12 replies
by techrep
I've found myself changing colours etc to please one of the client's staff, only to have others annoyed. By allowing a modicum of customization, my impression is that user's feel some sense of control over the tool they use all day at their jobs. To measure this desire in a generic way, walk around any given office and see how many people have left the default Windows colour scheme in tact.
vlad@agt.net
by vbhonda
There is a very good chance that I would still be employed if I had not fallen victim to this. I built two projects and crashed into the issue of gold plating. I built powerful applications with lots of features. One had to be stripped down as too cumbersome for the user. And it ultimately was scrapped because of a fatal bug. That probably would have been caught had the application been much simpler.
Another problem I encountered the extra features required extra programming time. My employer ended up viewing me as unproductive. I was layed off under the premise of work was slowing down. Funny thing, there was a new programmer at my desk the next week.
A painful lesson for an entry level programmer.
by dr ford
I don't see this quite so one sided. I agree Gold Plating is a very real project issue to be managed, but it raises at least two other issues.
One, that Project Managers appear arrogant or perhaps like a jobsworth ruling out anything not in the original specification, which doesn't lead to good project relationships and open dialogue.
Two, the assumption that we know everything at the outset, and nothing is learned during the journey (either issues or opportunities), which in my experience isfallacious.
It seems to me that Gold Plating needs to be directed into a disciplined Change Control process, that encourages ideas and doesn't "crush" the people involved. Leading to Projects that add value to the product, the people and the enterprise.
by talentonloan
by marko.hurst
Gold plating can and does get out of hand, or we wouldn't be discussing it. My experience has shown that having project managers (or equivalent) that are familiar with full cycle development is one of the largest challenges. While the basic principles of PM are carried forward, many things are unique to software & web development and should be treated as such. This is often a problem with in-house projects.
I cannot stress this enough, Know Your Audience! Allot of "Gold Plating" is eliminated by audience analysis, usability testing, and the equivalent to a UI Engineer/Architect. It doesn't matter how much you or even the clients like or want something, if the end-user cannot use it the project has failed. Also, often times 'business requirements' has a way of being placed in front of usability and that too is often a road to disaster.
As to personal preferences (colors, etc), some places it's permissible others it's not. But there are some good ways to create and manage those if it is necessary, CSS, portlets, widgets, etc.
by dbrodt
Sometimes the client loves the result of gold plating, especially if it's a cool feature; after all, it's more than he asked for at the same price. As project manager, though, I'm thinking, "This is one more thing we'll have to support when it breaks."
I find the most common reason for programmers to build functionality that was never requested into a project is because they can. They love to be creative and incorporate cool code they've come across on the Internet or played with at home. I don't discourage this, but I try to get across that the deadly sin is to present the new addition as a fait accompli; that is, to incorporate it without going through change control. It only takes a couple of times of making a programmer tear out 8 hours of code and rebuild a module to think twice next time. Even if this means we go over budget.
by mrhemphill
The desire for added features is simply human nature - especially among those who have less invested in the project (and the resulting work).
What I've done is to maintain a strict definition of what the ultimate goal of the project is, and then apply additional requests to it. If the new requests are valid within the scope and target frame, fine. If not, then it will be offered on the next rev - by which time it's usually forgotten (if it's fluff), or added (if it's valid).
If it really needs to make it into the current rev, and is outside the definition, then we have to reexamine the definition, and possibly change the scope, length, time frame and cost of the project.
YW
by louisavalos
Having developed many a project this is the best term yet for over delivering a project. Too many times we have over engineered a project based on the nice to haves rather the essentials. Considering that the 80/20 rules applies to software applications that is 80% of the users will use less than 20% of an application we insist on bullet proofing that other 80%. Users really appreciate applications that do the job they are suppose to do not allow them customize like the crazy their page or view the latest weather in Aspen. Users will find an easier application to use so they will abandon your application if it contains too much gold plating.
LA
by alanhoweusa
by akumar
I have worked on both product developments carried out inhouse and on projects awarded by external clients - i feel that with inhouse
projects the gold plating is necessary and is a part of the cyclic development effort that you make on the product. eg a simple server product expanded to support dynamic load balacing / load sensing etc
But the other case when you are developing to confirm to a deadline and probably assoicated payments , it pays to confirm to standards and keep the whole effort proffessionally upto standadrd look and feels etc but apart from that
feature wise gold plating is an absolute no no
by leeschoff
I've run up against a database design that includes tables, table organization, and layout that exceed the requirements as articulated, but which are strongly felt by the designer to be required in later phases even though the customer (a large gov't agency) may not know it yet. The agency data model includes much of this stuff in the total model, but the customer is one particular department of the agency and the project focuses on a small subset of the overall model.
Getting the customer to sit down and review the nitty gritty details is proving very difficult to do and of course there is a tight schedule.
I'm looking for inputs on how to resolve this. Internal discussions without the customer center on the abstract future use scenarios and the strident claim that it is much better to add this stuff in the original design even if it is not populated or coded for than to have to add it in at a later phase when "we knew it had to be there all along".
Any suggestions?
This discussion is closed to new replies.
Trending Web Development posts
Sort by Trending