Anyone who has visited Samsung’s office towers in Seoul, South Korea, will not be surprised to know that the Lee family – the dynasty that controls the conglomerate – runs a tight ship. The three towers, which dominate the landscape of the Gangnam district, were built to consolidate many of the activities of the firm; their imposing presence is emblematic of the company’s hierarchal culture. Inside, elaborate security procedures, long working hours, and deference to senior managers are all in plain view. Also in plain view is the firm’s recent decision to issue a massive recall of and then terminate the Galaxy Note 7.

Killing products isn’t easy. Engineers and managers toil for months, often years, to conceive, develop, and launch new products. They invest significant resources in research, marketing, and distribution. So to see those products go down in flames (literally, as in the case of Samsung’s Galaxy Note 7), often creates a very difficult choice: try to improve the quality and support for products or terminate them.

In our five-year study of global handset makers, we found that most firms pull products when they turn out to be obvious disasters such as the Note 7. What proves trickier is to yank lackluster offerings with sales that mildly disappoint — products that perform poorly but are not clear-cut cases for withdrawal. In such cases, managers often hang on in the hope of an eventual uptake. More often than not the result is a confusing proliferation of middling offerings.

Our research suggests that given the challenges of letting go of these weak products, companies like Samsung put important product decisions – like exit – in the hands of more senior managers. While centralizing decision making in this way can often overburden managers and be a drag on the creative freedom of designers, it has one clear advantage: it speeds the termination of products. Our research suggests that, on average, more centralized decision making structures are more likely to pull poorly performing products from the market.   

Centralizing product termination decisions does three things: 1) Focuses attention on the entire portfolio; 2) Manages the ripple effect that exit has on the firm’s ecosystem; and 3) Mitigates potential disagreements and delays owing to internal politics. Thus, centralized firms more quickly rid themselves of unsuccessful products – almost twice as fast as their more decentralized peers. These findings are especially useful for any companies facing a large portfolio of products, continuous technological change, short life cycles, and high levels of product obsolescence.

Know the big picture

Centralized structures speed product termination because they aggregate information about the whole portfolio at the top. Companies in our sample, which included all the major mobile handset manufacturers, had an average of 24 products in a major market at any given time. When you consider the number of all models across all regions, portfolios can become quite large.

However, lower level managers charged with managing one or two products, are not likely to have visibility across the entire portfolio. They are focused on a narrow set of products under their purview. Decision makers at higher levels are able to get a better sense of how well or poorly products perform relative to other products. Therefore, decentralized firms may have delegated termination decisions to a product manager, to whom the underperforming product might look satisfactory, centralized firms are more sanguine and ready to pull.

Also, the difference in responsibility and authority given to product and senior managers means they respond to performance problems differently. Lower-level managers may be less willing to give up on their products or give excess resources to a competing product team. Instead they would prefer to maintain the products and make adjustments to promotions or advertising. Senior managers, who have greater authority to move around resources, are more likely to withdraw support for flops and re-direct resources to winners.

For example, right after Sanja Jha was appointed head of Motorola’s mobile devices unit in 2008, he immediately took control of the portfolio, cutting the firm’s entire Symbian product line and focusing the firm on fewer products and the Android platform.

Terminating products such as mobile devices is a complex process that involves an ecosystem of design companies, carriers, applications developers, software companies, content providers, and manufacturers.  Product termination requires the adjustment of a range of interdependent activities such as product road maps, factory schedules, and supply chains. The presence of so many external factors creates the need for communication and coordination internally across functions and product managers, and with outside carriers and suppliers.

This makes it tricky for timing termination correctly. Firms might leave products on the market well past their prime. In industries characterized by short product life cycles and rapid technological change moving faster is generally better. Not only does this ensure “old” products are cleared out, but it makes room for new products, ensuring that managers can give them enough time and attention.