|User Reviews and Local Search Optimization |
| 8:13 pm on Dec 4, 2004 (gmt 0)|
One of the most important trends in local search is the use of users reviews. It is the ability to aggreate user reviews, distance scores, meta data, and mapping through a pure search environment which differentiates new local search solutions like Yahoo!; keeping the tools sticky and useful for the user.
Reviews present a special opportunity for optimizers and a special challenge to the engines.
Did you know that the user can sort Y! local search results on the basis of user reviews? When we think of traditional seo we think of on-site and off-site optimization. In the world of local seo, user reviews are a critical factor of (sorted) rank, and therefor a concentration of local search optimizers.
When is the last time you checked your business or your client's business reviews? When is the last time you submitted a review? How about your competitors (scary)?
This is not a directive. In fact, one must be careful in doing this. It is however, a reality in todays new world of local search optimization. In Yahoo's own words:
"Yahoo! Local makes it easy to share your opinions about the businesses and services you've tried. Just go to the business you want to rate and click on the "write a review" link."
Hmm. At one hand this might be a new SEM service, tactic, and comparitive advantage. At the other hand this might be a quality control nightmare for the engines to moderate and control.
| 3:15 pm on Dec 5, 2004 (gmt 0)|
|Hmm. At one hand this might be a new SEM service, tactic, and comparitive advantage. At the other hand this might be a quality control nightmare for the engines to moderate and control. |
Even with heavy automated moderation, if you really want to manipulate ratings data, it's just not that difficult until you can grow a community which has a checks and balances system. User reviews just take up too much time for the companies to moderate, they must turn review quality to the actual users as well.
Everyone already knows how to manipulate Alexa data. For a while, it was beneficial to manipulate the Alexa data for several purposes (and some still do it to sell links/advertising so they can claim a higher visitor volume than actually occurs).
However, there are several communities which give us some hope in this area.
When user reviews first stated on many site, most of the reviews were written by the owners, authors, publishers, and friends for their products. This created artificial rankings for many products. This will be the first step we see in local search - the artificial rankings for restaurants. Even now, it's easy to manipulate Yahoo Local, as the majority of restaurants still consist of 0-1 reviews.
As community sites grew, more people wrote reviews, and products & services started being rated based on real reviews and not artificial rankings.
This still wasn't a perfect system as there was no review quality, and the data could be manipulated by people dedicated to making their product always have a high review. Then the 'was this review helpful' was born.
This helped publicly show what many people thought of the review quality. The checks and balances system has served many review sites well.
Amazon took this a step further with Listmania. The idea was good (and to my knowledge, there is no affiliate IDs associated with any products) and the lists were truly altruistic. Again, the major checklist for a good list is the 'rate this list'.
With local search and user reviews, we're going to see another growth process where at first, the rankings are easily manipulatable. The question will be, can these local search engines/portals bring together a community that will impose checks and balances so that the reviews lend true quality to the community.
| 8:02 pm on Dec 5, 2004 (gmt 0)|
Using Amazon's review system, the best way we found to get good, natural reviews was to ask our happy customers to post one. In some cases, we offered a perk of some kind, especially for those listmaniacs who took the time to put together a list. This worked very nicely and created natural positive reviews - avoiding the stilted sound that people who have a business involvement with a product tend to use.
So I would imagine the same would apply to a pizza parlor, a general contractor, a real estate agent - anyone who already has happy campers in their corner and uses local search. And of course, if you don't make customers happy (or don't know if you do) then that would be step number one for the business anyway.
One of the key differences in local search would be the need for a physical business to exist, right? Online "pure-play" just doesn't enter into the local search picture, I would assume.
| 8:20 pm on Dec 5, 2004 (gmt 0)|
|One of the key differences in local search would be the need for a physical business to exist, right? |
Would a mail drop and a local phone number take care of that?
How is a search engine going to know if I've got a 10,000 square foot store, or a 4 inch mail box, as long as there is a local phone and street address attached to my info?
| 9:14 pm on Dec 5, 2004 (gmt 0)|
>>the best way we found to get good, natural reviews was to ask our happy customers to post one.
great pointer, tedster.
>>One of the key differences in local search would be the need for a physical business to exist, right?
it is very important when talking about local search to differentiate between tactics/or local search utilities.
for example, we all know that anyone can optimize in the traditional serps for geo qualifiers regardless of a physical location. we also know that on google adwords, one can do the same. overture localmatch on the other hand will require a verifyable local address, so too will some IYPs like yellowpages.com.
i think what we are referring to here is the new pure local search tools from G and Y....
>>How is a search engine going to know if I've got a 10,000 square foot store, or a 4 inch mail box, as long as there is a local phone and street address attached to my info?
the new pure local search tools from G and Y go to great lengths to verify a physical location. how do they do it? y ranks according to 'traditional' relevance but their relevance scores are based upon a core set of traditional data beyond the web, including infoUSA data. infoUSA is one of the largest data aggregators in the country. they have been around for a while, and got their start providing targeted traditional mailing lists. they aggregate rboc (regional bell operating company) data and they also have a team of telemarketing people who are continously working to verify information. even when submitting new info to infoUSA, they will take formal steps to verify the legitimacy of the data.
in the case of google, it is a little different, but equally difficult to manipulate. In G's words, "Google Local locates neighborhood stores and services by searching billions of pages across the Web, then cross-checking those results with Yellow Pages data to pinpoint the local resources you want to find." G unlike other local search providers does not disclose their YP data providers, but in this arena, there is only a handful of important sources. when you run a search on G local, you will see a link to referenced sources. the referenced source is an example page of where they are finding your business data. the sources are dominated by geo-vertical directories, and other geo-vertical based reference guides. google uses much of their traditional algo, specifically pr when crawling the web looking for local data to index and rank (although speculation is also abound about IP/hosting location of domain and links). They then cross reference the data for accuracy with the major YP data aggregators. then in their display mode, they will add a proximity score to the relevance score and present a result set. yet as google states (in their altruistic way) "Sometimes our search technology will decide that a business thatís farther away is more likely to have what you're looking for than a business that's closer." This is where traditional on-site and off-site optimization come into play.
>>so how is the engine going to know...
it depends on how determined you are to present comprehensive and consistant data about your business in a formal and informal setting.
this particular thread is about user reviews, and the reason these points may be relevant is because many users seek local information not on the basis of proximity, but on the basis of reputation. because of this fact, y permits you to sort results, as many do, on the basis of user ratings. it is here that for the near future, many will try to overcome their proximity deficiencies by promoting their ratings. many astute business owners will do so through concerted, even unnatural efforts.
<Added a couple of days later>
A different day today relative to getting listed on Y Local
[edited by: Chicago at 6:08 pm (utc) on Dec. 10, 2004]
| 10:30 pm on Dec 5, 2004 (gmt 0)|
|the reason these points may be relevant is because many users seek local information not on the basis of proximity, but on the basis of reputation. |
This presents a huge opportunity in some markets. Huge!
| 7:09 pm on Dec 6, 2004 (gmt 0)|
Chicago; I hope you'll pardon my ignorance on this issue, but I have a question, hopefully it's not off topic.
The "user review" is something I'm not experienced with, as it relates to local search.
I did a little searching this am as an experiment to see if I could figure it out better.
In the process, using Googles local search, I found a number of listings that had niche directories as a reference. Not having used local search to any noticable degree, that surprized me somewhat. It seems like in other threads the underlying local search info sources mentioned were all fairly well known larger outifts.
Some of these niche directories allowed visitors to write reviews or recommendations.
So the question is, as regards reviews, how large a role do you see for niche directories?
| 7:53 pm on Dec 6, 2004 (gmt 0)|
>>Chicago; I hope you'll pardon my ignorance
ken, anyone who claims they know it all in local search really knows nothing at all. even being in this space for years, it seems that every other day something changes. this change is an opporunity for us but it requires that we all are in learning mode.
>>The "user review" is something I'm not experienced with, as it relates to local search.
the most prominent use of the review can be found on the new pure Yahoo local search. there are many other examples. it is worth taking a look at what lawrence is doing over at -rate it all- as well. beyond the ratings, take a few peeks at the traditional serp value over there.
>>In the process, using Googles local search, I found a number of listings that had niche directories as a reference.
yes indeed. they crawl the web for local data then cross reference it to yp data.
>>underlying local search info sources mentioned were all fairly well known larger outifts.
they are a portion (a trusted, aggregator portion) of the equation - especially for backfill in iyps. there are many other sources of local data, including the directories you mention.
>>Some of these niche directories allowed visitors to write reviews or recommendations.
>>So the question is, as regards reviews, how large a role do you see for niche directories?
this is an important thread topic worthy of its own thread ken. my short answer is there is a role, especially for geo-vertical specialists that can aggregate rich meta data. this is something that the majors will be seeking over time and in many cases buying.
| 9:05 pm on Dec 8, 2004 (gmt 0)|
Like Chicago, I too believe that user reviews are going to be a make or break feature in the Local Search space. When user reviews work, it's a beautiful complement for local search. Why find a merchant based on alphabetical listing when you can find the BEST merchant based on the experiences of your community?
Speaking to the challenges of being a publisher of user reviews of local merchants, I would say that there are five principle areas:
Managing Rating Manipulation
With any user review system, there will be manipulation - especially with those user review systems that offer a clear financial reward for successful manipulation. The key defenses a site has against manipulation are automated tools (one vote per IP address), Mod / Administrator intervention, and the intervention of an active user base. However, the biggest protection a site has against rating manipulation is data. The more data, the less the impact of outlier manipulators. Unfortunately, the impact of manipulation is even more severe in the Local Search space because of...
Non Mainstream Subject Matter
Attracting user reviews for mainstream subject matter such as the movie, "The Shawshank Redemption" or "Britney Spears" is not difficult, as the universe of potential raters is huge. But how huge is it for Joe's Drycleaner down the street from your apartment? Attracting representative reviews of a merchant when there is only a rating universe of a couple hundred people is EXTREMELY difficult. Because of this, the eventual winner of this space will have some sort of a push strategy in place (much like BizRate's rate this merchant pop up windows from back in the day).
It is a pricy proposition to maintain accurate listings (phone, address, url) for local merchants around the country. This data decays rapidly as well, so to stay current you really need quarterly updates. Even having a bot at your disposal like G or Y doesn't protect you from having to spend to keep your data fresh.
Any time you are allowing anonymous users to post content that may affect the livelihood of the merchant being rated, you are exposing yourself to legal risk. Anyone who wants to play in this space needs to be prepared for this.
Achieving Critical Mass
For a rating community to take off, you need two things 1) existing ratings to seed the service; 2) an active user base that likes to share their opinions. These things take time - there is just no rushing into this space. For more info on this topic, check out RogerD's Building Community forum.
[edited by: Chicago at 2:50 am (utc) on Dec. 9, 2004]
[edit reason] removed url [/edit]
| 1:23 am on Dec 9, 2004 (gmt 0)|
thanks for your experienced insight into user reviews. much appreciated.
do you have any specific comments on Y local's use of reviews? what challenges lie ahead for them? sorting and optimization?
keep up the good work:)
| 2:29 am on Dec 9, 2004 (gmt 0)|
Nice summary Lorenzoinho2.
Banning by IP runs into problems with AOL, doesn't it, since AOL randomly assigns IP addresses?
Worse than that will be the local webmaster, who built the Donnie's Drycleaners website that just was trashed by a competitor. He/she will create accounts using an anonymous proxy and start to apply gas and matches to the rating fire.
Absent a very large local populous, ratings has all the markings of a porcupine or skunk in the making. I predict a bad end to it. I predict the outsourcing of reviews to Malaysia as a growing business model. Oh joy!
OBTW - The link drop? Tsk, tsk, tsk! Thou shalt not take advantage of the new mod Lorenzo!
[edited by: Webwork at 3:04 am (utc) on Dec. 9, 2004]
| 2:56 am on Dec 9, 2004 (gmt 0)|
OBTW, having met and talked with Lorenzoinho2 in Las Vegas and having seen his "rate it all" site in operation I've got to tip my hat to him for sharing his 'expertise by experience' in this area. Thanks for the contribution and please take my "Tsk, Tsk" in the spirit it was intended: A friendly jab for (momentarily) getting one over on the new guy from Chicago, who is so naive about the ways of the WWW. ;-P
| 8:01 am on Dec 9, 2004 (gmt 0)|
No worries WebWork. I guess I sort of thought it was on topic :)
Regarding user reviews and optimization, I think things will really get interesting when Yahoo decides to include merchant rating as a ranking factor for their default, unsorted local search results.
There would be no better way for them to spark the growth of their user review database than to let it be known that its a factor in how sites rank.
This would get local merchants familiar with the world of user ratings in a hurry, and would incentivize them to do the grunt work of seeding Yahoo's user reviews.
Cause the funny thing about user reviews is that nobody wants to leave one if there's not one there already. And from that perspective, who cares if this first one is a little on the biased side?
Another thing Yahoo has going for them is the ability to create the perception of transparency. When people read a user review, a typical reader reaction is to want to learn more about the poster. Yahoo can provide this info via user profiles, many of which are already public. Not sure if they're doing it already, but you can bet they're thinking about it - especially as its an area that Google Local could not compete.
| 1:39 pm on Dec 9, 2004 (gmt 0)|
I participate in the Yahoo Finance message boards. If the Y! Finance message boards are any indication of what's to come I rest my case for the proposition that the system is irreparably broken before being built.
I know of 1 microcap stock where there's no question that just 1 person employs at least 50 IDs (or more, not me.) Why do people on Y! Finance use multiple IDs? They do it to create the appearance that there's a consensus of opinion about a stock going up - or down, depending on whether the post is buyin or selling.
Could Y! do something about the multiple ID issue? Well, they could start by IP limiting registrations. Do they? Nope. Why? Well, I can think of at least one reason: They more times a person logs in and posts the more Y! gets to display banner ads for which they are likely paid on a CPM basis. So there's actually a financial incentive for Y! allowing a person to log in 100s of times - as different screen personas.
Will this get worse? In the finance realm it's hard to imagine that it could get any worse.
It's an interesting concept but I'm pretty certain that the 'acceptance curve' - of people relying on reviews - will be an up THEN down - way down - trend as people learn the system that they figurd was rigged is indeed seriously corrupted - unless some major patches can be installed.
Perhaps, if a website can create a credentialing process for it reviewers, the system's credibility - and therefore its utilization - will be salvaged.
| 2:03 pm on Dec 9, 2004 (gmt 0)|
I'd like to throw another spanner into the mix.
A lot of the traditional directory companies, Internet directory companies and including I believe the YPs are either now, or going to begin to, sell coverage area advertising whether or not a company has a physical presence in that area. All that is required is that the company be able to service that area.
In the B2B community, I believe this will have a big effect on 'local advertising'. I am not too sure how much this will effect B2C.
| 5:41 pm on Dec 10, 2004 (gmt 0)|
Webwork, it seems like a solution to your problem would be the following:
Social network + User Reviews = the user only getting the reviews of people that they value the opinion of (their friends).
It seems like it is a few steps away but at some point it will probably happen.
| 5:57 pm on Dec 10, 2004 (gmt 0)|
This is possibly one of the most important posts on WW as it highlights the growth potential and growing pains local search will run into in the next few years.
Could it be a case that the original review / information could be taken from trade associations local chambers of commerce and trading standards type organisations plus YP etc.
This is the first time I have had any understanding of what local search could do to the search landscape so thanks all those posting here
It also highlights how Yahoo and MSN could make large dents in Google dominance
Yahoo due to having been a portal for the longest time and MSN from the email and MSN messenger usage
| 6:03 pm on Dec 10, 2004 (gmt 0)|
>>Social network + User Reviews = the user only getting the reviews of people that they value the opinion of (their friends).
Friendster, Eurekster Team Up for Personalized Search [searchenginewatch.com]
|Results can also be influenced by your personal network ("friends," "friends of friends," and so on) or all Friendster members. |
Search results are prioritized with results viewed by anyone in your personal network appearing at the top of the list. These results are highlighted with a smiley face icon.
Its a good read on the topic with many of the topics we are addressing including quality control challenges.
| 6:09 pm on Dec 10, 2004 (gmt 0)|
>>Social network + User Reviews = the user only getting the reviews of people that they value the opinion of (their friends).
Epinions has done some cool stuff in this area as well with their "web of trust."
| 6:10 pm on Dec 10, 2004 (gmt 0)|
Some communities allow the users to rate other users - ie the slashdot model - people who are consistently guilty of "poor participation" in the community drift lower and lower in the ratings and their comments are filtered by the default setting.
Once you have several thousand people rating other users on a daily basis, even someone with 50 user IDs and consistently useless posts will fade into oblivion.
| 6:11 pm on Dec 10, 2004 (gmt 0)|
Web of Trust [epinions.com]
| 6:45 pm on Dec 10, 2004 (gmt 0)|
Just an aside: Anyone have reason to believe that a surge of spurious patents will kill the golden goose that we are discussing? I can just see the patent application: A system whereby people who know other people connect with one another via an electronic medium to exchange information . . .
| 9:12 am on Dec 11, 2004 (gmt 0)|
|This is possibly one of the most important posts on WW as it highlights the growth potential and growing pains local search will run into in the next few years. |
Conceivably, this is not just about local. Shopping search and online social networks have been mentioned, and they are not necessarily geographically defined. This might extend beyond metropolitan borders.
Some random comments...
I'm glad to see this. There have to be some alternatives ranking by distance from the center of a zip code.
It's inevitable that User Reviews, like links, are going to be manipulated. Before the Web, the "Best of..." type listings in local newspapers were obviously manipulated, and they still are. Rating the raters (possible online but not easy for newspaper votes) just adds an extra level of difficulty.
As lorenzinho2 points out, a larger sample size tends to give better results, or at least a better representation of the common denominator. I used to think that the larger the Google index got, the harder backlinks would be to manipulate... and, who knows, this may yet come to pass.
With local search, there's probably a point where the manipulation isn't worth the financial return, so we'll perhaps only see it big time in highly competitive local search areas... just as the really big linking networks in Google are generally applied to the big money areas.
I can see the hotel reviews being cranked out soon. The chicken and egg question is whether traffic or reviews come first, and how to convince a company this is worth the effort before the traffic or critical mass is there.
Also, how to pitch this without leading prospects right to it?
| 10:24 pm on Dec 11, 2004 (gmt 0)|
PS to the above...
|Also, how to pitch this without leading prospects right to it? |
In the wakeful light of day, I see that the indefinite antecedent "this" in the above can be misread.
By "this," in the context of this thread, I mean some sort of coordinated solicitation of reviews from customers... not the cranking out of fake reviews in sweatshops using different IP addresses around the world. I can see the latter happening, but I can see enough difficulty, compared with link swapping or outsourcing of link development, that it may not be worth the trouble.
| 1:05 am on Dec 12, 2004 (gmt 0)|
Robert, thank you for your thoughts. We appreciate your involvement in the local search forum, and we know that you can add seasoned insight into the topics that we are addressing.
>>how to pitch this without leading prospects right to it?
As SEMs we need to begin to understand that search extends beyond traditional on-site and off-site (pr) optimization on Y and G. Search is here to stay, but what will change is how result sets are served.
Gone are the days of one search fits all. Search is evolving and splintering. Search is getting personal, topical, and qualitative.
How will you pitch this? User reviews and ratings are a significant portion of how local search result sets are determined. What more is there to say?
Imagine speaking to a client and telling them that you will formulate a user review strategy for their business accross dozens of search resources. That you will monitor their reviews and ratings for inappropriate activity and will report it. That you will take advantage of current market timing to seek top rankings. That you will *clean* the default 10 tradition serp result sets and authoritative resources with favorable and useful information about the business including press releases, reviews, ratings, clean boosted date, site content, rich meta data, etc.
Imagine mainstream SEM two years from now. It is not just O and G PPC and Y, O, and MSN SEO. No way. Especially not for 22 million small and medium sized US businesses struggling for local market share.
>>I mean some sort of coordinated solicitation of reviews from customers
Tedster talked about it early in this thread. Think about your top 10 clients. Call them on monday and ask them to give you a rating and a review. It will buy you a three month+ lead in the marketplace.
Turn around and tell them how this movement is affecting them. Then tell them that you will clean and boost it for them.
| 7:18 pm on Dec 14, 2004 (gmt 0)|
Great stuff going on in this forum. Glad to see it going on.
>>how to pitch this
I think this begs the issue (perhaps fodder for a different thread entirely) of how prepared the local business owner is to comprehend the multi-layered issues of how/why to do Internet Marketing.
Joe's Dry Cleaner might not know his AOL screenname password from a hole in the wall, let alone want to decipher the ups and downs of user reviews, etc.. etc.. Even if Joe has a nice website that can enjoy SEO, the pitching becomes a potential quagmire of over-information. I think the true local business is many more steps away from being SEO savvy than pure plays and the like.
This is not to say that explaining the subtleties is not effective. Sometimes presenting the full story with a certain amount of dumbing down seems to help a "Joe's Dry Cleaners" reach a point where they can make a forward moving decision to engage with an SEO firm.
Sorry if I'm veering off topic (and in my first post none-the-less).
| 8:07 pm on Jan 12, 2005 (gmt 0)|
Interesting nonprofit called Checkbook.org [checkbook.org]
Looks to be trying to be the Consumer Reports of local merchant reviews.
I found out about them when I got a piece of mail from Consumer Reports inviting me to share my reviews of local merchants for Checkbook.org's service.
Anybody know these guys?
| 3:11 am on Jan 13, 2005 (gmt 0)|
If you think about it, reviews can make or break you in any aspect, from a consumer standpoint, not simply local search. Take Amazon - as a customer, I can't tell you how many times my buying decision has been affected, either positively or negatively, by reading the reviews of a product.
As a local search user (I use it 10 times a week or more) - a business with a good review is likely to stand out to me - and if the only difference between A and B is that A has a good review, guess who is going to get my business.
As an seo, I can understand the difficulty user reviews can present - abuse, legal ramifications, competitor sabotage and the like. I "know" that reviews can be faked and worthless as an seo, yet, as a consumer, I still will put some level of stock in them.
>>>How is a search engine going to know if I've got a 10,000 square foot store, or a 4 inch mail box, as long as there is a local phone and street address attached to my info
You've got a point there ken_b - see the bottom of message 44 in this post [webmasterworld.com] (supporters link).
>>>Social network + User Reviews = the user only getting the reviews of people that they value the opinion of (their friends).
Yea - sorry to defer to Amazon again, but they really have a review system that works as well as any I've seen over the years. Amazon allows regular users to decide if reviews are helpful and as a result, the reviewers "status" can go up or down. When I as a consumer read reviews, someone being a top 1000 reviewer can possibly hold a bit more stock for me. I think this could be manipulated some way by LS engines by making "ranks" for reviewers that relate specifically to a single city or even metro area and dubs them with a special title once they reach a certain level of "trustworthyness" as judged by the regular users of the engine. In turn, it could spark interest in people to take the time to do reviews in order to become a symbolic "leader" of an area. Hopefully that made sense. It does in my head, but it's late ;).
Finally, a little off topic, but an older article with more background info on Eurekster [searchenginewatch.com] to add to the link Chicago gave above.
| 8:50 am on Jan 13, 2005 (gmt 0)|
|Amazon allows regular users to decide if reviews are helpful and as a result, the reviewers "status" can go up or down. When I as a consumer read reviews, someone being a top 1000 reviewer can possibly hold a bit more stock for me. |
This sounds like a kind of PageRank in different clothes. I feel there's a strong implicit analogy to PageRank that's been running through a lot of this thread, and in fact through many of the discussions on this forum about local search. The same principles apply... but to different factors, and in an evolving algo that's not yet very mature.
Brin and Page...
|Academic citation literature has been applied to the web, largely by counting citations or backlinks to a given page. This gives some approximation of a page's importance or quality. PageRank extends this idea by not counting links from all pages equally, and by normalizing by the number of links on a page. |
It may be helpful to consider reviews as citations or links too. Ultimately, as local search databases and the elements that feed into them grow in size (note, I didn't just say sites feeding into these databases), there will be networks feeding into networks, and the local search engines will give the most weight to those factors that help them achieve user satisfaction.
(User satisfaction itself might be considered part of a review network of a different sort.)
I think that a critical mass of reviews is going to be key to really meaningful results, and I agree with lorenzinho2 that it's going to build slowly. Froogle is still struggling with critical mass for many searches. Big city hotel reviews, eg, will probably evolve more quickly than blurbs for barber shops. When the word gets out, there may be cottage industries of direct marketers with a little bit of web experience canvassing neighborhoods and spamming Yahoo.
I'll be really curious to see how existing niche review sites (like Zagat) are going to factor into this... whether they'll develop into little islands of their own, or whether Google might work out how to spider Zagat's data in a way that would be advantageous to both companies.