What do indicators really indicate?

[David Pogue] “How come cellphone signal-strength bars are so often wrong?”

[Reader response] “Like the battery indicator, the signal strength on a cell phone is deliberately weighted toward the high end. I worked on a phone development project several years ago. When the first units went to the carrier for approval, their first request was to toss the perfectly calibrated battery indicator in favor of one that sat at 4 bars for around 75 percent of the charge.”

Let me get this straight. Cell phone manufacturers fib about battery strength, to make it look (initially) like the phone holds charge better??

Should auto makers design a car’s fuel gauge so that it stays on “full” until a gallon or so away from empty? After all, customers will think - initially - that the car is getting great mileage.

[…]

There was a study about this a while ago. Sorry, can’t remember on which blog this came up. It seems the users used their cellphone a lot less (especially to initiate calls) when their battery indicator showed below 50%.

Of course this isn’t what the carriers wanted and so this was changed. Carrier happy, but screw the customer.

Posted by Phil on November 9, 2007 11:02 AM

This doesn’t match my experience. I spent years doing software for a cellphone company, and our group spent MONTHS making sure that the battery meters were perfectly calibrated. There were exact numbers (% of coloumbs left) for each number of bars shown. BTW, this is a very hard problem to do right, so it may just be that most makers don’t bother.

Posted by Andrew on November 9, 2007 03:05 PM

[See the entire post at goodexperience.com]

On fuel gauges, from How Stuff Works:

The main impediment to stretching your mileage is the fuel gauge on your car, which makes you think you have less fuel than you actually do. These devices are notoriously inaccurate, showing empty when there are gallons left in the tank and showing full for the first 50 miles.

Well, automotive fuel gauges straddle both sides of the fence, at least, they did the last time I checked (and a very cursory search of the current state of affairs indicates the same). The distance from Full to Half represents well over half the tank. I’ve heard it attributed to the shape of the tank and how the floater gauge moves in it (see also the How Stuff Works article for details); whatever the “real” reason, the positive sense that your car is still good to go adds to your positive feelings about the car. (In contrast to, “Oh my gosh, the car needs filling again soon. This thing just eats gas.”)

On the flip side, the gauge will show Empty before it really is, for the benefit of those who wait until the last moment before filling up (you know who you are). No one wants to actually run dry, and lengthening the grace period helps to ensure that you take care of business before it’s too late. The net effect is that it takes a long time to get from Full to Half, and then the indicator seems to race down to Empty.

But it all helps in giving the driver warm, fuzzy feelings about the car.

I’m not surprised (much) to read the claim here regarding cell phone battery life; I would be surprised if the tradeoff is worthwhile (customer feeling that battery holds full charge longer is more valuable than the risk of customer running out of charge unexpectedly). Perhaps the number of people who charge their phones every night, and never get below half-charge, is high enough to make that trade-off meaningful.

Contrast the four-bar cell phone battery gauge with the battery indicator options on an Apple PowerBook: visual icon, percentage of battery full, and time remaining. The indicator conveys detailed precision. How much correlation does the PowerBook indicator actually have with the battery status? I have no idea. Does the accuracy change over time, as the battery performance deteriorates? Also no idea; personal experience would suggest so.

That’s not to say that users don’t appreciate such design decisions; if we were aware of them, we might be very grateful for assumptions that gauges make (especially that early “Empty” on the car!).

But I think it’s fascinating to note the differences in agenda between the gauge designers and gauge users. Take-home lessons:

Knowing how much there is (or thinking you do) affects how you use the resource.

We were sitting with family the other night, and started to wonder what would happen if your credit card actually showed you how much credit you had left (or how much debt you had accrued). Would it affect your use of the card? You bet it would. That’s a fundamental difference between using cash or checks and using a credit card: it’s not just that cash or checks are payment debit versus accrual of debt, it’s that you see the impact of your purchase immediately (thinner wallet, smaller number in your checkbook).

The users assume that the gauge is there to serve their need to have a window on the guts of the machine. They take as a given that the gauge provides a reasonably accurate representation of the fuel status (essentially, use time left before shut down), allowing them to make informed decisions as to how and when they will use the device.

The designers assume that the gauge’s meaning in life is to make the users feel good about the device. If the device runs out of battery unexpectedly, the user will feel bad. If the device holds a charge longer, the user will feel good. If the device has no fuel indicator at all, the user will feel frustrated.