I just saw the email from HE about the new points system, which in theory I welcome, especially the idea that home will not have way more beds listen than are actually practical. But why in the world would be open to having pets in your home be a point driver? It should be the amenities in my home, not what someone else brings to it. We’re being penalized by not wanting other people to bring strange (allergy causing) animals into our home, not granted points because of what we offer. That makes no sense. I hope HE reconsiders that as part of the algorithm.
3 Likes
Thank you for your thoughts on the upcoming changes to the way that guest points are calculated. I will of course feedback your ideas to the relevant teams who are working on this. This is, of course, an ongoing issue, and as you can imagine, it can be challenging to meet everyone’s needs.
2 Likes
@across, I agree with you!
1 Like
@across
Respectfully, I disagree.
Your criticism about allocating a point-value to pet-friendly homes could be applied to any amenity that is not valued by a certain segment of the membership. For example, all-adult guest parties won’t value “child-friendly”, some people on vacation (like me) don’t watch TV, some people are travelling to visit friends or relatives and don’t care about being in a “world renowned tourist location”, some people don’t care about jacuzzis or bikes being available, parking, etc, etc.
Points are assigned for features of a home that appeal to a segment of the travelling membership. For some members, “pet-friendly” is a desirable feature and therefore should have a points value. In my opinion, the challenge is to assign points based on what percentage of the travelling population wants the amenity. This is where the current point-assignment algorithm needs improvement, in my opinion. For example, a pool may be highly desirable to only a small segment of the travelling population but is currently highly valued by the algorithm, whereas a refrigerator is highly desirable to virtually 100% of the population but is currently not highly valued in the points algorithm..
I’m glad that the point calculation algorithm is being changed. However, I do have concerns about how the member opinions were collected and used to prioritize the changes. I understand that veteran administrators, moderators, and support staff have used their knowlege based on user feedback on the facebook page, this forum, the platform chat volunteers, and emails to the support team. And this is good.
But, it would have been nice to do something like a poll of the membership to get a very large statistical sampling of actual data rather than anecdotal evidence.
For example, why not do a poll in which every amenity is listed and for each one, members are asked to indicate their level of interest in each amenity on a scale from 0 to 10, where
0 = No value or not applicable
10 = Essential
An email could have been sent to all 150,000+ members asking them to go to a website and do the poll within a period of time, let’s say 4 weeks.
It seems like a huge undertaking, but in this computer age, using commercial polling firms and statistical software, the results could be tabulated in literally seconds, with each amenity being given average and median scores, and other statistical attributes such as frequency distribution. This could have formed a solid statistical basis and starting point for the point-recalculation process.
I can already predict an argument that most members would not want to take the time to do such a poll… it would probably take perhaps up to 30 minutes or so, because there are many amenities; but even if only 1% of members did the poll, then there would still be hundreds of responses. I know that I would certainly take the time to complete it.
3 Likes
My point that it is not an actual amenity inherent in a home, but a choice by the host. I would say the same about “child friendly” too.
My point is that it doesn’t matter whether it is called an “amenity” or not. It is an attribute of a guest visit. A choice by a host most certainly affects whether or not a certain category of guests will even consider the house as a place to stay. All attributes of a home which provide a desired feature for a guest should be included in the scoring algorithm, whether it is a tangible physical item in the house like a bed, tv, A/C, etc; or a service such as a cleaning person, or a location in a desirable area, or a rule or permission which makes the house accessible to certain guests such as pet-friendly or child-friendly.
Not about your idea, just a data, it’s not 150,000, right now it’s around 264, 000 members.
1 Like
Hi Melissa. This post has a point. I have always wondered why my house on the beach in one of the most desirable locations has less GP than other random similar sized houses. It looks like the declared number of beds is a key factor and many list all sofa beds and lounge canapés as actual beds. While I understand there are sometimes large families, average HE travellers tend to be small parties. Could you consider changing the algorithm to discount any beds not in an actual bedroom? Thank you
How many of the 264,000 are paid members and how many are simply unsubscribed accounts?
I agree. Sofabeds and put-up beds are usually uncomfortable and in a non-private area of the house such as the living room. They should be awarded fewer points than the permanent beds. Also, I hope the new points system rewards number of bedrooms, not simply number of beds. Also, to be defined as a “bedroom”, a room needs to have a permanent bed. Some members try to get away with saying that an office with a futon or sofabed is a “bedroom”. I once requested a guestpoints stay at a member’s home which was advertised as a 2 bedroom apartment. In conversation with this member, I learned that the second “bedroom” was actually an office with a sofabed, but that they were currently not allowing that room to be used as a bedroom, therefore limiting occupency of the apartment to 2 guests. However, they didn’t edit their home description to say only 1 bedroom and 1 double bed, so the algorithm calculated their guestpoints as if they had 2 bedrooms, 2 double beds, and sleeps 4 guests. They did not offer to reduce their GP rate. I’ve also seen people who say that they have only 1 or 2 bedrooms but say that they sleep 8 or more people. It’s clearly an abuse of the points-calculation algorithm.
I’m not saying that these abuses are common, but the current points algorithm makes it easy to artificially inflate a home’s GP total., and it sometimes happens.
3 Likes
264,000 are the paying members. Total members is more than this, I don’t know the number but I guess you Can probably do X2 or something like this
1 Like
Hello Canuck,
On an IT point of view, the number of click on amenities boxes is accountable.
It is not a perfect system as a pool is usable during the hot season only but it gives an idea of the most clicked boxes.
I suppose the fridge box is rarely clicked because people think it is a normal amenity.
As far as i am concerned, only the location and the number of real bedrooms and real beds matters 
and lying on these points is quite upsetting.
Nathalie, thank you for your message.
When you say that the fridge box is rarely “clicked”, I assume that you mean “clicked” by people who are searching for a home.
The search function only displays a subset of the amenities that a home may have, and “fridge” is not one of them. So, instead of “rarely clicked” as you say, it is actually impossible for a member to search for a home that has a fridge because that amenity is not offered as a search parameter. So of course statistics would show that a fridge is “rarely searched”. In my opinion, if the points algorithm calculation considers the number of user “clicks” in the search function, then it should offer ALL of the possible amenities of a house as searchable parameters.
This is a justification for the development team to do a poll of members asking them to evaluate the importance of ALL possible amenities of a home when calculating guest points, rather than the incomplete statistics gathered by search function “clicks”.
1 Like