Varying Success Rates among Mobile Apps in Prebooking Taxi Services Excessive Access Right Opening to Risks of Sensitive Personal Data Leakage

15 March 2018
Forward
Email this page

Varying Success Rates among Mobile Apps in Prebooking Taxi Services  Excessive Access Right Opening to Risks of Sensitive Personal Data Leakage

Technological advances have changed the business landscape as never before.  The emergence of e-hailing service platforms, including mobile applications of taxi hailing services, has also brought changes to consumer transport behaviour in recent years. How do consumers go about choosing taxi apps? To evaluate and compare their performance, the Consumer Council has conducted a test on 7 of these more popular prebooking taxi apps, with an investigation team of nearly 80 members taking out a total of 429 actual trials.  
 
The investigation found varying performances – the overall success rate in prebooking taxi services spanned over a wide range from 52% to 98%. The success rate was affected by a number of factors including the time for the booking, the distance of the journey, the offer of tipping and its amount, or even the cross-harbour tunnel opted. But the most disappointing was in short haul services – the success rate was as low as below 20% in certain taxi apps. Prebooking taxi apps are intended to offer fast and convenient transport service, without having to wait in the streets, but if they are so dependent on varying factors affecting their performance, it might leave a negative impression on consumers in the use of prebooking taxi app services. 
 
Furthermore, the test revealed the practice of the majority of taxi apps in requesting users for the right to collect and access their data totally unrelated to e-hailing of taxi services. The Council is deeply concerned that this is intrusive to consumer personal privacy and may even lead to sensitive information being leaked out.  Consumers are urged to study and compare the various taxi apps, before download, their right in collection and access of data.
 
To the customers, their satisfaction level of taxi apps and the driver service are inter-dependent, besides increasing the success rate of the apps, the industry must also strengthen the overall taxi service quality. Only then could the industry maintain its competitive edge in a rapidly changing personalised prebooked transport service.
  
In the past 2 months, 74 members of the investigation team posing as customers made trial use of 7 taxi hailing mobile applications, each with at least 60 times totalling 429 times. The trials were conducted at various time periods – peak hours (148 times) and non-peak (281), and at various distances – long haul (140 times), medium haul (180) and short haul (109), in order to compare the actual performance of these apps under different circumstances.
 
Vast variations in success rate of booking and service quality
 
On the issue of most concern to consumers: the success rate in taxi service booking, the results indicated that out of 429 trials, the success rate being 79%, but vast variations existed among the apps – the best performer achieved an overall success rate of 98.4% while the worst only 52.4%. Despite 2 of the apps attained success rates of over 90%, some apps were consistently lower than 60% regardless whether it was peak or non-peak hours, or long, medium or short hauls. 
 
To find out the relationship between the booking success rate and the time periods required for service, the investigation team made booking during peak and non-peak hours and found the success rate of booking in 1 app dropped drastically from 92.5% during non-peak periods to 72.7% during peak hours.   
 
Besides prebooking during peak periods, the success rate was affected also by the differences in journey distance. While the success rate in booking for medium and long haul services varied by less than 10%, in the majority of cases, the success rate in booking for short haul journeys was clearly lower than medium and long journeys; in 7 apps the short-haul success rate was differed by 2% to nearly 70% lower than that of the long haul. In the case of an app offering 15% discount, the success rate for long haul taxi booking was a high 85% but dropped to below 20% for short haul booking.
 
For booking journeys involving cross-harbour tunnel, without specifying the choice of tunnel, the success rate was generally about 70%. However, the success rate in the case of choosing the Cross-Harbour Tunnel would be 64% whereas choosing the Western Harbour Tunnel or the Eastern Harbour Crossing, the success rate rose up to nearly 90%. 
 
Tipping could be an incentive to increase the chances of success, the more generous the tipping the higher the chances. Without any tipping offer, despite initial booking failure, the success rate was down to less than 30%, with offers of $10 to $20 the chances rose to nearly 40%, and with $30 to $50, to a higher 60%.
 
 
To consumers’ most concern would be no taking of orders from drivers or taxis picking up late, yet overall some 55% of the drivers arrived later than expected. The ratio of late drivers from the various apps was around 44% to 63%, the average time of delay was 3.8 to 6.5 minutes; in the worst case, the driver showed up 30 minutes later than the estimated time, a most unsatisfactory performance. For the waiting time for the taxis to turn up, the results also showed considerable variations from 5 to almost 14 minutes.
 
Excessive access right of apps collecting unnecessary user data
 
In comparing the apps for their practice in data collection and access, the Council found most apps would seek to obtain consumer information irrelevant to e-hailing taxi operation, such as requests to access/alter memory card contents, access to the user’s photo, contact person data, records of phone communication, or even download the customer file without their prior knowledge. The Council is seriously concerned about taxi apps in the collection of personal data well in excess of their actual needs.  
 
App service providers are urged to refrain from such a practice, and to clearly list out the information needed for the choice of consumers. In fact, one app was found to need only the location and other service-related information. This clearly reflects that app developers have no actual need to collect large quantity of personal data to be able to effectively provide e-hailing taxi services. Consumers are, therefore, reminded that before they download a mobile taxi app, apart from service quality consideration, they should also carefully understand and compare the various apps for their personal data policy, to safeguard personal privacy.
 
At the end of taxi journey, investigating team members would complete a questionnaire, and give their evaluation and level of satisfaction. The findings pointed to room for improvement in various areas of performance.  
 
In terms of convenience of use, the apps varied from 30% to over 95%, the worst apps were judged to give insufficient information, leaving the consumers uncertain if their orders were successfully accepted; and likewise consumers were not duly informed when their booking failed. Further, upon confirmation of booking, the consumers would be notified of information about the drivers, including the taxi licence number. But in 6% (21) of the trips the actual licence number was found to differ from the number on the app. Consumers are reminded to ensure the accuracy of the taxi licence number as they might need it to track down the driver in the event of leaving behind personal belongings in the taxi. They should therefore verify, before boarding a taxi, if the number on the taxi licence plate is the same as shown on the app.
 
The service and attitude of taxi drivers have long been an issue of consumer dissatisfaction. In nearly 340 trips, 90 or over one-quarter, investigation team members were displeased with the drivers’ behaviour, including rudeness, frequent changing of car speeds or use of mobile handsets. On the basis of the individual apps’ success rate, the chances of the customers encountering an unpleasant experience of one app could reach a high 36.4%, even the app with the least chances of an undesirable experience was 14%.
 
Among the 7 apps, 3 offered automatically a 15% discount, of which 2 used just the numerals “85” to denote implicitly the discount offer. But if customers do opt for the 15% discount their success rate in hailing a taxi would be reduced by 10% and even if the driver did accept the discount request, more than 30% of the trips turned out in the end with less than 15%, the actual final discounts were more likely in the region of  4.2% to 12.4%.
 
In accordance with the Road Traffic (Public Service Vehicles) Regulations, “no person acting or purporting to act on behalf of the driver of a taxi, shall in any manner attract or endeavour to attract any person in order to induce such person to make use of the vehicle”. The Council has referred the apps in question to the Transport Department.  On completion of the test, the Council went over the apps again and found some to have already removed the choice column for discount service, nevertheless consumers could still put in requests for discount under a different column for miscellaneous matters.
 
The Consumer Council reserves all its right (including copyright) in respect of CHOICE magazine and Online CHOICE.