转自：上观新闻
Follow us
Swipe for Chinese version
It is a common practice to check review sites before settling on a new restaurant for a meal. But can we always trust these ratings? And does a higher score always mean better quality?
More and more patrons, disappointed by the quality of numerous highly-rated restaurants, are now gravitating toward more modest eateries in hopes of uncovering unexpectedly delightful dishes.
Hashtags such as "Youth flocking to 3.5-point restaurants in a 'revenge culinary' trend" and "The best flavors might be hiding in less acclaimed dining spots" are buzzing on Sina Weibo, and accumulating millions of views.
Amy Wang (pseudonym), hailing from Qingdao, Shandong province, has a knack for discovering these "hidden gems" and she shares her insights on rating apps.
As a part-time food blogger, Wang has visited over 200 restaurants in just six months.
According to Wang, the trend of favoring lower-rated restaurants reflects a growing distrust in the rating system, especially among young diners. Ultra-high scores are not always reliable, as they can be artificially inflated through various means.
1.offer incentives like complimentary dishes or discounts for positive reviews;
2.train staff to leave reviews on behalf of customers;
3.pay for professional services to raise their scores.
"Even labels like 'locals' favorite' might be obtained through deceit — some places hire internet ghostwriters using local IP addresses," Wang explained.
Reading through comments, especially the negative ones, can also be enlightening, according to Wang. "It helps you figure out if the criticism is about service, ambiance, or food quality, and whether you can overlook it. But you should also be wary of overly detailed comments with lots of words and pictures because they are often insincere," she said.
As Wang pointed out, circumstances vary greatly in different places. In major cities like Beijing and Shanghai, where nearly every aspect of life is tied to the internet's rating system, ratings can be reliable. Surprises are more likely in less-developed areas, where online ratings aren't as meticulously managed, and customers aren't as keen on sharing their experiences on social media.
Wang's most memorable experience at a "3.5-point restaurant" happened in a small town in Hanzhong, Shaanxi province. "I visited a place with only two reviews on Dianping, and to my surprise, I had the best remipi (rice slices) of my life," she recalled.
Eat? Or not?
Zhao Yige, 27, is a novice manager running a Chongqing hot pot restaurant in Beijing. Since its soft opening in early January, the store has received about 10 reviews on Dianping, all positive with at least 4 points. However, the overall score stands at 3.8 points, which Zhao finds disappointing.
She explained that review sites use complicated formulas for generating scores, which are not solely based on customer ratings. In this digital age, even a reduction of 0.1 points in rating can lead to reduced exposure, resulting in a drop in customers and revenue.
"I can tell that the review sites want restaurants to encourage customers to leave comments and make payments through the platform, both of which would boost scores," Zhao said.
Highly-rated places also get certain privileges. For instance, only those with 4.5 points or above can sell group coupons on the platform.
Upon hearing about the trend of lower-rated restaurants getting more popular, Zhao welcomes it. "This shows that customers are becoming more discerning and hold more diverse standards," she said. "I hope this trend continues, pushing more shop owners to focus on what's truly important for a restaurant — taste and service."
Tan Binbin from Chongqing noted that the trend of choosing moderately-rated eateries hasn't extended to the field of takeout.
Tan, formerly a manager of a leading takeout app in China, is now working as an independent consultant for takeout restaurants. He explained that the logic and algorithm of ratings differ between review sites and takeout platforms. For example, the rating system for food delivery platforms like Meituan and Eleme is much simpler — it uses a dynamically changing model that calculates the average points received over the past 30 days.
"This means the scores are mainly decided by customers rather than complicated rules set by platforms. It also allows restaurant owners to quickly respond to low ratings and improve within a couple of days. Because of this, ratings on takeout apps are usually higher than those on review sites," Tan explained.
"We live in a world surrounded by ratings. We check scores and reviews not only when choosing a restaurant or ordering takeout, but also before online shopping and watching films," Tan said. "Ratings are supposed to be helpful. They just need to be fairer and more representative."
外出就餐第一步
打开评分软件
可是评分这么高
怎么踩雷啦？
外出就餐第一步，看看评分。但你有没有想过，评分的可信度又有多少呢？评分高就一定品质好吗？
消费者频频在“高分店”踩雷，所以把目光转向“低分店”，希望“低分店”提供超出预期的菜品。
“年轻人报复性挤爆3.5分饭店”“评分低的店可能更好吃”等词条冲上热搜，引起热议。
来自山东青岛的Amy王，在寻找这些“沧海遗珠”方面有自己的技巧，她也针对评分机制分享了自己的见解。
图源：受访者供图
Amy王，副业是美食博主，短短6个月里去了200多家餐厅。
她表示，消费者对于低分餐厅的青睐反映了评分机制的信赖危机，尤其是失去了年轻消费群体的信任。商家可以通过各种手段实现高分，从而高分也无法代表餐厅的好坏。
1.提供折扣或者免费餐品
2.让员工假装客人评论
3.找专人提高店铺评分
Amy王介绍，甚至标着“当地最受欢迎”标签的餐馆也不一定是真的，因为有的店会雇本地ip水军评论。
评分和餐厅质量不对等？
这可难不倒
美食爱好者们
他们反其道而行
Amy王表示，看评论是个了解真实情况的好办法，尤其是那些差评，可以帮你辨别差评是针对餐厅服务、环境、还是餐品质量的，是否是无伤大雅的小问题。她还提醒要注意辨别带多张图的长评，这种一般都是假的。
Amy王指出，不同地方情况会有不同，像北京上海这种大都市，生活的方方面面都会和网络评分绑定，评分就相对可靠些。但一些没那么发达的地方，消费者懒得在社交媒体上分享自己的体验，就会有惊喜。
她就在陕西汉中的一家3.5分餐厅吃到了自己的“人生米皮”，尽管这家店只有两条评价。
27岁的赵一歌，是北京一家重庆火锅的新手经理。自一月试营业以来，店铺在大众点评收到约10条评价，都是4分以上的好评。但是店铺总分却只有3.8分，赵一歌感到沮丧。
图源：受访者供图
她说评分机制很复杂，消费者评分不是唯一考评因素。大数据时代，评分哪怕降0.1，都有可能影响店铺曝光，降低客流量，影响收益。
为了改变现状，赵一歌购买了平台会员，获取更多商家权益，更好地管理店铺。
她发现，评分平台希望商家能引导消费者发表评论或者通过平台付费，这两项都会提高店铺评分。
高分店铺会有一些“特权”，比如只有评分4.5以上的店铺才能售卖团购券。
听说低分餐厅的流行，赵一歌欣然接受，表示这是因为消费者有自己的判断，他们的评判标准也越来越多样化。她说：“我希望这种趋势能继续下去，让更多的店家关注餐饮真正重要的东西——味道和服务。”
图源：受访者供图
重庆的谭彬彬表示选择低分餐厅的风还没吹到外卖平台。
谭彬彬之前在中国头部外卖平台当经理，现在创业给外卖店家做咨询。他说评分平台和外卖平台的评分逻辑和算法不同，以美团和饿了么为例，它们的评分机制更简单，即用动态变化模型计算最近30天商家收到评价的平均值。“所以外卖评分主要是
他表示评分无处不在，人们外出就餐或者点外卖之前都会先看评论和评分，购物和看电影之前也会看。这些评分应该为人们提供帮助。它只是需要变得更加公平，更能代表店铺的真实情况。
Source: 21st Century
Edited by Wu Wangtiancheng
