Why some applications use not real progress scales

Original author: KAVEH WADDELL
  • Transfer
The author of the material analyzes the performance indicators and reasons why these interface elements are intentionally programmed to incorrectly display data.



From the point of view of productivity this year, I decided to sit down early in order to bring my taxes in order. Working with them was more than usual, so I took some time to put all the information in TurboTax and make sure that everything was done correctly. In the course of work, the online tax reporting program regularly assured me that it had already helped me identify every tax break that suited me, and promised that thanks to it, there were no errors in my declaration.

However, watching one particularly pleasant animation depicting gradually filling yellow and green lines, I wondered if what I saw at that moment on the screen reflected the progress of a real task performed in the background. Is it true that the “detailed analysis” of my income, which the page, judging from the information on it, performed at that moment, really takes a lot of time there? Didn't TurboTax make all the necessary checks before, even when I brought them into the program?

I asked my friend Andrew McGill to help me deal with the processes that animation is meant to accompany. We went through the source code of the TurboTax website and soon confirmed my suspicion: the animation was a ready, immutable object. We did not find any attempts to interact with the site servers at the time of its display. Every TurboTax user saw the same picture, of the same duration. We found a similar process on at least one other page of the site, which was supposed to display the course of testing the applicability of “all possible tax exemptions” using three animated scales.

But why is this done? What was the purpose of creating a false impression about the time of the process, taking the time from the user?

The point is not that TurboTax like to mock their customers. According to Ethan Edar, a professor of computer science and computer science at the University of Michigan, artificially slowing down the load on the site is one example of a phenomenon that he calls "benevolent deception." In 2013, he and two of his colleagues from Microsoft researchers published a paper that described a wide range of design solutions that trick users, but ultimately do it for their own good.

A benevolent deception can hide uncertainty. For example, Netflix automatically downloads general recommendations if user bandwidth is not enough to get individual recommendations. This approach is also useful if you need to mask and smooth out small technical hitch (progress bar, which is filled evenly and regardless of the speed of performing real actions). Another case - the desire of developers to help users get used to a new type of technology. For example, fixed animation in Skype, designed to indicate that the conversation is still on, and the other party can still hear you at that end of the line.

The word "deception" has a negative connotation, and lying to users usually causes criticism. But according to Edar, this kind of deception, if properly applied, becomes a useful tool. Designers have been cunning with interfaces for many years, and users usually do not pay attention to it.

Curiously, the TurboTax animations are different from most of the similar practices learned by Edar. Instead of hiding the slow work of the service, they introduce an artificial delay where it is virtually absent. This technique is used to create a user’s sense of reliability of the product to which he has just entrusted his financial information.

“The process of filing a tax return is often associated with at least a little excitement,” said Rob Castro, an official spokesman for Intuit, the parent company of TurboTax. - To compensate for these sensations, we use a set of various design elements - content, animations, movements. This allows us to reassure our customers and instill in them the confidence that their declaration is filled out correctly, and they deserved all their money. ”




Edar made a similar decision in an experimental game that he invented almost twenty years ago. Its essence was that two people had to agree on a price, talking through two separate mobile devices. The final step was quite complicated: the participants' proposals were encrypted, transferred over the wireless network and compared, after which the program informed on the basis of this information whether the transaction could take place.

Despite its complexity, this sequence of operations was performed almost immediately in the first iteration of the program. However, this speed led people into confusion. “They reacted in the spirit of 'Wow, is that all?“, ”Says Edar. “We really didn't like it.” As a result, Edar came up with a trick: instead of instantly completing the final stage, an animation appeared on the screens with asterisks instead of symbols.

And this imaginary cryptographic representation worked. “The application’s work has become more pleasing to the participants of the experiment. It seems that their confidence in the result also increased. ” (This assessment, however, is based on the testimony of only a few people; no formal study of user reaction has been carried out.)

The practice of creating artificial hitches is not a favorite topic among designers for conversation, but it is quite common. Last year, Fast Company author Mark Wilson discovered a similar trick on the Facebook security page. He then led other similar cases. For example, the application for approval of applications for a loan, holding some pause before the announcement of the result, to remove the extra suspicion of users. Or a website that provides individual recommendations on cellular tariff plans, slowing down the results in order to convince users that they will indeed receive a personalized answer. Similar examples are often published on someone else's twitter account, as is the case.with a scale of progress Verizon, which turned out to be a simple javascript-timer.

In his material, Wilson cited a study of two Harvard professors of 2011 who had studied in detail this approach, which they called the "illusion of work." They found out that websites, which simplify the processes going on them as much as possible, received less satisfied reviews from users. “When websites go into a transparent mode of operation, notifying users of the program’s work, people can give preference to websites with a longer delay than those that provide results instantly,” the authors wrote, are the same. "

But not every case of a benevolent deception is thought so that people think that the situation is under control. One of these tricks in particular was aimed at introducing uncertainty in the visual presentation of data, and almost caused heart attacks across the country.

This is the online panel The New York Times, which covered the results of the voting during the US presidential election in real time. It depicted three circular scales reflecting the forecasts of the publication regarding the distribution of votes and electoral colleges, as well as the chance of Hillary Clinton and Donald Trump to win the elections. Throughout the night, the arrows on each scale danced from side to side, first signaling a serious advantage to Clinton, and after midnight, indicating Trump’s clear victory.



The counters' arrows constantly moved one way or the other, adding sharpness and uncertainty to the moment. Several enterprising readers studied the source code of the page and, finding that the shooters changed their position by chance, expressed their anger on Twitter. The word “irresponsible” was often used in their posts .

A week later, Gregor Eish, one of the Times' designers who worked on the online panel, explained the logic behind choosing such a decision on his blog. According to him, the arrow at any time fluctuated within the limits of the permissible prediction error. This technique was conceived as a reflection of the living, ever-changing nature of the forecast, and embodied the uncertainty of the result. Gradually, during the voting, the forecast became more and more accurate, and with it the degree of deviation of the arrow during oscillation decreased.

I asked Aisha whether his team would like to reconsider its approach in the future due to a similar response, and received a negative response. According to him, the visualization correctly represented the information and performed its function, and he would use the same approach in a similar situation. Negative feedback, in his opinion, most likely represent anger at the election results as such. “The point is simply that on election night, we were the first destroyers of hope for a large number of people,” says Aish, “and so it happened that we took the fire on ourselves.”

The only adjustment that Aish would like to make is not to display the candidates' chances of winning as a percentage. In the view of the majority, the 80% chance of Clinton meant a definite victory, whereas in fact her victory was not as likely as it seemed. “No one will ever trust contraceptives if their chance of failure is 1 out of 5, but our idea of ​​probability made many people believe that Clinton had a clear advantage,” says Aish.

When Twitter users removed integuments and Aish’s cunning was exposed, some people felt that they had been deceived with malicious intent. And therefore deception works best when nobody knows about it.

I asked Edar about a possible edge, having passed that, deception from benevolent turns into harmful. He outlined three basic rules. First, designers should prefer fair and transparent solutions. Secondly, the positive result of deception should be quantified. And thirdly, if you ask the user about his opinion, he should give preference to the deceptive option. Although, of course, most designers do not have the opportunity to ask their users about such things, so they have to make their own decisions.

However, deception for the benefit of the user does not necessarily have to complicate the work of the designer. In fact, according to Edar, a good deception usually benefits all parties. Satisfied customers receive a more thoughtful solution and continue to use the product, and it is also possible to pay for it, which allows the designer to understand that he has done his job well.

Take, for example, the same TurboTax. Techniques used in his designcan reduce customer experiences during the tax reporting period and give them more confidence in their own finances. Instead, they will continue to use the service and pay for its services. However, TurboTax has another incentive to use such animations. This service positions itself as a friendly guide to the thorny jungle of loans, benefits, relaxation and the forms that Americans have to deal with every year. And the preservation of this jungle in their most inhospitable and complex form is very beneficial for Intuit. The company regularly lobbies for maintaining a complex US tax code in its current form and opposes proposals that would lead to its radical simplification.

Therefore, a few extra seconds of animation give you the impression that TurboTax is diligently working to process information about your income, thereby improving your well-being. However, in addition to this, the program makes it seem to you that it works wonders. And in the end, when she asks you for $ 50 or $ 100 for her efforts, these few extra seconds make it easier for you to part with your money.

image

Also popular now: