There's a right way to ax your products - and many wrong ones.

By Verne Kopytoff

FORTUNE -- A company announces plans to kill a once-popular product. Executives lay the blame on slumping sales and changing priorities. Such decisions are a routine part of doing business. They also require ample forethought to avoid alienating customers, violating contracts, and racking up extra costs.

Google's (GOOG) announcement last week that it would shutter Reader, a service for creating a feed of blog posts and news articles, is a prime example. To minimize the impact on users, Google said it would wait more than three months before pulling the plug. During that time, users can find an alternative service. Those who want to export their data from Reader can do so.

Fans still went online to criticize Google's decision and ask that it be rolled back. But the company's outreach at least gave it some cover.

In deciding the fate of their products, companies generally follow a standard process. Analyzing important metrics like revenue, profit, and customer engagement against specific goals is routine. "Product life cycle analysis" or, more informally, "kill analysis" are just some of the names for the job. The point is to come up with a dispassionate, unbiased recommendation about how to proceed.

MORE: The tech IPO drought is only getting worse

Calculating how much a product costs a company can be complex. Any attempt must factor in employee salaries across a number of departments like sales and customer support. Bandwidth costs can be sizeable for an online product. So can upgrades, fixing bugs, and making products compatible with new technology such as the latest tablets. Basing decisions on emotions and internal corporate politics is dangerous. But it is all too common. A founder may feel an attachment to a product and resist shutting it down. Product managers may fear that admitting failure will hurt their careers.

"Without a doubt, when you have products that are the babies of the founder -- that come with an emotional attachment -- those are the ones you really have to come in having done your homework about why you're making the decision," says Chris Brown, vice president of product management for Apartments.com, an online apartment rental listing service that is owned by online classified giant, Classified Ventures.

In most cases, "life cycle" reviews show that sustaining a product is fine. In others, the analysis may conclude that the company needs to ramp up investment for new or improved features.

Products deemed outright failures are, in fact, relatively easy to deal with. Big money losers have few defenders. Things get more complicated with products that have gained some customers but are nonetheless stagnant or in decline. Many executives dislike sacrificing short-term revenue, even if it is for the long-term good of the company. "Boy is it tough to tell a CEO that you want to take revenue off of the table," says Brown, who has shut down a number of products and features over the years.

MORE: Silicon Valley looks for love, at $50,000 a pop

A shift in corporate strategy is another common rationale. A product that was once central to a business may no longer be after a few years. Weighing the competitive landscape and market needs is also important. "If we're doing something that's so far removed from our strategy as a business, we better have a good reason," says Kirsten Butzow, an instructor with Pragmatic Marketing, an executive training and consulting firm in Scottsdale, Ariz.

One option is to pull back investment and let the product die a slow, natural death while using any profits to fund other projects. But leaving a product limping along can also confuse customers and distract employees who could be working on the next big thing.

Pulling the plug more quickly is the other alternative. Quick is a relative term, however, because of all the necessary considerations. Companies may have agreed in customer contracts to supply a product for years into the future. Regulatory issues may also apply. Lawyers need to get involved early in the process and, if it's a key product, so should the marketing team and finance to make billing changes and adjust employee sales goals.

MORE: Electronic Arts CEO resigns abruptly

Giving customers ample notice is critical. A software company, for example, may promise to continue to support its products for a year, but that it will no longer upgrade them. Losing customers is always a possibility. To minimize defections, a company can offer customers free upgrades to any successor product and help them make the switch. If getting out the business completely, a company can suggest alternatives.

Of course, companies don't necessary have to shut down unwanted products and can instead try to sell them to another business. The corporate world is littered with examples. EBay (EBAY), for example, agreed to acquire Skype, the online calling service, in 2005 for $2.6 billion. After failing to do much with the acquisition, eBay decided in 2009 to unload most of its stake in Skype to a group of investors that later resold the business to Microsoft (MSFT).

Butzow, who has shut down and sold a number of products over the years while an executive at various technology companies like Fujitsu and Pearson, says, "Just because something is not working for your business doesn't mean it's not highly applicable to another."