Statistician, Department of Statistics & Data Science, Consumer Reports, focused on Machine Learning and Natural Language Processing
In August, 2020, Consumer Reports (CR) investigated fee transparency with food delivery apps. We inspected four of the most widely used online food delivery companies – DoorDash, Grubhub, Postmates, and UberEats. Our research focused on seven cities : Chicago, Jersey City, Los Angeles, New York City, Portland, Seattle, and Washington DC. All of these cities attempted to protect consumers by placing restrictions on these companies. Our investigation found that the companies we studied can do better to improve fee transparency and increase company accountability.
To raise awareness regarding the questionable company practices highlighted by this work, Consumer Reports launched a Share Your Story campaign on the Consumer Reports site, providing a link through which our 6M+ members could tell us about their personal experiences with these services. We asked consumers, delivery contract drivers, and restaurant workers to share their experiences with food delivery apps. This campaign yielded 400 Consumer Reports member responses in which they provided their state, city and zip code information along with a description of their experiences. Figure 1 below provides a sample of the stories shared by CR members.
Figure 1: Stories about food delivery app experiences shared by CR members
These voluntary responses reflect only the views and experiences of CR members who shared their story; this work is not nationally representative nor representative of all CR members.
As the effects of COVID-19 continue to decimate large portions of the economy, respondents indicated that food delivery app companies are taking advantage of the changes consumers have made to their everyday lives. People are reluctant to leave the house and much more likely to order delivery for fear of contracting the virus. One woman from Los Angeles, CA stated the following,
“I’m so disappointed in the food apps I’ve used during Covid-19. They really could have stepped up during this crisis considering their increased profits. I had no idea that in addition to the charges I pay they are also charging the restaurants and grocery stores such enormous fees. I have made a firm commitment to stop using them and pick up my orders myself.”
Others claim these companies try to capitalize on pandemic-driven lifestyle changes through increased fees and price markups and even further, that they take deceptive measures to prevent the consumer from realizing the source of the increased cost. They feel this lack of transparency allows the food delivery companies to avoid the responsibility of fairly pricing their services. For example, a gentleman from Morgantown, WV, stated,
“Due to the pandemic, I saw a raise in fees from multiple services. I wish the fees were more transparent because I would prefer to think the increase in cost is to help these businesses to keep up during the pandemic crisis, but I also saw some restaurant workers complaining about it, so not so sure where the money ends up. Services I used: GrubHub, DoorDash, DubVEatz, EatStreet.”
The word clouds below highlight the language most frequently used by the CR members who shared their stories; the larger the word, the more it appeared in the stories. It is clear from Figure 2 below that the words fee and tip appear in the stories more frequently than any other words. We discuss these themes in more detail later.
Figure 2: Word cloud of most frequently used words
Figure 3 is a sentiment word cloud. This word cloud shows the positive and negative words used most frequently throughout the stories.
Figure 3: Word cloud of most frequently used positive and negative words
To get at all the issues in the stories shared by respondents, Consumer Reports employed natural language processing and advanced statistical topic modeling using RStudio’s tidyverse package through an RShiny app designed by CR’s Statistics & Data Science team for a robust investigation and quantification of the responses. Our work revealed trends in the frequency of issues and the companies most commonly associated with these issues.
Figure 4 below highlights the counts and corresponding percentages of unique mentions of the specific problems faced by CR members who used these services and shared their stories with us. Each bar represents the number of times a particular issue was mentioned in the stories. Note that the counts are unique at the sentence level, but not the review level. For example, we found a total of 291 sentences that were about fees, but the same review may contain a sentence about delivery fees as well as a sentence about a service fee.
Figure 4: Frequency of mentions of specific issues
Figure 5 below is an interactive chord diagram which shows the relationships between the top ten most reported issues by CR members who shared their stories. The width of the bands indicate the number of stories that mention two connected issues. Also, when hovering over a band between issues, the number of stories that mentioned both is displayed. For example, the thicker pink band that connects fees and tips indicates that there were a relatively large number of stories that mentioned both of these issues. From this chart we can see a connection between multiple issues reported by consumers who shared their stories.
Figure 5: Relationships between top ten issues
We detail below four key issues found in the experiences shared with us by responding CR members.
Key Issue 1: Too many fees
The largest number of respondents, at least 127/400, indicated fees as the most notable pain point in their experiences with food delivery services. For example, many report that the cost of the order is significantly greater via second- and third-party websites. Some expressed feeling taken advantage of by this practice because it is unclear if fair compensation is provided to the restaurants or delivery drivers. A simple delivery order may accrue charges such as service and delivery fees, small order fees, tips, and taxes in addition to the cost of the food. The story below, shared by a gentleman from Eagleville, PA, outlines the exorbitant nature of these fees.
“As a customer that uses multiple food delivery apps: Grubhub, DoorDash, Uber Eats, I find that prices for each menu item are higher through the apps than if I ordered directly through the restaurant. Some apps also then charge a “Service Fee” and/or “Delivery Fee”. I have no idea how much of those added charges go to the restaurant, driver, or app company. It is not clearly outlined at all. Plus I need to tip on top of all those extra charges to make it work the driver’s time and effort. So, it quickly becomes ridiculously expensive to order through the apps.”
Respondents report that even individuals who pay monthly or yearly subscriptions for delivery services often pay reduced fees. Moreover, delivery fees are generally fixed in a given area, so a person living across the street from a restaurant is charged the same delivery fee as a person that lives miles away. Perhaps most disturbing, many respondents claim fees have steadily increased with demand for food delivery.
Figure 6 below provides a depiction of the counts and corresponding percentages of unique mentions within the category of fees. Each bar represents the count of CR members who mentioned the labelled issues. Note that a respondent may be counted only once within a bar but may be counted again between bars.
Figure 6: Frequency of mentions relating to the key issue of fees
Key Issue 2: What to tip?
The second largest number of complaints from the CR members who shared their stories, at least 86/400, relate to the tip. Many report feeling reluctant to tip well given the multitude of fees. According to the shared stories, this reluctance to tip well then affects the willingness of drivers to accept deliveries. For example, another gentleman from PA stated,
“As a delivery driver, it is really only worth doing if the customer tips really well. Without tips, a 5 mile trip that takes 20-30 minutes will only pay about $3. Meaning we’d only make $6-10/hour if no one tipped – assuming we are constantly getting orders, which we aren’t. And when you factor in that we are using our own vehicle, paying for gas, insurance, maintenance and quick depreciation, $6-10/hour is not nearly enough. So we rely on tips very much. Many driver’s will not accept a delivery request if there is no tip or the tip is not high enough.”
Additionally, some respondents feel tricked into over-tipping when the preset tip percentage is 20 or 30 percent of the total, including the service and delivery fees, and they do not realize this before placing the order. In some cases, they even report that a portion of driver tips are taken by the app services. There is clear confusion by app users as to who is actually receiving the tips.
Figure 7 below provides a depiction of the counts and corresponding percentages of unique mentions within the category of tips. Each bar represents the count of CR members who mentioned the labelled issues. Note that a respondent may be counted only once within a bar but may be counted again between bars.
Figure 7: Frequency of mentions relating to the key issue of tips
Key Issue 3: Lack of transparency
Another item of contention discussed by the respondents relates to differences in pricing between food delivery apps and the actual restaurants/grocery stores. They report significant mark ups in the app services during COVID-19 and as the demand for delivery service rises.
Many of the CR members who responded believe the delivery app companies use dark design patterns in order to obfuscate the fees. This includes unexplained fees, fees that aren’t itemized, unannounced price mark-ups, and false advertising. For example, a gentleman from Visalia, CA stated,
“I’ve frequently used DoorDash, GrubHub, and Uber Eats. All of them have different prices for the same restaurant. All of them have multiple ‘fees’ added to the cost of the food, and none of them are clear on where the money goes. If the delivery fee goes to the restaurant, why is it different for each service for the same restaurant? Why should there even be a delivery fee for the restaurant–they don’t charge one if I pick up my food there–if I’m paying someone else to pick it up, why does that impact the restaurant at all? Does any of the fee that the service charges go to the driver, or is the tip I pay all they get? So many questions, and so few answers.”
Transparency is also lacking regarding the commission taken by food delivery apps and the potential damage this does to restaurants and drivers. Only a handful of respondents expressed that DoorDash, Instacart, Grubhub, Vons and Smith’s Food and Drug were fairly transparent and even fewer claimed to receive itemized receipts from Grubhub and UberEats.
Figure 8 below provides a depiction of the counts and corresponding percentages of unique mentions within the category of transparency. Each bar represents the count of CR members who mentioned the labelled issues. Note that a respondent may be counted only once within a bar but may be counted again between bars.
Figure 8: Frequency of mentions relating to the key issue of transparency
Key Issue 4: Risky business
Fraudulent activity is another major concern facing the respondents who use these delivery services. Some indicate they were subjected to charges for food they did not order. Others claim food delivery apps deliberately avoid disclosure of data breaches that put personal consumer information at risk of theft. For example, a gentleman from Dallas, TX stated that,
“Because of underlying med conditions I use the services provided by Instacart and wholefoods—-but folks aside from what seem higher costs it ain’t the same—-I have received 5 pounds of bananas when I ordered 5 bananas or a bag w/rotten peaches or way past due grapes—-and with both delivery services the receipt of your purchases is purposely not included in your bags——furthermore instacart failed to inform me when they were hacked and my and many others credit info was stolen—-is that legal?”
Some CR members who responded also report unknown charges appearing on their accounts. Perhaps most egregious, some online ordering websites, like that offered by Google, imitate the appearance of an actual restaurant website in order to attract consumers who are then subsequently overcharged for their order. Respondents claim these instances of potential fraud are often inadequately resolved by representatives that are evasive and unwilling to issue a refund. Ordering services have also been accused of advertising unofficial menus without the consent of restaurant owners.
Figure 9 below provides a depiction of the counts and corresponding percentages of unique mentions within the category of potential fraud/identity theft. Each bar represents the count of CR members who mentioned the labelled issues. Note that a respondent may be counted only once within a bar but may be counted again between bars.
Figure 9: Frequency of mentions relating to the key issue of potential fraud/identity theft
How widespread are these practices?
According to the CR members who shared their stories, the practices outlined above are widespread. In nearly all states, fees are the predominant topic of complaint followed by tips. But the majority of respondents from Florida and New York mentioned Instacart and indicated pricing discrepancy and transparency as the most significant pain points. The majority of respondents from California and Texas mentioned DoorDash and indicated fees and pricing discrepancy the most. And the majority of respondents from Massachusetts mentioned Grubhub and indicated fees and transparency as the most significant pain points.
We filtered the stories further to study six cities in particular – Chicago, Los Angeles, New York City, Portland, Seattle, and Washington DC – based on their claims that they enacted unique local rules or ordinances to protect consumers and small businesses. As with the complaints across states, fees and tips are the most numerous topics found for these cities, which calls into question the effectiveness of these restrictions.
In conclusion, it is clear from the stories shared by CR members that the concerns previously expressed in our Consumer Reports September article are well-founded and should be acted on immediately by all food delivery companies and services. We strongly urge all these companies and services to implement fair and transparent practices that do not exploit their customers’ struggles during these challenging times.
I would like to express my sincere gratitude to the Consumer Reports team members who helped me complete this blog. Listed alphabetically: Kimberly Fountain, Dina Haner, Stephanie Nguyen, Kristen Purcell and Michael Saccucci.