Review Stuffing services : Really worth it ? An exploratory analysis
A month or so ago, I got some physical spam mail : a rarity in this day and age. Ironically, the physical mail was advertising a service perfectly fit for the digital era — A review stuffing service for sites like Amazon. The spammer also guaranteed that my listings would gain the coveted “Amazon Recommends” badge — a system whose algorithm details are still unknown.
Needless to say, my interest was ..piqued.
Large “managed marketplace” companies like AirBnb and Amazon often have such metrics, like reviews and “SuperHost” status. These are accorded by unseen algorithms and, presumably, factor into customer purchases and rentals.
AirBnB claims a certain set of criteria must be fulfilled for the “Superhost” label. This makes the review stuffer’s claim rather interesting — can they really game criteria like minimal cancellations and hosting guests ? And what will such an effort be really worth ?
AirBnB host income depends on two factors : Price of the property and the average occupancy. Taking some data from AirBnB, specifically the Seattle AirBnB listing data from 2016, I decided to check one of these dimensions, the Pricing.
Q1. Does a “SuperHost” status in AirBnB actually enable you to charge more ?
Price is in many cases, the main driver of rental decisions. Would someone pay more for a “Superhost” as compared to a non-superhost rental ?
My first instinct was to compare the pricing per square feet of real estate between Superhosts and Non-superhosts. Unfortunately, most of the data did not have such a metric. I therefore chose to differentiate based on the rental type : Full apartment, Private Room, or Shared.
The Results :
It does seem like the “Superhost” tag allows for higher pricing, if only slightly. However, this assumes all other things remain equivalent, but do they ? What about amenities provided, or the neighborhood the listing is in ? What about visibility ?
I guess this warrants a further look.
Q2 : Do Neighborhoods matter significantly when it comes to pricing ?
There is an old saying in the hotel business : “Only three things matter in a hotel’s success: location, location, location”. With the figure above, we disregarded a major factor : the neighborhoods affecting the pricing.
Lets take a look :
The graph makes it quite clear that different neighborhoods have different average pricing, and the range is large with high overlaps. This muddies the signal we got from the Superhost graph.
Q1 & 2 : What about combining them both ?
Lets take a closer look at the West Queen Anne neighborhood, but this time see the difference caused by the Superhost Status :
As we can see, the picture is quite different from before — quite opposite, in fact.
It is clear that the Superhosts are not able to command higher prices in this neighborhood. In fact, their low pricing may be key to them keeping their Superhost badge.
Q3 : What about Positive reviews ? Do they not help boost earnings ?
While natural positive reviews would imply additional effort or costs to the host to justify his high pricing, review stuffing would not follow that rule. But is review stuffing really worth it ? Lets see:
There seems to be a significant upward trend of Pricing to Review Score. Assuming that reviews are independent of the Location effect (a logical assumption), we can say that having a higher Review Score does enable a host to charge a higher price for his/her property.
There is so much more to explore on this topic. We have yet to explore occupancy rates, which may actually get a benefit from the visibility given by the Superhost Badge. We have also not explored other facets like the amenities provided, the photographs and the descriptions — all of which is likely to enable higher pricing. But for now, I will stop here.
The conclusion ? It can be argued that review stuffing services do have a place in a hoster’s arsenal. If you are not worried about ethics or the risk of getting banned due to ToS violations, that is.
The Superhost badge though ? Probably not worth paying for.
You can find the details and the code in my Github repository here. Feel free to reuse/explore/extend the code, and generally, have fun with it.
What is your opinion of such review stuffing services ? Do you feel forced to resort to such tactics, because “every other person is doing it”?
I would love to know your opinions !
