A year ago, my partner and I moved to Seattle. I moved from New York City, and she moved from Grinnell. It was a long-distance move, and we both wound up having to buy a lot of things again once we arrived in the area. While we were shopping in Target, a shiny new Brita pitcher caught my eye — the Stream.
Unlike my old Brita, the Stream’s standout feature is its “filter-on-pour” design. Rather than the old “filter-on-fill” method that restricted the fill speed to the filter’s processing rate, the Stream fills as fast as your faucet can go. In principle, on-demand filtering is a plus for the owner (who can fill the pitcher with less time and hassle), and the drinker (who gets their water on-time, and can drink any quantity without having to worry about a tedious refill). In practice, real-world constraints make mincemeat out of this promise.
The Brita filter that is employed in the Stream is not fundamentally different from the classic filters Brita has sold for years. They incorporate active carbon and ion filters to scrub impurities out of water as it passes through, with a throughput lower than that of many kitchen faucets. This works against Brita both ways: whether the water is flowing in or out, it can’t move faster than the filter can handle. And when you’re holding a pitcher over an expectant glass, the delay can be excruciating.
Many products suffer from an apparent disconnect between the design and engineering teams. The Stream looks to have been born of such a schism: on the one hand, the designers clearly had a vision of a hassle-free improvement over the basic Brita model. On the other hand, the engineers did not have the technology to match that vision, outside of changing the appearance and shape of the filter and a relatively leak-proof lid (which, to be clear, is really quite nice). The result is a product that offers a bad user experience at a higher price point; not ideal at the best of times, and definitely not helpful to anyone looking to replace their old Britas (we wound up buying a “standard” Brita and using it for most of our needs).
Given the practical constraints of Brita’s filter system, the design team should have more carefully considered relevant user stories. On the one hand, people need to fill their pitchers with water; while tedious, the Stream has a 10-cup capacity and is unlikely to need to be refilled after each use. On the other hand, the basic use case for a pitcher is to pour a liquid when needed. The larger the capacity of the pitcher, the larger the ratio of pours:refills becomes. Efficiency improvements in the most common use cases have a disproportionate impact on the ultimate user experience; users are apt to remember a frequent annoyance.
The Stream’s flow rate starts out alright, but rapidly degrades to an almost unacceptable trickle. If there is insufficient water in the reservoir, the only way to maintain a decent flow is to tilt the pitcher at increasingly steep angles. While this is true for any pitcher, the Stream’s throughput declines far faster because water must force its way through the filter before being dispensed. In contrast, “standard” pitchers provide a uniform rate of flow so long as the spout is fully covered with water.
The end result is that the Stream pours slowly for half or more of each fill cycle, making for at least five cups’ worth of standing with increasing frustration as water trickles into a glass. This is one case where the design is half-empty, not half-full.