-11618

072424

Commentary: Blogs become key tools for developers

	Title	Becker	Ness
	Secrecy	Secrecy has long been a hallmark of the software development process: Let too many people know too much about what you're working on too early, and somebody might steal your ideas.	Surely a correct description, and true of more than just software.
	The Open Door	But a growing array of big-name software developers are finding they can make better software if they leave the doors open, by sharing information with potential customers from the start and incorporating their feedback into development decisions. While developers of games software have used this method for years, business software makers are now also catching on.	This seems to me to assume the case that needs to be proved. The notion that some `leave doors open' is surely true today. But, whether it leads to `better software' or not surely is not proven. So far only relatively little of the software that has been built this way is available, and virtually none of it has yet been around for long enough to make me comfortable that it is a `better way'.
	Blogs	Web logs (commonly known as "blogs"), message boards and other online forums are becoming increasingly important vehicles for developers to attract customers--and development talent--well before an application even enters the beta stage.	I guess so.
	Kapor	Mitch Kapor, founder of software pioneer Lotus and creator of its breakthrough 1-2-3 spreadsheet program, started a development blog early on in his quest to build a smarter personal information manager. He said the blog has been a vital conduit for him to communicate with users about the project and to solicit their ideas.	OK. Kapor probably believes this, and he probably has said it. So I better not argue.
	Open Feedback	"Some of the world's smartest software people are interested in this project and communicating with me," Kapor said. "The more open feedback there is, the better we can incorporate those ideas into the product."	And some of the `world's smartest software people' are not communicating with him. Since we have only a bland statement, we don't know which group is more important. And whether the ideas are worth incorporating remains to be seen.
	Updating Users	Kapor uses the blog to update potential users of the manager on development progress and his ideas. Readers are invited to share their views via e-mail, private and public chats and other means.	True.
	Communicating	"It was a conscious plan of communicating to people about what we're doing," he said. "It's part of a long-term process of building a user community. Every process has its advantages and its disadvantages, but if you have an open process, you can get much better feedback, and you get stimulated by new ideas."	And you will have to waste a lot of time responding to bad ideas, too. Telling the difference between the good and bad feedback is the trick. Always remember when you ask what `the public' thinks, you're not just going to hear from the smart ones.
	Bricklin	Blogging has also become an important part of the development process for Dan Bricklin as he works on the SMBmeta specification, his idea for a giant online business directory that would open the Web more to small and medium-sized businesses. Bricklin, co-inventor of VisiCalc, the first spreadsheet program, said his blog and other public communication conduits extend the possibilities for user feedback beyond beta testing, the traditional approach in which developers send early versions of a program to a select group of testers.	I guess I'd agree. They do `extend' possibilities. But whether they do so in a useful way still remains to be seen.
	Beta Testers	"I remember trying really hard to find beta testers by saving all the business cards I collected," said blogging pioneer Bricklin. "We had to find people, call them and beg them to be beta testers and mail them the software. We had to call them every week to see how they were doing.	This makes me worry a bit. There's a difference between going out and getting beta testers, and being willing to take those who self-select. I have no reason to believe that a self-selected group will lead to the same results that we might get from a group we had to solicit, and I certainly have no reason to believe that the results obtained from such a group would be better.
	Blogs as a Tool	"A Web log is a simple, inexpensive tool to get communication going and do it so much more efficiently," Bricklin continued. "It's unbelievable how much wider an area you get feedback from. What Web logs do is let there be more serendipity of ideas. I'm already finding bugs and looking at ways to do a better job based on what people have shared with me after reading the blog."	Wait until you develop a real wide audience. Then the vast majority of the feedback you get will be, well dare I say it `average'. Just like the population of people reading your, or anyone else's, blogs. Early adopters are `a different breed' (by definition) and you should be very careful about generalizing about any future results.
	Games	While business software makers are starting to pick up on the value of early public feedback, games developers have known about it for some time. Traditional offline games still adhere to a typical pattern of prerelease secrecy and limited betas, but developers of multiplayer online games have learned to court potential customers early.	Games developers and their audiences strike me as quite an atypical group of people. I wouldn't want to generalize much from their experience or process without thinking about it in considerable depth. And there's no indication that is going on here.
	The Gaming Community	It takes a community	Sort of like `... it takes a village...' But the whole relevance of this discussion about the gaming community to the situation here is problematical.
	EverQuest	The social nature of online games makes it essential to build a user community early on, said Scott McDaniel, vice president of marketing for Sony Online Entertainment, publisher of leading online game "EverQuest."	Before making any judgements about `games' I'd like to have some notion of what the half-life of the games we are discussing is. It may be essential to build a user community in the early days of a game's life, but that would be particularly true if the life was short and there weren't many `later days'.
	Community Priority	"It's hard to get a community going if you don't make that a priority from the beginning," McDaniel said. "You want to make sure people have the information they need to get excited and evangelize the game...We do a lot of advertising and promotion, but we found that the No. 1 reason people sign up for 'EverQuest' is word of mouth--a friend told them to try it."	While this may be true of games, I'm not sure that I'd stretch the point to apply to any other particular domains without some research evidence that such a generalization is warrented. The game world is a pretty `odd' world, and I'm not convinced that other behavior is likely to parallel it in any substantial way.
	Star Wars	"Star Wars Galaxies," the upcoming online role-playing game to be published by Sony Online Entertainment, has had a Web site full of active user forums for almost two years--well before the earliest stages of beta testing. Developers use the site to update fans on the progress of the game, to conduct open chat sessions with readers and to solicit feedback through discussion groups and other forums.	This would be an interesting point if there was any evidence that the kind of feedback obtained this way was (a) different; and (b) better than that obtained by other more conventional methods. I don't know if such evidence exists or not, but it most certainly isn't presented here in any case.
	Galaxies	McDaniel said early feedback from fans has played a significant role in shaping the development of "Galaxies." "We started off asking really basic questions. The answers led us in directions we hadn't thought of," McDaniel said.	And how were these questions asked? The role of the Web in any of this is completely unclear.
	Earth and Beyond	Rade Stojsavljevic, senior development director for Westwood Studios, which creates games for leading publisher Electronic Arts, said the development team for "Earth & Beyond," the studio's new intergalactic role-playing game, began soliciting feedback well before there was any test software to send around.	And here too. Was the Web involved somehow? Was this involvement successful? How did they measure success?
	Feedback	"We spent a lot of time in there when the community was really small, soliciting feedback from players," he said. "We hired a full-time community manager early on--her job was to filter feedback to the development team."	Another uninformative remark. The usual questions apply.
	Structural Changes	Early feedback resulted in major structural changes in the game, Stojsavljevic said. The initial concept for the game didn't include "avatars," or display appearances, to represent individual characters--players instead were supposed to focus on customizing their spaceships.	Enough about games. The stories that have been presented may have some amusement value to game players, but they surely contain nothing useful with regard to design issues.
	What to Think About	"Our idea was that you wouldn't spend any time thinking about who was in your spaceship," he said. "The thing we didn't realize in the beginning is that it's hard to get attached to a piece of metal. The user groups were very clear and unanimous and made a really good point about that. After that, we spent a lot of time developing an avatar system."	A little colloquial story telling, but no useful evidence at all.
	Filtering	Filtering the chat	Apparenly the user ideas obtained need some `filtering'.
	Liabilities	Inviting the world into your software project includes some liabilities. It can be a lot of work to sieve through discussion group postings, e-mail messages and other submissions to cull worthy ideas. Westwood has five full-time community development specialists for "Earth & Beyond," who spend much of their time sifting through user input.	It would be interesting to have some clue about whether this sifting activity consumes much time. It would surprise me if it took more energy to sift through bad ideas than to generate good ones de novo.
	Funnel Ideas	"You can easily get a mess," Stojsavljevic said. "I think it's critical to have someone there to dig, find out which ideas the most people are talking about and funnel those to the right people on the development team. At the height of the development cycle on this game, we had 150 people working on it. It's just impossible to have everybody know what's going on overall."	And this is pretty much begging the issue. We replace the problem of discovery about our original problem with the one of setting up the proper channels for handling it. It is not at all clear that this is done at a profit.
	Unproductive Threads	Developers also need to be able to pull the plug on unproductive or exhausted discussion threads. Otherwise, projects can bog down in a paralysis of ideas.	Yes. It's very easy to make the mistake that user feedback will be `good feedback'. While some of it is---particularly in the early stages of a process---quite quickly lots of the feedback becomes useless and distractive. This can then slow down progress on the project to a crawl.
	Anti-Authoritarian	"You have to be prepared to make some clear statement--we're doing this, we're not doing that, or this issue has been settled, and here's the reason why," Lotus founder Kapor said. "If you don't take stands and communicate them, the discussion never ends. And you need to do that in a way that respects the community. If its feels authoritarian, that doesn't build a good dynamic."	Well, that's just the problem, isn't it. Being authoritarian without being authoritarian is tough sledding, and some people just can't pull it off. There is no clue here about how any group that doesn't posess this kind of capability would end up getting it.
	Thick Skin	It also helps if the developer has a thick skin--to withstand correspondents who express their ideas in harsh language--and enough self-awareness to be able to admit they might be wrong.	Thick skin is useful in lots of different aspects of life. I'm not convinced that software development is at all remarkable in this respect.
	Humility	"Being a software developer, you need a combination of ego and humility," Bricklin said. "There are so many bugs and dead ends, you need the ego to keep going. But you need the humility to learn and make changes."	Sounds like `normal life' to me.
	Success	With the right attitude and attention, however, public participation can make a software project a success before it leaves the developer's cubicle.	Some products sell well from their first day on the shelf. Others never move. That's life in the marketing business.
	Early Adopters	"If you do this right, you've got early adopters, you've got evangelists, you've got a lot of early support," Kapor said. "The train has left the station and is gathering steam before you do a final release."	This may be particularly important if the systems you are building have a very short life. And lots of the one's discussed here, Kapor's among them, are quite likely to face that fate.