Follow the passion

A blog for people who love to think

Long tail Feature Requests

Recently Chris Brogan wrote about LinkedIn features he needs. After reading his post and comments along with Highrise customer forum I came up with the following:

A LOT of people want features which are extremely valuable, but only for a small group.

And they are willing to pay for them. If it’s possible to create a sort of Features API for a particular App then a user can pay to a “features vendor” for developing and hosting features, which are relevant for the user.

“Feature vendor” here is a third-party developer or company not affiliated with the App vendor.

This is how everyone benefits. But we have a lot of problems here: conflicts, privacy…
usability!

I dunno…What do u think?

Advertisements

May 31, 2008 - Posted by | future, tech | , , , ,

10 Comments »

  1. If I wanted a special Facebook app and you were a feature vendor for Facebook then how do I trust that you can understand exactly what I want and deliver it in the time I need it?

    Let’s say you deliver something and I say that’s not what I wanted. How many changes are you willing to make and how much time/patience do I have to give you? It varies a lot from person to person and based on situation.

    Or maybe after you deliver it I simply disappear and you never hear from me…

    Custom software development is not for everyone.

    If I was willing to pay a few thousand dollars to someone to develop a Facebook app per my existing specifications (which the developer must ask for and include in contract) and if I put the money into an account that is managed by Facebook or a trusted third party then I can see that working, but this requires me to be able to produce a spec, negotiate a contract and accept/reject the finished product based on whether or not it meets the spec.

    So to make a long story short, I don’t think this idea could succeed at this time. When we have “bots” that build features and understand human language then … that’s when your idea would work well.

    You would need to build a bot that can understand some intermediate software specification language (somewhere between NL and UML) and that can build entire applications (using templates, etc)

    🙂

    Comment by evolvingtrends | June 15, 2008 | Reply

  2. Hi Marc,

    Thank you for your comment!

    Basically you’re right. It’s not for everyone. And it’s not for Facebook. It’s mainly for small to mid companies already using the platform. For ex. imagine that a small company (industry design, for instance) uses Highrise for staying in touch with partners and customers, but they want to add some functionality which is specific for their industry. So they just would hire a feature vendor (negotiating a contract, etc.) to do that. That’s the point.

    You don’t need to be able to produce a spec, because hiring feature vendor to built a feature is much like hiring web design company to build your corporate website.

    Automate feature creation has no value, because it cannot create an app or feature better than a human being can do without any programming at all (for ex. if you need a feature which collects and represents some stats, then you can collect the data manually and represent it as spreadsheet). And if it can, then it’s not automatic feature creator, it’s AI ))))))))))

    Comment by bloodcarter | June 15, 2008 | Reply

  3. I understand the applicability of the feature vendor idea to say Salesforce.com but that is already an EXISTING business, nothing new. Salesforce.com development shops are too many to list here and they’ll do custom development for you (to customize your CRM or even build one from scratch) using the Saleforce.com platform.

    But I think what you’re saying is to give “individuals” (as opposed to a “development shop”) the power to be feature vendors.

    So assuming that down that path, my response would be:

    AI is what AI does 🙂

    Think of declarative UI oriented languages, where you describe the “what” not the “how” and the implementation is generated automatically. Also, think of advances in UI design where components can now be built directly from Photoshop images, with the user able to specify component behavior and data binding all visually.

    There is a paradigm called “prototype based” programming where you start with a programmable “work engine” where you can define/parametrize a generic “work task” and the task can call itself (iterate) thru conditional loops so you could implement any type of “work process” … I built something like that back in ’99 and it led to 80% of active changes to the application being done thru the work engine’s visual IDE rather than via hand coding.

    Take that many steps further and you can see that a spec can be transformed into a working, but human customizable, application, then the human can come and customize it.

    Of course, a programmer will always be needed but they will be needed less than 20% of the time. The idea is to SHIFT production to the consumer.

    If you haven’t noticed (I’m sure you have) there is a huge trend now to move production from producer to consumer so consumer becomes “consumer-producer.”

    Places like Ponoko.com where you are both a consumer and a producer is a great example, and there are/will be many such places and application models where the consumer IS the producer.

    So the consumer is their own feature vendor. Now that’s beyond the long tail. Everyone can produce what they consumer and everyone can trade/barter their productions.

    ….

    (this reply was written in a hurry so consume at your own risk)

    Comment by evolvingtrends | June 15, 2008 | Reply

  4. Hi Marc,

    Thanks for your comments!

    I didn’t know about Salesforce’s devshops…And thanks for pointing out Ponoko.com!

    Now I see what you mean. This is a really complex, long-term goal.

    Could you show what did you build back in ’99? It’s very interesting.

    Comment by bloodcarter | June 16, 2008 | Reply

  5. What I built (with a team of 13 developers and 4 testers!) was an enterprise workflow engine that was part of a commercial product called FlastFlow(tm) (you can probably still google PR links about it) which was targeted at “business analysts” at big telecom companies. It was web-based but the product was owned by the company I worked for so I don’t really have a way of demoing it. It has a run time piece that was web-based and a design piece that was desktop based. All written in Java.

    The closest thing to it is Microsoft’s Workflow Foundation. I am not sure how visually oriented theirs is but the design strikes me as similar, at the high level. They define the same model basically.

    Comment by evolvingtrends | June 16, 2008 | Reply

  6. Hi Marc,

    Thanks for sharing this!

    Comment by bloodcarter | June 17, 2008 | Reply

  7. Vladislav, what you describe closely resembles plugins. If an API is available, anyone can develop them and use them with their software, even if it is web based (hosted at Amazon, for example). While it does bear a mark of sometimes lower quality (especially, when there are thousands of plugins, i.e. WordPress plugins), it does get the job done. I mostly see positive things in open sourcing and providing an API for the software, really.

    Marc,

    As a matter of fact, there are FB 3d party developers that do exactly that: develop applications for customers.

    To successfully build an application – or a website, as Vladislav has pointed out – the company simply needs to have previous experience in building the same type of application or significant expertise to venture into a new app type. It may not be easy to find the company that you get your ideas, but it is still possible (choosing by previously done work and recommendations, for example).

    To order an app, you only need to clearly understand:
    – what you want to achieve with it
    – what you expect your software users to do
    – the reasons they are going to do that

    Creating a spec is a job of a professional, while you only need to get your ideas right, because changing them midway will be costly.

    Frameworks aren’t always required to create software, but they certainly help.

    Everything isn’t hard, if you work hard 🙂 If you expect to get what you want without thinking, well, you’ll need to wait till developers learn to read your mind.

    P.S. Vladislav, I’d suggest getting a domain and hosting your site there on WordPress. Then, you’ll be able to use it however you want, if you decide to add something to your blog or start selling something, for example. Building your personal brand on a free 3d party hosting isn’t really that great.

    Comment by Yura | June 20, 2008 | Reply

  8. Yura,

    Thanks! I really appreciate your comment.
    Your plugins analogy is really smart. There is nothing new with features and plugins. I just said that *nowadays* such APIs is not a mainstream. Most of webapps have just get/set APIs, allowing your to build on top, but not changing something inside. Logic pieces, user experience, etc.

    Hmm…does it look like I’m building my personal brand?))))
    I never thought about it. I just write about things that matters for me.
    Anyway, thanks for your advise. Maybe you’re right.

    P.S. Good luck you with your company, Effectize!
    P.P.S Currently, I live in Novosibirsk too)

    Comment by bloodcarter | June 20, 2008 | Reply

  9. Whatever you are doing, you are building your brand 🙂 Just about right way to write, too.

    Btw, APIs are just entering the mainstream. Woopra, an analytics tool, is being released with an API from the start, for example. The problem is, there’s no single standard (as far as I know), but with Open Social and other standard open APIs, it’ll get better.

    P.S. I may be right, because plenty of other people have found this out the hard way, when they needed a personal blog, even if it touched on a business. Then again, you’ll probably need to think of a better name, but it’s another story.

    What you can try is simply blogging from genomepeople.com/blog. If you decide to have a separate blog, you can always do that. That way, you’ll get the visitors to your project (subscribers, at the moment), too.

    It is just a matter of flexibility of blogging from a domain with more trust, really.

    P.P.S. In addition to my previous comment, companies can start developing plugins for money and it’d be their long tail business. It simply removes any incompatibility/inconsistency problems and adds a revenue stream for the company. Then again, open sourcing is probably a better way to grow software, if the goal isn’t making the most money.

    Comment by Yura | June 20, 2008 | Reply

  10. Thanks Yura! I’m working on this.

    Comment by bloodcarter | June 20, 2008 | Reply


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: