Every year Google makes around 300 changes to the their algorithms which help searchers find the most relevant, highest quality content for any search query. Historically almost all these changes have brought more and more people to wikiHow, which makes sense since our goal for wikiHow is to produce the single highest quality page on any topic on the web. Unfortunately about a month ago, one of these 300 algorithm changes (known as Panda 2.2) significantly reduced the number of visitors wikiHow received from Google. The detailsIn the month before Panda 2.2, we had close to 36 mm unique visitors per month, the month after we are down to 25 mm unique visitors. This sets us back to roughly where we were around December, so we lost 7 months of growth. Sadly, the decline in our readership also feeds directly into a decline in the number of new editors we attract. Since new editors join us primarily after finding us in a Google search result. Given the ambitious goal of our mission, we need to take this decline in our traffic and new editorship very seriously. Ironically the first versions of Panda (Panda 1, Panda 2.0, and Panda 2.1) rapidly increased our traffic. It made me think we were being rewarded for our high quality. It was a surprise to get hit by Panda 2.2. On the good news front, a few days ago Google rolled out another algorithm change called Panda 2.3 which increased our traffic by around 10%. So hopefully next month should be a tad better still. (BTW thanks to the folks on this thread http://forums.wikihow.com/discussion/1382
who waited patiently for me to write this forum post!) Why did we get hit?The Panda algorithm has a unique twist that previous Google algorithms lacked. Google now penalizes *all the pages* on a site if also hosts just *some* low quality pages. In Google’s words: “low-quality content on some parts of a website can impact the whole site’s rankings, and thus removing low quality pages, merging or improving the content of individual shallow pages into more useful pages, or moving low quality pages to a different domain could eventually help the rankings of your higher-quality content.” http://googlewebmastercentral.blogspot.com/2011/05/more-guidance-on-building-high-quality.html
This is a particular challenge for a wiki. Since the goal of a wiki is to turn “trash” like the 2006 version of: http://www.wikihow.com/index.php?title=Make-a-Vodka-Watermelon&oldid=261985
into treasure like the current version: http://www.wikihow.com/Make-a-Vodka-Watermelon
While wikiHow has tons of great articles that we all love, it also hosts many articles we aren’t quite proud of . And the mere existence of those articles probably pulled down the rankings of the great content all over the site. What can we do about it?First and foremost, I’d say that we keep on doing what we have always been aiming for here: Our goal all along has been to produce the web’s largest and highest quality how-to manual. We want each individual page to be the single highest quality resource for that topic on the web. Second, I’d like for all of us to think of ways we can accelerate the process we turn “trash to treasure” faster. With that in mind, I’d like open a brainstorm in this forum about changes to wikiHow we should make which will more rapidly accelerate the quality of our articles. The web keeps changing, and we are going to have to keep changing with it. If you have ideas on how we can better improve our low quality articles faster, let’s hear it! I’m not especially interested in ideas which are meant only to trick or please Google. But any ideas that will help us improve the quality of our articles and make the experience of readers on our site better would be good for this brainstorm. Third, let’s not panic. We are on a multi-decade journey to build something the world has never had: A fantastic how-to manual on every topic. Setbacks like this are to be expected. We will overcome them.
Let’s brainstormI’d love to hear any and all ideas you may have on how we can improve the quality of wikiHow articles.
Hmmm… Even the worst articles can be fixed up to beauty as seen on some of our most popular articles, How to Do Nothing
, How to French Kiss
, How to Kiss
, How to Get Six Pack Abs
, and How to Lose Weight Fast
. All of these are featured, and the most popular articles on the site, which started out as stubs, nfd incompletes, and jokes. The stub and format categories are very high right now, and I think everyone could pitch in. One thing, a lot of stub articles may turn out to be better to be nominated for deletion as incomplete. As harsh as deleting is, it is sometimes necessary to create a better wiki. Remembering back in sixth grade, my teacher told me not to use Wikipedia for essays, because it was “unreliable”. I like to believe that Wikipedia, and furthermore wikiHow, are pretty accurate. While some of the things here I can not test, and do not want to test, (i.e. How to Survive a Long Fall
), many things I have found and used on wikiHow work well. (Although there was the time I didn’t follow the steps quite
right on a certain article.) Sorry for rambling off here, but I think we just need to continue following wikiHow’s original mission
, to create the world’s greatest how-to manual. If others don’t understand the cooperation and teamwork put into this, then that’s their
problem. (And they should look up on article on wikiHow about how to recognize that :P) I’ll try to help sort things out as much as I can. Am I missing something else unknown here? ~Andrew~
system
3
Quite. Well, I think instead of starting more articles for now, we should (as Andrew pointed out) fix the articles in need of it first. Personally, I think this is fun, I do it quite a bit. IDK how you other peoples think this should be done, but I’ve done it just by searching “horses”, and then clicking on the articles which look interesting, and from there re-writing, formatting, copyediting, categorizing, and adding pictures. REVISION: I should’ve said, writing articles is awesome too, but if you are going to write articles make them the best quality possible (without spending a lifetime on them, of course).
system
4
I sense a vocal opinion in a future conference call \o/ It’s really sad to see the up’s turn into down’s in readership, as well as editor rates. But as volunteer editors, we can all do so much. I agree with Andrew on pitching into the stub and format categories. Researching goes a long ways - not only that you’re helping an article grow, but you also learn something new that you haven’t prior to looking information up. I admire the people who use the quick-edit feature beyond the norm random spelling corrections or removing the extra step numbers - by actually revising the complete article (format, copyedit, etc), whatever is needed. Just a little “nudge”, I’d like to see new articles being done in the same format via RC instead of patrolling it “as is”. If you can’t, a fix-tag template is sufficient. Also, people needs to know how important patrolling for quality really is. In this time and through this issue, it’s sad to see things being patrolled when they’re not supposed to be, and that deprives the site in quality information and “assists the burden”. Every “job” on the site is much valued as another. If you can’t think of something to do, make small pieces of paper and pick something out of a bowl. Bored on something? Do something else.
system
5
Hm… We’ve already adjusted our article intake methods considerably from just a few years back. They now check for duplicates, check for length, and prompt the authors to preview and review before hitting the publish button… Is there anything else we can ask there? any way we can tweak our intake to screen out stuff with poor grammar, all caps or all lowercase, texting style spelling, or things of that nature? I think we might ask NABbers what their current fix up trends are and continuously adjust our intake to both make their jobs easier and to screen out more and more of the NFD worthy stuff. And with a couple hundred thousand requests in the wings, can we plan a method of plowing through those periodically to screen out anything obviously self serving to advertisers, or anything with horrible grammar, or anything that is a close duplicate of what we already have? That would save us some time later. I’m sure I’ll think of other things, but those two would make a good start for us.
system
6
@Loiswade42
There is the http://www.wikihow.com/Special:ManageSuggestions
as far as those suggestions that would be duplicates, joke topics, etc.
It seems that most of us are answering this question from the perspective that those of us who are reading this comprise the manpower which would be needed to a bring a huge range of articles (and requests) into better shape. I think one of the answers is going to have to involve finding ways to recruit and keep many more active editors. This is a big job and there are not all that many people reading this thread.
True. Citizendium is having the same problem when it comes to retaining contributors. Other problems are seemingly involved at wikiHow too. wikiHow offers a supportive and friendly atmosphere most the time and it used to offer far more real life how to articles. Lately it seems to be less practical fare and more “understand” or other articles that are not strictly how-to. As a volunteer organization there is absolutely nothing that can be done to fix this as WH can not assign articles to be written as there are no paid writers here. The only way to make it clear about what works is to continue to RS and FA WH articles written that match the organization’s mission. MA
On the other hand, one sure way to limit the number of problem articles, would be to stay on top of the NFDs. I will try to get back into checking into those from time to time.
KIS I’ll do what I can to vote on the NFDs to help move them a long.
Thanks, MA! Voting on the NFDs helps a lot. I am sweeping now, and finding that many of these go all the way back to May. http://www.wikihow.com/Category:Nominations-for-Deletion
@KnowItSome
@Mash317
KiS, Mary, it would be great if you could help trial the NFD Guardian tool. Discussion: http://forums.wikihow.com/discussion/comment/19404/
Tool: http://www.wikihow.com/Special:NFDGuardian
It is very unfortunate that wikiHow has been afflicted by a Panda system.
What leaps out for me in Google’s advice is their comment, “Does the site have duplicate, overlapping, or redundant articles on the same or similar topics with slightly different keyword variations?”
Well yes, that pretty much describes a lot of our subcategories for me. This is an area that could be concentrated on. The rest of their advice is interesting and I guess the backroom should be, maybe are, considering how this is ranked/calculated algorithmically.
system
16
I have a “trivia” open question, probably on the same lines as “high vs. low quality”. Where do we draw the line of the “it should be obvious advice” being kept? While I was adding links to http://www.wikihow.com/Make-a-LEGO-Garbage-Can
, I couldn’t help but notice this --_== Warnings == *Do not chew on legos or plastic objects: Choking or teeth damage may occur._I don’t know about you guys, but wouldn’t that be obvious to anyone older than 6 years of age? I also keep seeing things on recipe articles concerning “Don’t burn your hands” or the like. What is the targeted audience for this? 9 year olds or 36 year olds?
I have to be honest here, I stopped choosing wikiHow from search results when needing how-to advice, because I didn’t need to be told to ask an adult for help. http://www.wikihow.com/Special:Search?search=“ask+an+adult”&advanced=true
Wait, wait, wait. Doesn’t Wikipedia have stubs and NFD’s too?
system
19
Result 14:_Make Easy Peanut Butter Cookies (2,101 bytes) 29: *Be careful when taking cookies out of the oven. If you are a child and need assistance, ask an adult._THIS is what I’m referring to. IMHO, kids really shouldn’t be using ovens or the like to a certain age without parental assistance, so I’m asking, is it really necessary to have this in wikiHow articles? http://www.wikihow.com/Appreciate-Mice_==
Warnings == *If you are young, ask an adult to help you research. Always use Safe Search functions if you are underage._What does this have to do with appreciating mice? http://www.wikihow.com/Motivate-Yourself-to-Bake_==
Warnings == *Be careful using the oven *Ask an adult for help if you are younger._Again. I could really easily make a novel from this, lol.
*cough* We have a target age. *cough*
