|Vertical based Ratings and 360 Identity|
| 1:17 am on Aug 21, 2005 (gmt 0)|
As part of the recent release Y Local has introduced vertical based ratings and Yahoo 360 personal profile identification as part of the user generated ratings structure.
Restaurant/Bar Ratings look like this:
Hotels Ratings look like this:
A very natural and nice progression in ratings.
As far as 360, well.. here we go.
| 5:47 pm on Aug 21, 2005 (gmt 0)|
I see hotels as a complex of trade-offs related to the purpose of the trip. I suspect I'm not alone. So why the same review for everyone? What I really want to know about a hotel room:
1. Proximity/centrality to what I want to visit.
2. Rate + tax + (parking charge + other gotchas).
3. Clean and well-kept. (Once you get over 100 rooms you're bound to have a room with issues for an occupant.)
4. View, where relevant, which includes Seattle's SpaceNeedle. ;)
5. Proximity to speedy transport. (Boston! Arghh!)
6. Child friendly or not: Pool size and quality, clientele, etc.
7. Nature/cost of in hotel dining as convenience when traveling with kids.
Would reviews yield a different value if they were structured to address questions (or need for info) based upon the expected consumer:
A. Business traveler needs/values.
B. Adult no-children pleasure trip.
C. Family traveler.
D. Other, for example "reason for travel": Reunion, death in family, visit hospital, etc.
I think that approach would be more valuable by far.
Besides, what I really want to know is the quality of the mattress that I'm going to sleep on because the longest stretch of time I typically spend in a hotel is whilst I'm asleep - or attempting to sleep. Where's that in the list?
Oh yeah, and what about:
1. Water pressure in the shower and whether the bathtubs drains quickly?
2. The taste of the water when I take a drink or brush my teeth?
3. The noise generated by the surrouding businesses or traffic?
4. The smell of the room cleaners they use?
5. How quickly calls for assistance are answered?
6. What size is the desk and how easy is it to plug in PCs, etc.
7. Do the windows actually open so I can get some fresh air?
8. How slow are the elevators and are they all working right now?
9. Are there any mass bookings for the time I'll be attending, such as they international extreme sports association?
10. How easy is it to adjust the air conditioning?
11. Is the mattress firm, very firm or like a slab?
12. Do the room doors slam like jail cell doors, so I can listen to them slamming day and night?
13. Are the walls paper thin so I can listen to the T.V. in the room next door?
14. Is the pool picture nicely angled to make it look Olympic sized when it's more like tea cup sized?
15. Is the hot tub constantly down for service?
16. Are the number of parking spaces barely enough?
More often then not, when a hotel experience had issues, it was about something transient - like the hotel restaurant being closed for renovations. (Dang, now where am I gonna get breakfast for the kids.)
Okay, obviously I have issues about the appearance of value in reviews . . . so I'll stop here . . . for now . . . but truthfully I see most reviews for what they are: Almost nothing, certainly nothing to base a decision upon and I'll start by assuming that the better ones were written by some relative of the owner.
The factors that really matter seldom get mentioned anyway and, despite the number of variables I've mentioned I haven't mentioned several of the most important ones - such as hotel brand. A well defined brand - consistency of experience - carries a lot of weight, which includes "fixing it" when something about your experience is off the mark. For less well known hotels often the info to be aware of is the availability of other hotels in the area at the relevant time. ;)
Where's that mentioned in the reviews?
| 12:56 am on Aug 22, 2005 (gmt 0)|
Vertical based ratings vs. one overall rating has been debated since the early days of online rating.
Vertical based ratings was originally championed by Deja - who brought it to the extreme, so that users were forced to rate, for example, the beat, the guitar, the vocals, etc. when reviewing an album.
It was ridiculous and artificial, imho.
I think that well thought out, customized vertical based ratings can work, but can also raise challenges (how to weight various categories? who decides on the proper ratings categories? can the ratings platform easily absorb various structures, or are you limited to a certain number per vertical?, will all the fields scare potential reviewers away? etc.).
Integrating user profiles is also nothing new in this space, and is a no brainer. We've allowed readers of our reviews to click through to the profile page of the reviewer for almost five years now, and sites like Epinions have done the same - I think it's a critical piece towards giving the reader the tools required to reach a comfort level (or not) with the review.
And if the site has personal messaging available through the profile page, a demanding consumer like Webwork may even be able to get some of his very specific questions answered by someone who has stayed at the hotel. I think it's unrealistic to expect every review to meet the customized demands of every reader.
My "this is nothing new" skepticism aside, Yahoo is doing a fantastic job of using all the pieces they have available to them (photos, blogging, reviews, maps, yellow pages, IM, events, etc.) to create a very unique local experience. It's impressive.
| 10:24 pm on Aug 25, 2005 (gmt 0)|
I have to chuckle - one man's paradise is another man's hell...
What someone else has to say about a hotel has nothing to do with whether or not I will like it is basically true, in small numbers.
However, what I want to see is a profiling exchange where members of the exchange share (in an anonymous way) personality characteristics...once enough data is collected, when people offer reviews, you can see only the reviews from people who match your personality profile. Those reviews will be far more relevant.
Why does this work? A study done at the Univ. of Minneapolis years ago proved that your could take a random group of people and ask them to rate movies on a score of 1-5 (where 1 is hate and 5 is love). When you found a group of people that loved 3 movies in common with you, the probability that you would like a 4th movie recommended by any member of that group was over 80%. For statistical based recommendations that is pretty good.
Now imagine a local search engine that can review restaurants, hotels, etc. and make recommendations to you in a meaningful way. I want that local search engine.
| 11:52 pm on Aug 25, 2005 (gmt 0)|
welcome to WebmasterWorld and the local search forum, liceor.
stay around and share your thoughts. we appreciate it.
| 2:41 pm on Aug 26, 2005 (gmt 0)|
Hey Chee-ka-go. Thanks for the greetings.
happy to ;-)
actually - I find it hard to keep my big mouth shut...so this will be a lot fun...