Help Wikitravel grow by contributing to an article! Learn how.
New users, please see Help or go to the Pub to ask questions.

Wikitravel talk:Business listings reliability Expedition

From Wikitravel
Revision as of 11:48, 14 September 2011 by Globe-trotter (Talk | contribs)

Jump to: navigation, search

Starting out

Per Denis' I took a stab at expanding this. It seems like there might be two good ways to get this going:

  1. Start gathering info about where things stand now - what documentation exists, what processes exist, etc.
  2. Start brainstorming possible improvements and ideas on this talk page. Anything from redoing documentation to eliminating business contributions to adding "verified" badges to listings, etc.

To the second point, a couple ideas that occurred to me are making article status more prominent (perhaps revisiting Template:Title-icons and doing a bit more with those) and possibly implementing icons that could be added to listings such that a rollover would show users who "verified"/"endorsed" the listing and when they did so, for example:

  • The Shiz, 1234 Shiz Street, +1-555-555-5555, [1]. With four fireplaces and a view of the Magic Mountains this place is great for a romantic evening, and the food, while pricey, is memorable. Verified by User:Wrh2 (24-Aug-2011)

The "verified"/"endorsed" icon would obviously benefit greatly from using a template, should probably use something other than a star, and ideally should change the image depending on the number and freshness of the people verifying/endorsing. -- Ryan • (talk) • 20:49, 24 August 2011 (EDT)

That little verified icon is awful cool! I might call it an "approved" icon instead, and maybe use a check mark? We would need to restrict its use, though, perhaps to autoconfirmed users.
I think a small template that we could place either in a section, such as "sleep", or even at the top of the article (ideally in the header) that would show the article was reviewed and deemed reliable, and would show who reviewed it and when, could be very useful as well.
I also like the idea of adding title icons to guide and usable templates (I assume that would be the best way of doing it?). --Peter Talk 21:23, 24 August 2011 (EDT)
I already thought of this "verified" listing idea, but I worked it out in a slightly different fashion (see [2]). However, your design is probably even better. The problem is who decides which listings would be endorsed, and how much of these listings we have. I'd say it should be limited to one per price category, else it'd lose its significance. --globe-trotter 15:54, 26 August 2011 (EDT)
I was envisioning this less as "endorsed" as in "this is the best" and more as "the people or person shown in the alt tag have verified that this listing is accurate as of the dates shown". I think it would be undesirable to have to try to come to a consensus on a single choice for any given article, although if ten people choose to "verify" a listing then it might be a good indication that it's a place worth visiting. Obviously some criteria would be needed as to who can "verify" a listing (must have been active for X months, have a user page, etc?) but that's probably a topic for discussion once the goals of this tool are established. -- Ryan • (talk) • 16:12, 26 August 2011 (EDT)

(Re-indenting) To summarize where this discussion stands so far, there seem to be three proposals:

  1. Create some sort of "verified"/"endorsed"/"approved" icon for listings.
  2. Expand the use of title icons to include things like usable/guide status. I'd also think docents, Wikitravel Press books, and other indicators of quality articles might be good icon candidates.
  3. A small template that we could place either in a section, such as "sleep", or even at the top of the article (ideally in the header) that would show the article was reviewed and deemed reliable, and would show who reviewed it and when, could be very useful as well.

Any other suggestions? Is there enough interest to start figuring out details of the above three and to then see if there is enough support to actually start moving ahead with them? -- Ryan • (talk) • 20:35, 29 August 2011 (EDT)

I think the list of three all have merit and are worth developing. For the listing icon, I would greatly prefer a verified category rather than trying to highlight the best. The latter is extremely subjective, the former is more of fact/quality check.
I really like the third idea. For example, recently I was delighted to hear that first time visitor to Nusa Lembongan used our guide and nothing else. Last year I had similar feedback about East Bali and its sub-articles. That's extremely encouraging for a writer as well as providing credibility for future users.--burmesedays 09:13, 30 August 2011 (EDT)
On #1, I would clarify: in addition to "added by X" we need "this listing [originally added by X] was verified by Y on date Z"
And one more suggestion--hopefully won't duplicate anything from the above.
4. What if we mark listings with "originally seemed to be added by business" and "originally seemed to be added by client"? Make no discrimination for some time, only to track how each type of source correlate with endorsements and non/validation later. For the cleaner experiment, we can even make no visual distinction--just add a template that expands into nothing when reading--but can only found when editing/diffing?
--DenisYurkin 15:40, 31 August 2011 (EDT)
I've added sub-sections below for discussion each of the three items mentioned above. I'm a little concerned about the "originally added by business" idea as that seems subjective - per Ikan's comment below, we already remove or de-tout anything that is overt advertising, so further trying to discern whether someone is a real user or a business trying to be sneaky might be difficult to do with any accuracy. For the "originally added by X", that's available in the article history - my hope would be that the "approved" icon would make it irrelevant who originally added the listing since we would now have other users giving their seal of approval. -- Ryan • (talk) • 13:13, 4 September 2011 (EDT)
*Bump* While discussions of ashrams and DotM templates are obviously grabbing the most attention right now, comments and suggestions for moving this expedition forward would still be greatly appreciated. -- Ryan • (talk) • 22:05, 13 September 2011 (EDT)

Approved icons

Following up from the above discussions, here are some issues that I think need to be resolved in order to move ahead with the above approved/endorsed template:

  1. Are we communicating that the users adding their names in this template have verified that the information in the listing is accurate, or that they are additionally saying that the information is accurate and they would recommend the business?
  2. What are the criteria for endorsing/approving a listing, and how is that enforced?
  3. What are the technical issues - icon, template, etc?

Regarding #1, I would suggest it means that the information is accurate and the person(s) listed would recommend the place.

Regarding #2, I would suggest the following:

  • Multiple listings can be approved - trying to choose just one or two would be really difficult.
  • Create a "Wikitravel:Approvers" (someone please come up with a better name) page that describes what it means for a listing to be "approved" and includes criteria such as registered user, will not approve a business that you own or work for, will not approve more than X listings per article section, where I'd suggest "X" is 2-3. The user should then add their signature to the bottom of the page. This ensures that users have read the guidelines, but is a fairly low barrier to entry for new users.
  • We include a comment with the "approved" template within listings such as <!-- please read http://wikitravel.org/en/Wikitravel:Approvers before approving a listing -->.
  • Any "approval" that is over two years old should be removed. Users are welcome to update the "approval" date yearly if they so choose.
  • If a listing changes significantly - for example "under new ownership", etc - all approvals should be removed.
  • It might make sense to limit the number of approvals for any given business to 5 to avoid having things get out of hand. If someone else wants to "approve" a business with five approvals then they should remove the oldest one first.

Regarding #3, a "checkmark" icon has been suggested, and I might propose that we come up with a few versions that could reflect how many people "approved" an entry (1-2, 3-4, 5, etc). The template can be an inclusion in a listing tag of the form <!-- please read http://wikitravel.org/en/Wikitravel:Approvers before approving a listing -->{{approved|User:Wrh2, 4-Sep-2011|User:Shiz, 3-Aug-2011}}.

Hopefully that's enough to get the discussion moving along... -- Ryan • (talk) • 13:13, 4 September 2011 (EDT)

While i general I agree with your proposals, I'm not sure if a Wikitravel:Approvers would be a good idea. I think it's a high barrier of entry for new users who just want to quickly approve a listing. I'd say just having a certain amount of edits (like 10 or 20) should be sufficient. I even hope in the future we'd have a technical solution so that a user can approve a listing with just one or two mouse clicks. Also, what listings are we talking about here? It seems mostly applicable to Sleep listings, but what about restaurants, bars and nightclubs? --globe-trotter 07:44, 14 September 2011 (EDT)

Title icons

Some ideas for expanding use of title icons:

  • Add an icon for articles with docents. Currently the left-nav link is more-or-less invisible.
  • Add an icon for guide articles.
  • I'd also suggest tweaking the display slightly to make the icons more eye-catching - suggestions for doing so are welcome.
  • Re-work Template:Title-icons to make it more flexible, so for example Template:Startopic could automatically add the icon without the need for an additional {{title-icons|star-icon}}.

Additional ideas/comments? -- Ryan • (talk) • 13:13, 4 September 2011 (EDT)

Reviewed templates

Per Peter's comment above, it would be helpful to have a unobtrusive template that could be added to article sections or full articles indicating when the article/section was last reviewed for accuracy, and who did so. Currently articles that are used for Wikitravel Press books get reviewed fairly often, but a user reading the article would have no idea that an extensive review had been recently done. The issues to resolve with such a template are similar to those for the approved icon:

  1. What does this template mean? This template should indicate that someone verified every phone number, URL, price, description, etc for a section? That all listed businesses are still open? Something else?
  2. What are the criteria for users to use this template, and how is that enforced?
  3. What are the technical issues? What does the template look like?

For #1, those who review articles for Wikitravel Press are probably best qualified to say what their current criteria are and we can then build from there.

For #2, I'd suggest something similar to the Wikitravel:Approvers process, where we have a page that outlines requirements and ask people to sign it before using this template. What those requirements are is up for discussion.

For #3, I'm sure there are lots of suggestions :)

-- Ryan • (talk) • 13:13, 4 September 2011 (EDT)

Who posted the listing

I would like to argue that any listing using "we" or "our" should be not edited but summarily deleted, as obviously self-interested and, therefore, unreliable. Like any other rule, this would be subject to exceptions on a case-by-case basis, but I think that whenever it's obvious a listing was posted by a self-interested person, whether because it's 100% obvious it's touting ("well-appointed rooms," et al.) or because "we" or "our" was used, unless other, non-interested people can vouch for the place, it should not be listed - and I mean probably never listed. Ikan Kekek 23:05, 24 August 2011 (EDT)

What's wrong reliability-wise in a listing saying "We had a dinner here and liked is so much that returned here twice during our week-long stay. We definitely enjoyed their T-bone steaks and cheesecakes."? Yes, we should edit first-person pronouns, but otherwise it's very likely a first-hand real-experience reviews--the only better thing in terms of customer-experience reliability could be only a summary of several real-experience first-hand reviews. Why remove them altogether? --DenisYurkin 17:20, 30 August 2011 (EDT)
You misunderstand. I'm talking about "We offer luxurious, well-appointed rooms," not "We had a great dinner" or "We liked this hotel." Ikan Kekek 17:54, 30 August 2011 (EDT)
I think it's fine to remove listings for this reason, but it should also be just as OK to keep it after detouting. While touts do come across as infuriating little dissemblers, most are just totally clueless. The owner of the Welcome Manor Inn in Bronzeville is a really cool guy who runs a great business. The listing he originally added, though, looked like this [3]—it now looks like one of the better listings in the article. --Peter Talk 16:29, 26 August 2011 (EDT)
now looks like one of the better listings in the article
You mean "one of the best-written" rather than "I tried many alternatives in the area, and this hotel is definitely the best in its price range", right? --DenisYurkin 17:20, 30 August 2011 (EDT)
I mean, there are quality of services and quality of description. In my belief, we should focus on former here--and only after succeeding with that, aim on the latter. --DenisYurkin 15:30, 31 August 2011 (EDT)
I even think business owners placing their own businesses are an asset to Wikitravel. I get plenty of spammy/touty hotel listings in the Bangkok district pages, but after de-touting they make fine additions to the articles. I don't see why we should delete these listings while they are useful additions in general. --globe-trotter 23:56, 26 August 2011 (EDT)
Most of my early edits here involved adding business listings, and I am sure I cocked them up big time :). I am of the same view as Peter that most business owners are clueless rather than deliberate abusers. There are of course notable exceptions to that.--Burmesedays 01:02, 27 August 2011 (EDT)

Here's the problem: Unless someone without financial interest in a business can vouch for its being a good one, there is good reason to distrust self-interested listings, even if detouted in style by a subsequent editor. What is your solution in the very numerous cases that an administrator has no knowledge of the establishment other than an obviously self-interested listing? I can detout, but I still don't trust the detouted listing if I am personally unfamiliar with the establishment. Ikan Kekek 02:14, 27 August 2011 (EDT)

There is no fool-proof solution, which is why we do not seek to be review site. Contact details and prices can often be verified by a website, but we cannot go any further than that, nor should we pretend that we ever could.--burmesedays 03:05, 27 August 2011 (EDT)
So what is this expedition about? Purely reliability of contact information and pricing, not descriptions in any sense? If so, is there very much to discuss? Ikan Kekek 04:10, 27 August 2011 (EDT)
A tiny percentage of business listings could be verified in the manner Ryan proposes I guess. However, there is no site-wide solution I can think of other than attempting to ban a business listing made by anyone with an interest in that business. Aside from the impossibility of policing that, there will never be consensus reached for such a draconian action. --burmesedays 04:29, 27 August 2011 (EDT)
So is your proposal to essentially give up, close the discussion, and go on to something else there's more consensus on? Ikan Kekek 04:39, 27 August 2011 (EDT)
 :=). Nope. I would be delighted if a practical solution was found.
Going back to your original suggestion about first person writing, it is already a reason for deletion which any editor has the discretion to enforce. I agree with Peter on that point.--burmesedays 04:51, 27 August 2011 (EDT)
It seems like right now, the touting policy is enforced either by: (1) deleting the listing more or less for good and all, (2) detouting it, or (3) deleting it with instructions and a chance for the offender to detout the listing him-/herself; and that which of these three reactions is taken is purely at the discretion of the editor, with no general case on which of these approaches one would normally take. I, too, have varied between these different reactions, depending on how bad the touting was and how much time I felt like taking to edit someone else's touty listing. My proposal above, though, was to establish a rule that the general case would be to simply delete touty listings and keep them deleted, with the two other types of reactions I list above constituting exceptions to the rule.
Is this worth discussing? Ikan Kekek 05:06, 27 August 2011 (EDT)
(Re-indenting) I'd suggest that we allow some discretion - if the person patrolling a listing thinks it is worth de-touting then they should be permitted to do so, as should someone who chooses to simply revert it. If the listing is being deleted I would suggest a guideline that feedback is provided for the user in question, either via the edit summary ("please do not write promotional, first-person listings per Wikitravel:Don't tout") or a talk page message; repeat offenders could obviously be reverted without additional comment. I don't disagree with Ikan that the reliability of a listing posted by a business owner is suspect, but it would be a shame to make Wikitravel even more anti-business since business owners can often make valuable contributions if properly motivated. -- Ryan • (talk) • 20:28, 29 August 2011 (EDT)
My actions are in fact along the lines you outline, but that's in keeping with policy here. If the policy were harsher, I would summarily delete and then not allow cleaned-up versions to be posted. Ikan Kekek 06:20, 30 August 2011 (EDT)
I'm in minority here, but I'm with you in proposing to remove business-added(?) reviews alltogether--rather than detout them. Because once we detout, we'll never know whether any single customer ever liked this place, or it's purely a product of its marketing. --DenisYurkin 17:23, 30 August 2011 (EDT)

Variants

Actions

Destination Docents

In other languages