The new All search in NOT broken

Phil,

THANK YOU! I am at the top of the rankings in the new searches in most of my categories...why? Because I took the time to learn how they work, made the necessary adjustments and it is paying off. Those categories I'm not coming up high in are simply because I haven't taken the time yet to analyze why yet.

I admit, working in web design I have an advantage many do not, just as you do with your experience, but you have done an excellent job of educating others. Well done.

Regardless, there are still some screwy things with the system that should be corrected by LL and the fact that my store wouldn't show up in places search for two weeks was a definite bug. So is the 5 page problem. Overal, though,l it's simply learning how the new search works.

Times and business change. You can make the adjustments to accommodate it or you can watch while the competition marches past you when they do.

They spoke of this on the blog, Phil.

They said it would only return five pages, and they would probably have a fix for it "in a few months."

That can be fatal for people who are making something extremely popular, like hair.

I think this is absolutely atrocious and completely unacceptable. Not to put too fine a point on it.

coco

WOW! Thank you Bradley

I'm really pleased that you've made it work for you. Hopefully others will too.

I think Bradley must have done that before this thread started, but that's the point of this thread. The All search leaves a lot be desired, but WE need to tailor our end of things to make it more useful for people, and for us. If we don't do it, then the results suffer.

When you're still on page 4 the bottom of the search shows pages up to 10, but on page 5 it acts like the last page. If you go back to page 4 and click on 6 you'll sometimes get page 6, and sometimes you get page 5 again (or maybe it was actually 6 but still said 5 with no further pages listed?). After a couple of attempts at getting beyond page 5 I just said screw it and closed search.

I agree. 50 is fine for most things, but not for items where there are lots of places in SL. It's odd that I saw the 50 limit change though. Maybe they changed it briefly, and didn't like what they saw for some reason.

I also saw something else that came and went and came and went - the use of operators such as 'link:'. I don't spend time watching these things but I did see that come and go and come and go.

This thread gets the award for "Only Useful Thread Currently On The First Page of Resident Answers".How many residents do you think will take the time to do that? Anything beyond page 5 is toast.

This is precisely why I got my landlord to separate the land into separate parcels and name my parcel for my store. This could mean a major change in the real estate market. Those that are capable and/or willing to do this may see themselves much more successful then those that don't.

Those that don't or can't, such as malls, may pass on or offer rock bottom rates. Again, one of those things that happen when change occurs.

Sorry, I was using imprecise language - what I meant is that the pick will result in the location being returned for additional non-relevant search terms.




No - but the location now displays for some additional search terms which aren't relevant to the location. Someone looking for a location those additional search terms may find that location to the detriment of a location which really was relevant.

A false postive in a search may push down the ranking of true positives.

If my top of the head figures are anywhere near accurate - only 30% of picks contribute to true postives, whilst 70% of picks contribute to false positives which could push the rankings of true positives down.





My experience with a Google appliance (albeit from a couple of years ago) was that it worked best with documents which were primarily (or had a lot of) structured text, and ideally a document which was part of a textual hyperlink network (i.e. had links to and from other documents whose significance could be determined by surrounding structured text). This is a very different environment to the one LL is trying to index.



Suppose I create 1000 alts, each batch of 100 designed to get a different location to appear for the term "SexGen". There is a real chance that by doing this I could force real SexGen stores off the first and maybe even the second page of results.



It is tuned for structured hypertext environments - SL isn't really that sort of environment.



Don't get me wrong, I think your post was a good one and may help things to a degree. However, fundamentally, I believe that LL is trying to get a square peg into a round hole - we can mitigate against that a little, and we can disagree whether the new search is better or worse than the old search (which also has its problems), but I don't think we can get around all the problems with the new search.

Matthew

Lol - well I think "most things" in SL ARE "most things"!

I mean, what do we mostly have: Clothing, Hair, Houses, Furniture, Weapons, Vehicles, a few other biggies maybe, that I'm not thinking of.

Now if it's something small and specific, sure! Like irons and ironing boards. Me and the few other people who make those are sure to come up on the first page!

(I would imagine. I haven't checked this yet.)

coco

Well, bang goes multi-level malls then!

Matthew

Well, I think a lot of people DO look further; I know I do.

In any case, though, we should at least get the option! Just obliterating anyone who doesn't show up in the fifty returns is INSANE.

coco

P.S. If Google did that, I would consider Google entirely useless.

Actually, the problem is more the number of pages than the number of results. Nobody clicks "next" after about five pages. In the old search, you could quickly scan a lot more results and eliminate the useless ones by name. This is more of a UI issue, but it's a crucial UI issue.

I was fortunate that I opened my business about the time that the new search came on line so I wasn't stuck in my old ways. I tweaked and worked within search guidelines, read alot in these forums and made adjustments as recommended. Basically, I did my homework.

I am lucky that I am a person that has always embraced change. My former VP said I like to change things for the mere sake of change. That gives me a major advantage when things like this come up because I may do a little whining here and there but I always take action as well. Part of the thrill for me.

Bottom line: New Search sucks! Period, end of story.

I agree with all of that.

I agree with that too. It's not an ideal search engine, imo, because it can't be used in the way that it's designed to be used.

I agree with the rest of your post too, btw

Not necessarily. Pick a popular mall... let's say Goddess of Love Mall. This mall takes up a whole sim. If they optimized title & description, and every tenant listed the keywords the right way in their prims for sale, AND every tenant put the mall in their picks (with description largely unchanged), it would still have power because of the power of numbers - of tenants and prims.

Instead of fighting for people to find you as a new skinmaker, you are part of that mall, which should come up high because of the number of inbound links to that mall thanks to the tenants. It shouldn't BE the death knell for malls. The numbers and inbound links should work for the malls.

To be blunt: yes.

To be more blunt: nobody older than a month TP's to places listed as "furniture,animations,sex,houses,shoes,hair,scripts,skins,eyes,sunglasses,art,fish, boats,iguanas,crabs,panties,penguins,biplanes,cliques,lawyers,guns,money,lions,tigers,bears."

Quality retail space rentals need to be their own parcel and the landlord needs to work with their renters to make sure their store gets listed appropriately. Easier to keep a happy renter than find a new one......

If it counted, there would have to be a link somewhere and there isn't one on the page for a group except for the group founder. In the list of links, groups never came up either for group-owned land so I'll personally go with "groups don't count" .

---

One thing people seem to be missing is that "optimize for the new search" only works right now because most other businesses aren't. Any benefit you get from that will turn out to be short-lived since eventually everyone else will catch on and then we're right back at whoever games search the most will end up on top.

I can also see someone automate the whole process: "pay this sign L$xxx, drop a landmark on it and 500 alts will keep you on their picks for the next 30 days unless you renew". Or someone with his/her own alts will subdivide their land to create 10 smaller parcels (10 picks/avie) so they can either flood search with identical results, pushing people off to the next page entirely, or they can keyword spam to a much higher degree.

It's the fact that search results can just be manipulated that makes it "broken" to me.

I don't know if it would work or not, Cristalle. I think the days of the mall may be over. Whether they are or not, though, it's refreshing to see someone being creative and thinking out of the box rather than whining "It sucks!".

Thank you!

lol but this is precisely why everyone thinks that search is broken - because these people are a large source of income, and they don't know where to go, and hey, one stop shop!!

I put this idea out before, but a themed mall that sells only one category (much like the fashion islands, or Hairspray) should be successful because all they sell is skin|hair|weapons|sex toys|etc. and can use the numbers built in with tenants. It really all depends on how you leverage the numbers.

I actually had "lions, tigers and bears oh my" at the end of my old parcel description, as I couldn't quite bring myself to be serious about it.

I've been suggesting the same thing for a long time as well. Combine your vendors into mini-malls and take advantage of the cross traffic. Put the appearance related stores all together and list it as a Fashion Center. Put Home & Garden together and list it as a one. Put avatar accessories together and list it as one. The point of search is to match up someone looking for something with the thing they are looking for - which might be a new skin or it might be a Goth nightclub.

I see a big win win with this idea. A big win with the smaller start up businesses who can't afford much land to compete with the big guys out there. Also a big win for the consumer who will have more options in one place then one big sim with items all by one maker.

The other big win is you'll have different creators competing against each other all under one roof which I would presume would improve quality for everyone.

I'm sold back on malls again...in this manner. Sounds like a great idea to me and one that can definitely take advantage of the new search.

LANDMARKS! Those are also IBLs that I forgot about until Kitty mentioned them. And what is their text? That land name again. More power to the land's name.

Yes and that's a very reason to do it now, and get a head start.

Doing things to improve search engine rankings has been around since soon after the first search engines appeared. It will never go away, and it's something that has to be accepted as long as search engines direct people to things. Unfortunately, search engines will leave a lot to be desired for the forseeable future, and improving ranking is something that a person must do if it matters to him/her.

This particular search engine can be improved for its users if people with places use the land's name and description effectively. It is helping both the engine, the users, and the place. The engine needs it to improve its effectiveness, users need it to better able them to find what they are looking for, a place may or may not need it depending on the owner's desired.

I don't think the search is "broken" so much as LL's logic on it is, heh.

The links may work. I think they'll be gamed as much as traffic, but we'll have to wait to see.

Personally, I'd like to see traffic redefined as "TPs/walks into the parcel" and get rid of the more points for longer stays. This would remove the parked alts (though moving bots may still be used *sighs*).

Add to that a time-stamp on the parcel. That is, the ability to sort finds by how long the parcel has been owned by the current account/group. Same with Classifieds; to be able to sort by how long they have been running consecualtively (however ya spell that lol).

The reason for the time stamp is two-fold. One, longer own buisnesses that have made a name for themselves will rank high, as they should, for sticking it out. However, one could sort the opposite way to see what "new blood" has come into the community.

A BIG help would be the ability to use multiple search critera. ie an Advanced Search. Where in we can select various fields for finds to be sorted by. EXample: Search for "Furnature", selecting oldest timestamps for the the longest run stores, unchecking traffic as "irrelivant", and limiting the search to ignore any find with the word "sex" in it.

The new search only upsets the system for the older consumers. The sellers who were gaming the last system will learn to game this system faster than people will learn to use this current system. It is a change yes, but I see no benifit in it. Simply a change. Sort of like changing the snapshot feature, or the Communicate UI. Change is good, but it ~should~ be for the better, not just cause it's different.

~Jessy