Fog Creek Software
Discussion Board




Opinions on demand-driving product design?



One approach to product design is to try SELLING the product *before* you build it. I.e., a "sales/marketing" directed product development.

This is more likely to develop a product that is sellable, more likely to please the customer, and more likely to generate income.

One of the issues making this collaboration easy. Doing this over the internet seems like a good idea.

Has anyone ever tried this?
Do you think this idea has promise?

For example, I (and I think many people) have a need for one program to organize: projects, task list, and a knowledgebase.  I've looked for one for years and found nothing.  I don't have time to write such a program but I'd be willing to participate in a design discussion.

Perhaps if enough customers could "discuss" and "agree" on a feature set someone might come forward to build it.

(If we come, they will build it).

JoS discussion of the above program.
http://discuss.fogcreek.com/joelonsoftware/default.asp?cmd=show&ixPost=122635&ixReplies=19

The real Entrepreneur
Thursday, March 11, 2004

You can't sell something before you build it.

So, to get around this problem, the common thinking is to build as little a possible, that is still saleable, and then sell that while learning what you did that was right and wrong. Let the early customers and/or rejections dictate the next phase of project life.

There is some danger in this. Not all customers will want the same thing. It takes a lot of work to try to sell to enough people so that your sampling yields meaningful results. You can't simply do what the first paying customer wants, thinking that everybody else will want that. And, most importantly, you need enough money for the very long runway it takes for a process like this, which almost assuredly means very few people making almost no money.

Brad Wilson (dotnetguy.techieswithcats.com)
Thursday, March 11, 2004

"Perhaps if enough customers could "discuss" and "agree" on a feature set someone might come forward to build it.

(If we come, they will build it)."

Why don't you build it?  Or why doesn't the JOS community build it?  I've often wondered how a company comprised of just the JOS community would operate.  Part of me wants to say that it would be a total success for the simple fact of being smart and getting things done.  The other part of me says too many Type A personalities getting in the way of doing anything productive.  I bet we'd all just sit around and talk about the best way to do things and not get anything done.

If you spec it out.  I'll build it.

Elephant
Thursday, March 11, 2004

> You can't sell something before you build it

*cough*Microsoft*cough*


Thursday, March 11, 2004


Problems mentioned above (and responses)

1. You can't sell what doesn't exist.

You can get an advance order.  The goal isn't to make money on the sale (immediately). The goal is to get SOLID marke research. It's one thing to get someone to  SAY they want something, but another to get them to (at least tacitly) agree to BUY it if it has ABC. 

Then you turn them into a beta tester.  I worked for a company that did exactly that.

We sold one bank of the idea of a video kiosk. They paid for a prototype. They were a *partner* who paid for the product and was a beta tester. He then sold more companies on the idea, with each company adding some feature that they wanted (and that others might want).


Here's a whole inc magazine article about that. (GREAT article, BTW)
http://www.inc.com/magazine/20020201/23855.html

2.  Can't get customers to agree on features.

Agreed, this is difficult. That's why I was asking about how one might do this online.

I think that if you had lots of SMART early adopter programmer types,  you might be able to reach consensus.  Some sort of multi voting, with discussion.

The HARDEST part of creating a successful sellable product is knowing what the minimum feature set is.


Elephant - are you really serious about building the product if I can spec it out?

The real Entrepreneur
Thursday, March 11, 2004

Send me an email with details, I'll let you know if I'm serious after that.

Elephant
Thursday, March 11, 2004

If you are serious about a JoS thing, count me in. The concept could absolutely fail, but it could also be fun trying.

Kind of sourceforge-ish, and yet without all the fanfare.

Aussie Chick
Thursday, March 11, 2004

I was dead serious about that Aussie Chick.  Send me an email too.

Elephant
Thursday, March 11, 2004

I work at a manufacturer of household goods.  There's a common flip-flop in our industry between the textbook definitions of sales-driven (sell it, then build it) and marketing driven (build it, then sell it) orientations.

One of the problems inherent in sales-driven organizations is that the market doesn't provide *coherent* direction over the long term: the aggregate of all your market data telling you what to sell will be a confused, reactionary mess.  So you'll sell lots, but your overall product base will have little rationality to it, and with that comes increased costs for managing the business and sales.

The marketing driven approach tries to correct that with strong internal design of a product line--the drawback, of course, being that you're not selling people what they asked for, but what you think they need.  That's great if you're prescient, but you're really risking warehouses full of dead inventory.

In software terms (and in terms of the project being born here), that means that the risk is a confused spec without long term direction.  Always scratching the immediate itch will make you end up with a kitchen sink application that has no clear personality, theme, style, or principle, that becomes a maintenance nightmare down the road when you're trying to add yet another customer request.

So, Elephant, I'd urge you to do your best, when taking the specs you receive, to find some underlying principle or design goal in what you're building.  You don't have to even tell your customers, just have it available so that the project is always, from the inside, clearly organized.

Justin Johnson
Thursday, March 11, 2004


USE JOS AS A FOCUS GROUP

A commercial version of Sourceforge

I've always liked the idea of the open source collaborative projects at Sourceforge, etc.

BUT.... they're never oriented to making $$ so the projects tend to stall because they're hobbies.

Would it be possible to collaborate on the *requirements* and design of a sellable  product, using JoS as a focus group.

The JoS folks are early adopters who are able to envision a product that doesn't exist. A perfect group for market reserach.

So, rather than collaborating, for free, on the PROGRAMMING, collaborate, for free, on the specification.  Then whoever writes it gets to make the profit, etc.

The real Entrepreneur
Thursday, March 11, 2004

"I'd urge you to do your best, when taking the specs you receive, to find some underlying principle or design goal in what you're building."

GOOD advice.  Use the market feedback as SUGGESTIONS. 

I view customer requirements as  needles through which you're trying to thread a program.

The goal is to shift the needles around enough and wiggle the thread enough to get as many needles threaded as you can with the minimum amount of program. (I.e., stretch X program into as many needles as you can)

The real Entrepreneur
Thursday, March 11, 2004

Argghhh... Aussie chick beat me to the punch on the Source forge comparison.


Actually, I wrote it before she did, but I was interrupted by a sales call <g>.

Well.... great minds run in the same path.... or fools think alike :-)

The real Entrepreneur
Thursday, March 11, 2004

SUGGESTION FOR THIS PROJECT

We need to have some way to :

1. List goals
2. List features to attain those goals
3. Discuss them via the internet.
4. Do this as easily as possible (without a lot of overhead).  We don't want the discussion tool to interfere with the discussion.

I like JoS's board, but I think it's too hard to keep track of something ongoing like this.

Suggestions?
(Simple discussion board, each feature is one root topic?)

The real Entrepreneur
Thursday, March 11, 2004

Hands up everyone who is willing to actually attempt to sell this product.  On the phone, or in person.

Anyone?

If no one volunteers, then you can all just forget about making any money off this venture.  Treat it like a fun hobby if you want, but don't kid yourself there will be any money in  it.

whatever
Thursday, March 11, 2004

I will.


If we can create a product that I think will sell well and that folks on JoS (my test market) are interested in, then I'd be willing to set up a website, online ordering, and credit card payments, etc.


But that's a little way down the road.  It think we should take this a step at a time and regard it as a hobby until we see what's being produced.

HOWEVER, I think the person(s) writing the program should have ownership of the source code and THEY get to decide what's done with thier program. They can sell it, give it away free, or I can market it for them (for a %) or help THEM market it (for free or for a smaller %).


BTW, I think the market for this is just emerging.  This is the perfect setup for some one man company to  own.  A big company won't be able to maintain the cash burn rate to wait for this to be a million dollar niche. But a small company can survive on $100k a year in sales.

But take it a step at a time and HAVE FUN WITH THIS to start with.

The real Entrepreneur
Thursday, March 11, 2004

Real Entrepreneur, have you looked at the open source project, Chandler?

http://www.osafoundation.org/Chandler_Compelling_Vision.htm

From the article:

    Chandler lets the user keep track of lots of
    concurrent, ongoing activities. Managing activity
    extending over time requires the ability to collect
    just the right sets of related items.

It doesn't do exactly what you want, but it's likely to be sufficiently extensible to be  made to meet your needs.

Burninator
Thursday, March 11, 2004

If I remember correctly, Joel commented on the Chandler project in one of his articles and it wasn't positive. I think he said they created too much hype and now are trying to catch up with it all... Not that it's a bad thing if you like the product.

entell
Thursday, March 11, 2004

Burninator,

Yes, I'm familiar with Chandler. I tried it once before and couldn't get it to run. Just tried it again. One two XP computers. Same result.  This is either an indictment of Chandler or of using Python (which introduces extra complexty)

I don't know of anyone who's actually tried Chandler.

It certainly *seems* like it might be useable.

Honestly, my *impression* of chandler, which I first heard about a couple of years ago I think, it that it's a big beauracracy.  They have all of the elements of a real software company. Except for...ummm... a product. 

Thier website has been around for a couple of years (since mid 2002) and still not product I can get to run.

I started my company the opposite: none of of the overhead of a software company EXCEPT for the product.



A lot of fluff but nothing actually usable seems to come out of it.  I think the best innovative products come from small groups of smart people.  (I know Mitch Kapoor is probably a smart guy).

The real Entrepreneur
Thursday, March 11, 2004

"So, rather than collaborating, for free, on the PROGRAMMING, collaborate, for free, on the specification.  Then whoever writes it gets to make the profit, etc. "

This is an interesting point.  It's often mentioned here that ideas are cheap, it's implementation that's hard.  Well, we can throw around the spec for free, and then get down to work.

Assuming, of course, people actually do this.

-Rich

Rich
Thursday, March 11, 2004

"Would it be possible to collaborate on the *requirements* and design of a sellable  product, using JoS as a focus group."

Unless you're making something for programmers, I doubt you want JoS to be your focus group.


"I like JoS's board, but I think it's too hard to keep track of something ongoing like this."

It's kinda out of scope, don't ya think?

Rich
Thursday, March 11, 2004

>But that's a little way down the road.  It think we should take this a step at a time and regard it as a hobby until we see what's being produced.

I think it is a cool idea.
And I agree, the thing that has always stopped me from ever joining a sourceforge product was that is was only a hobby, ie if I am putting in effect I want something in return.

But JoS does, IMO, have something that SF doesn't, it is as far better site for networking, perhaps this is a good enough motivation for now.

JoS (ie one forum with topics that disappear at the end of the week) is a shocking place to run a project like this from, but doesn't it just sound fun! There wouldn't even be a central place to store file. Lovely. It would  be an incredible show of teamwork.

Aussie Chick
Thursday, March 11, 2004

"JoS (ie one forum with topics that disappear at the end of the week) is a shocking place to run a project like this from"


Hmmm.. a "sliding window" would help keep the project moving, eh? 

The real Entrepreneur
Thursday, March 11, 2004

good point. *grin*

Aussie Chick
Thursday, March 11, 2004

Nb. If this gets done (or even attempted) someone better write an essay on the process, it could make a good read....bunch of anonymous people attempt to design and create a brilliant niche piece of software, revolving around a simple message board with disappearing messages, and no shared file space. This is not sourceforge, they could make money, they don't know who would get the money, and the whole thing could dissolve into a bunch of lawsuits as the money is divvied up and identities are revealed....but boy they had fun in the beginning.

Aussie Chick
Thursday, March 11, 2004

GET TOGETHER TO WORK ON THIS

Ok, if you're interested in working on this, either spec'ing or coding,  email me.  Let me know if you're interested in the spec'ing or in the actual programming.

I have one programmer who wants to write this program.  He'll keep all rights to whatever he creates.

If we have enough people interested in spec'ing the software, then I'll set up an online message board where we can list all the features, etc.

Thus begins the great experiment with collaborative specification......

The real Entrepreneur
Thursday, March 11, 2004

I would prefer that you didn't clutter JOS with this project. I don't come here to design products for other people. Set up your own off-site discussion forum.

x
Thursday, March 11, 2004

4 people including myself interested.  I do my own hosting, so I'll put up a message board (from the ones suggested) and we can continue the conversation there.  I'll let you know when I've gotten it up.

Elephant
Thursday, March 11, 2004

I've posted some completely awful board software for the discussion to continue offline from JOS.  If you have a desire to participate or want to know where I've posted it, shoot me an email and I'll set you up.  I dog food everything I do, so I don't want to post the link here to generate any unnecessary traffic.

Elephant
Thursday, March 11, 2004

Yes, Chandler is playing catchup to the hype. No it's not even near a 1.0 release.

Mitch Kapor 2/26/2004:

"This week we released Chandler 0.3. The team did a great great job; my hat is off to them. If we were building a house, we'd be at the point of having a foundation, the framing, and the plumbing roughed in. Now on to the interior partitions and finishes."


and this:

..."using Python (which introduces extra complexty)"

s/b

...using Python (which *reduces* complexty)

fool for python
Thursday, March 11, 2004

How are we doing with the spec and the coding? I just sold a copy.

Just me (Sir to you)
Friday, March 12, 2004

I very strongly suggest that you target an extremely specific market that has specific needs.  Otherwise your customer driven spec will go berserk.

Also, just because a few people scream loudly about needing a given feature doesn't mean you should automatically do it.

Personal anecdote (yeah, I'm prone to those):  I was on a committee that was supposed to be attemptng to standardize data formats across the pharmaceutical industry.  We were moving along the requirements phase pretty well, but there were a few requirements that popped up from a rep from one company that's pretty big.  There weren't many objections, so most of them we kept, but every now and then we got "Well, we absolutely have to have this in the data, and we have to have it this way."  Sometimes this directly conflicted with another company's needs, but we worked around it mostly.  It wasn't until this had been going on for nearly two years when I thought to ask the entire group "So, once we get this done, we have committment from your company to start using this, right?"  It turned out that the larger companies who had the most sway in the group were the least likely to actually use the thing.  The smaller companies with less invested in the old way of doing things could rapidly switch,. but they had less of a say in how it got done.

So...just because someone apparently big and important screams for a feature, it doesn't mean they will buy it.  When confronted, they may just say, "No, we already have a lot of money invested in our internally developed system and won't be switching."  They are not going to be a client, so their input isn't as important as they think.  Sure, they may have solved that problem already, so they may have some useful input, but not as much as they want you to believe.

Aaron F Stanton
Friday, March 12, 2004

"How are we doing with the spec and the coding? I just sold a copy."

Sorry for lowering the SNR here, but that was hilarious!

Rich
Friday, March 12, 2004

Aron,

The comments you made are another good reason to ship early and often, with an evoving product.

You never know exactly what features people will want until they actually put down money. So selling a version 1.0 is a great way to find out what people are willing to buy. If they all ask for more "reporting features", then you expand that. If they all ask for more "export and import features" then you expand that. If they hate the UI, etc. etc.


Joel has already written about this ("Shipping is a feature") so I'll let it rest.

Mr. Analogy  (formerly The real Entrepreneur)
Sunday, March 14, 2004

*  Recent Topics

*  Fog Creek Home