Adding features does not mean adding value

Features and value are two different things and sadly, it can be slightly confusing.

Building a new feature doesn’t necessarily mean more value to the end-user and feature overload can even lead to users’ confusion, which we all want to avoid.

By not associating “features” with “value”, it’s easier to make decisions about what to and what not to build.

All in all, the challenge really is into figuring out what will add value as I think it’s really easy to fall into the “let’s build this cool feature (even though the users don’t really care about it)” trap and waste precious time.

Figuring out what will add value can be fairly complex and I’m not sure it’s something that can be guessed that easily. For this reason, one approach is often to release a lot of “beta” features and eventually kill the features that do not work. By doing this, you avoid supporting useless features and don’t spend months fine tuning features that’ll never be used.

This is only one method among many others and is out of scope anyway for this post. I really just wanted to make the point that adding features doesn’t necessarily mean adding value, figuring out what will add value and what features to build is another problem.

What do you think?

I’ve been evaluating a lot of e-commerce solutions lately for a particular project I’m working on and I’m really disappointed by the open source offering in general: most e-commerce solutions out there are just heavy and/or a total mess to customize.

Because we’re talking open source here, most platforms are built using php. I wouldn’t say that the reason they ended up this way (messy) is because they were built using php, but for me it was a great opportunity to look for other e-commerce solutions built using newer programming languages like Ruby and Python.

This quest for e-commerce solutions built with new technologies and best practices in mind led me to a couple of very promising platforms built with Ruby on Rails or Python Django.

I’m not going to review the solutions in this post as my goal is only to raise the awareness about the simple fact that if you’re not satisfied with the current classic php platforms like OSCommerce, X-Cart or Magento, there are some decent alternatives.

Let’s have a look.

Spree shopping cart

I have to admit that so far, Spree is my favorite solution. It’s built using the popular Web framework Ruby on Rails and it’s very lightweight and clean. One could argue that it’s lightweight because it’s still missing a lot of features most php carts have, but it’s actually pretty complete for such a young product. You can have a quick look at the features list here.

Among the features, you will find localization, extensible design, use of css frameworks, custom tax and shipping rules, advanced product categories, SEO friendly, one page checkout and many others.

Another great thing with this solution is the community. Like any open source project, one of the most important aspect is the community support and also plugins built by this community. No problems for Spree, it delivers. There’s an active discussion board and also an extension repository with around 40 extensions ready for download and installation.

Visit the project’s website or have a look at the demo, it’s worth a look.

I haven’t tested spree yet in a production environment, but so far I’m impressed by what I see in my development environment.

Substruct

Substruct is another interesting alternative based on Ruby on Rails, but it is not as complete as Spree shopping cart. The project is hosted on Google Code and while that’s absolutely fine with me, the problem is that the project doesn’t have its own central website (with extensions, basic information, support, etc.) like Spree. It’s not as mature as Spree is.

If you have a quick look at the demo admin section, you’ll also find out that there are a lot of features missing and that it might be a problem to build a decent e-commerce website with all the functionalities you would expect.

For very simple needs though, it could end up being a very lightweight and easy to maintain solution.

For more information (download, demo, support), visit the project’s Google Code page here.

Satchmo

Satchmo is the third non-php e-commerce platform I’d like to cover. Unlike the other two solutions, Satchmo is built using the popular Python Web framework, Django.

I’ve heard a lot of good things about satchmo, but it’s a little bit painful to set up for the first time as you need quite a lot of dependencies. Usually I wouldn’t bother too much about that aspect as once your initial setup is all set, you’re good to go, but it could end up being a problem.

The problem here is that because of all the required dependencies, I would be worried about deploying this in a production environement on a traditional shared hosting. Feel free to have a look at the impressive requirements list.

The requirements might be impressive and the whole thing not that polished, but the feature list is pretty interesting. Among the features, there are some you won’t find in a lot of e-commerce platforms like support for downloadable and subscription products, support for rating and comments, related products suggestions and many more. Feel free to have a look at the features list for more information.

Conclusion

As mentionned earlier, this post was definitely not a review of every e-commerce solution out there: it’s more of a reminder that there might be some pretty good solutions that aren’t built using php.

I love php, but the code base of e-commerce platforms using php usually end up being really messy. In fact, it’s not just the case for e-commerce platforms.

Newer programming languages like Ruby and Python offer a lot of advantages over php and it’s definitely worth considering these newer languages for your next projects, being e-commerce or not.

Thanks to @kinaze on twitter for discovering (or letting me know about) that one.

The Yahoo Developer Network have released a short guide on the best practices for speeding up your website. It is divided in seven categories and contains 34 great tips to make your pages load faster and improve user experience.

There are some good tips in the guide and every web developer (or website owner) should have a look at it.

Here’s the guide: Best Practices for Speeding Up Your Web Site

There is a great quote from Henry Ford on the first car he built:

If I’d asked my customers what they wanted, they’d have said a faster horse

It is by far one of my favorite quote as it relates very well to the Internet industry and the whole customer-business relationship. It’s easy to take it very seriously and to think that what Ford meant with this quote is to ignore your customers and just do what you think is right, but it actually goes deeper.

I’m pretty sure that if Ford had asked his customers what they wanted, they’d have said something like faster horses and the reason is fairly simple: they couldn’t imagine anything else.

In fact, they didn’t want faster horses, they wanted a faster personal transportation method. It’s as simple as that. For Henry Ford, achieving this goal was absolutely impossible with a horse, so he came up with the idea of building a car that everybody could afford. Nobody knew they needed a car before they saw the Model-T (and knew they could afford it).

Twitter anybody? Twitter is the tool you never knew you needed before you had it: nobody wakes up in the morning with the needs for a tool like twitter, but the aggregation of different persons communication needs eventually led to the creation of twitter.

Twitter is the first obvious one that came to my mind, we can find plenty of other examples.

Customer needs on the web

How is this all related to customers needs on the Internet and to the whole customer-business relationship (note that the term customer is used in a very large context and can be your website’s visitors, readers, the people you build websites for, etc.)?

In fact it’s very simple:

Users will ask for tons of very precise features and it’s important to understand that they ask these features because it’s hard for them to think outside the constraints your product, website or service impose them.

What’s really important is to understand what lies behind these requests.

Do your website’s visitors ask for a certain feature because they simply want more exposure (egosystem)? In this case, you might think of a better or different idea than the feature they’re asking for. A feature that really answer their need and that is not limited by the external view users have on your product.

Do the customer you’re consulting for wants a certain feature on his website just to better communicate with his potential clients? If so, what the customer want is to better communicate: the feature he’s asking for might not be the best solution.

Innovation and feedback

The problem when you innovate and do things according to core customers needs and not feature requests is that you have no clue how it’s going to be received. That’s why you need feedback.

I posted a talk by Kevin Rose last week, and he said the worst thing you can do as a website owner is to pretend you really understand what your visitors want. To avoid this problem, he recommends to build features very quickly, release them and then test (feedback).

All in all, I guess the key is to act according to core customer needs (not feature requests) and put in place a good feedback loop so you can react quickly and make good decisions. Agree?

I got that one from Problogger today and I think it’s worth sharing. It’s a talk Kevin Rose gave at fowa (future of web apps) in which he’s sharing some tips on taking your site/web app from one to a million users. You know, kinda useful.

If you don’t know who Kevin Rose is, he’s the founder of the popular site Digg.

Honestly, it’s not rocket science, but I really liked the part where he talked about hacking the press, getting some coverage from big news networks (WallStreet journal, TechCrunch, etc.), using events to demo your product and getting influencers to talk about you.

Whether you build web apps, websites or blogs, there are surely some useful tips you can get from this talk.

Enjoy.

Taking your Site from One to One Million Users by Kevin Rose from Carsonified on Vimeo.

I read a post about 80Legs today on Mashable and the only words that came to my mind when I read the article was “they’re changing the game”. They are. Here’s what they do:

80Legs is a service platform for web crawling and processing web content. We put over 50,000 computers to work for you to deliver exceptional crawling performance at incredibly low costs. Our service is easy to use and completely customizable, so you can crawl and process web content however you want, whenever you want.

To summarize: you get your very own web crawler at a more than affordable price: 2.00$ / million pages crawled.

Why it’s game changing

Web crawling is a very complex and expensive process: crawling a web page, extracting content and scale the process to million of pages is far from being trivial and cheap. You not only need to build your crawler or use an existing one, you must have the infrastructure in place to support it all.

Now, anybody willing to build a niche search engine, extract specific information from a series of websites or build any web app involving crawling the Web at a large scale can do it at an affordable cost, without much technical knowledge.

It’s not 80Legs that’s game-changing, it’s the whole concept of democratizing web crawling and giving a chance to anybody with a great idea but without the money or the technical skills, to just make it happen.

It’s similar to blogging: it’s not about the platform you use (WordPress, blogger, TypePad, etc), it’s the fact that anybody can now have a voice and have an impact. Platforms die, ideas don’t.

It’s dead simple

I don’t want to talk too much about 80Legs because I haven’t really tested it (except creating an account and playing around) and I don’t like talking about stuff I haven’t tested, but it really looks dead simple. The process may look simple, but yet it seems to offer some decent advanced features.

I won’t go too much into details, but it allows a lot of customization so you can extract the information you want. Of course, you will need some technical knowledge if you want to use the crawled data and build a web app around it, but the hard part involving scaling issues is covered by a tool like 80Legs.

Spam is now more affordable than ever

As much as I’m enthusiast about democratizing Web crawling, there’s a huge downside: spam and content scrapping is now more affordable than ever. It’s an easy way for amateur spammers to build email lists by crawling websites and extracting email addresses. I’m sure we can find dozen of other spamming issues, but I much prefer to focus on the positive aspects.

Overall, the concept is extremely interesting and anybody willing to build a web app involving web crawling should be looking into this.