constraint. Today because of that I have issues with in-memory indexes when all they will ever occupy is about 1 MB at a time. I have to fight myself from doing crazy stuff like that or caching unnecessarily (database access is so fast nowadays that you may be better off hitting it than caching). At the same time, it’s important to understand the nuances of the newer language sets like C# where string manipulation, because of managed memory allocation, has to be looked at in a completely different way (i.e. no “A” + “B” + … , but using a StringBuilder instead)
(Sorry for the techie note)
Having said that, I have to get back to trying a new Ajax funda and living with vile javascript. Later!
Hmmm.
Very, very nicely summed up by Deepak Shenoy especially regarding
the hosting options available in India. This week I’m moving from
one US hosted shared server to another less expensive and with a more bandwidth option US hosted shared server. Site speed is becoming an issue as I try to balance the scale of the database for search queries vis-a-vis adding features. Search is an area where you need multiple servers more then anything else but for the time being a single shared server works as I try and further tweak the back-end.
And this is my fourth UI redesign in the last 1.5 yrs. I’m sure skilled designers are out there somewhere but I guess it takes more the just a startup allure to attract them. Isnt that what entrepreneurism is all about:-). Do more with less. huh?
So while I’m focusing on adding features and tidying up the UI, site speed is slowly becoming an increasing concern.
Great discussion. Thanks for being so candid about where you might be going wrong yourself. Some comments/ resources:
Deepak
– I used to have a hosting service out of USA, and around 2001, decided to shift to India. We evaluated reliance and netmagic – both great suppliers, and price competitive on dedicated servers. Worth checking out.
– Users can articulate demand for features. They dont necessarily articulate the demand for better usability – doesnt mean it is not required. Look at your clickstream and see (a) what are people doing versus what you would like them to do, (b) how many people % drop off at every page. If your numbers are anything like what mine were, you will know that you need to fix this before anything else 🙂
Ram
– Agree, free markets will fix it. Efficiency of free markets depends on fluidity of information flow. If this is indeed a problem, how fast this will get fixed depends on how fast this is recognized and priced.
Kiran
– Point well taken – things are improving. The question is “are they improving faster than the technology is evolving”? Are we bridging the gap, or are we falling further behind? Your views on this would be great! Also, HFI might gain a lot by evangelizing this issue a lot more – some data on loss of business due to poor UI might be great to have even here on venturewoods.
Rehan
– I differ somewhat – once you make allowance for bad quality, that is what you get. Sure, when measuring, different apps might have different bar (because they stand to lose less), but at a level of defining a discipline, I think we need to keep the standards high.
Alok
I’m guilty of this as much as any other site founder – our site isn’t quite “usable” yet. We have now figured out what “doesn’t” work – standard ajax libraries like the Microsoft Ajax for ASP.NET is just not tight enough, and we have to now learn a lot of javascript to get our stuff usable. Time that wasn’t quite planned for, but what else is business about anyhow.
The other thing that hits speed is that hosting in India (or close) seems to be awfully expensive for the options we get. I’m paying $20 a month for our site and we get PHP, SQL Server, .NET, MySQL and a one click wordpress install, plus about 10Gig of space. This kind of option may work a few days (it’s a shared server) but eventually we’d like to get our own server – though honestly getting a virtual server hosted in India with a fat pipe does not look very possible (Am I wrong here ? Any good options?)
So we live with the US round trips and sub-optimised code. That’s ok for now because all we want is something out there and we’ll refactor, optimise and fix it for the back end, and we’ll work with some people in the US to design the UI and usability – not that it’s not quite there in India, I just find way more options in the US for something like this.
One interesting thing is that usability doesn’t quite seem to be as much an issue (or an issue at all) as the demand for features, so we’re focussing on features right now. Maybe it’s because the majority of our users don’t really care about usability as they’re fairly used to getting kicked around in other sites. or, it’s just that they’re too nice to tell us – either ways, we get a lot of “looks good and when do you thing feature X will be out” rather than “can’t use the darn site, dammit”.
Still, usability is an issue for us (as “techies”) but we want to keep business first. And speed is a major pain point, and we absolutely have to solve that, because we want people to not have to be scared of clicking on a stock link.
What a site would like to be is : easy for a newbie and as you get more used to the site, the ability to do things faster. A big problem I’ve noticed in a lot of sites is – they take way more information that you need to provide. I would at most want to register with an email and a password. Ok, a captcha at the most. But if you ask me my full name, my address (?), my city of birth, my passport number and shirt size, chances are I’ll get ticked off, and I can type more than 40 words a minute. For our site we’d like to implement the concept of search queries to do everything: typing “compare Nifty and Reliance” to bring up a quick comparison chart, for instance. We’d like to make it easy to add stocks to portfolios – just drag and drop, but to not lose scale as people go nuts with this.
Some of these interface things are subtle, but still need to be thought of thoroughly – and that really can’t be outsourced. What you can outsource to take your feature set and make it more “usable”.
So does India have the skill set? i think so. Are they out there willing to work for startups? I think not quite. You may actually find more options in the US, and with the rupee where it is, the US could actually be cheaper.
Correct me if I’m wrong..am happy to admit so if you can give me links.