Why today local search fails – and how to fix it

by on December 9, 2007

How today’s local search and social local websites fail to cater my needs and how to solve the problems and win the race.

I’ve learned in quite some years working in and for the Internet industry that it is always a good idea to check whatever you are going to do against the real life scenario that closest matches your online venture.

For instance if you want to build an online shop and make it successful then go out there visit the brick and mortar shops and have a good look around.

Dig deep till you know what makes them successful and try to port that to the online world.

The things that you might recognize are good customer service, good location, skilled sales people and so on and so forth.  

Now looking at what social local websites are trying to mimic or replace is the way we go about decision making for questions like.

-    Where do I take my girlfriend for dinner
-    Which plumber do I call to fix my dripping pipes 

2096823849 3209a4880a m Why today local search fails   and how to fix it

The social local websites – learn from others

Social local websites like Yelp or Qype do have a lot of followers and users. These sites allow you to write reviews and comment on businesses and places so that you can share your experiences with others.

Here you can learn from other peoples experiences about a new and great restaurant or an awesome plumber. And it may also help you to avoid making the same mistake that another person who hired an electrician that was just pricey and incompetent had made before.

Now that’s a great improvement to the old school yellow pages scenario – one might think – where the one business with the biggest ad, who simply paid the most or the one with the catchiest name that happened to be on the top spot of the page would always make the race.

But does all that really help us to make decisions and even more important to make better decisions?

Lets try and compare that with the real life scenario and see what we can learn.

In real life how do we really go about choosing the right handyman or the best restaurant for a posh night out?We will ask our friends and more generally people we know and trust to give us advice and suggest a solution to our “problem”

Now you can say that this is what Qype and Yelp are doing as well. But that’s only half the truth.

Everything else aside I see two big problems with this approach.

One is the problem with too much choice and information overflow.

I’ve been to San Francisco and a friend had booked a table in a restaurant. Being curios of what the place might be like I checked it out on the web and came across the yelp reviews of the place.

There are more than 460 reviews to that place on Yelp.

Reading this amount of information in the real world would mean that your friend would talk for approx 2.5 hours about this place and I reckon you would be confused and dizzy afterwards.

If I had had to choose the restaurant the dilemma would have even been bigger.

A search for a French restaurant near the union square in San Francisco returns 80 places to choose from with an average of more than 100 reviews each. So I could be ending up spending days to figure out which one might be the best for me.

How is that supposed to help me make a decision?
This is adding to my dilemma and not solving the question of where should I go.

That’s when I give up and go and ask the concierge of the hotel I’m staying for advice.

(Note: Barry Schwartz has done a great talk about the on how too much choice is a bad thing. Really worth the time to watch)

Imagine you go to a big drugstore and you want to buy toothpaste – they will have dozens of different ones to choose from. The Yelp approach to solve the problem of finding the right one is to read the information printed on the package and learn as much as possible to make the right choice – or better an educated guess.

You would still be left unsure if your decision is going to be the right one.

So in reality what has happened is that you loose time.

What we are all looking for – and that might be an illusion – is one answer to our question and only one.

And the same goes for all these social local websites. They might help you to make better choices but at a significant cost. Your time! – and quite a lot of it.

2096823605 9ca1867122 m Why today local search fails   and how to fix it


Just search to solve the problem!

So another way to try and solve my problem with the restaurant in San Francisco is to go and try a search for the answer.
And guess what Google Maps does provide me with more than 2.500 answers to my question not to mention the 220.000 that the web search does delight me with.

The experience with Yahoo! is similar – web search informs me about 1.2 million results and our local search has more than 190 answers to my question.

Even worse is that here the way to narrow down the results for both Google and Yahoo! is far less adequate and will hardly help me with getting to a more “digestible” amount of information.

So that failed for me as well. Why?

Well for local search there are a number of problems that lead to a rather poor user experience – well lets say felt user experience.

1.    The data basis to base the algorithm on for local search is poor in comparison to e.g. web search

  • The core data set.
    In a core business data set you will only have a business name a category and after geocoding long and lat to the business location.
     
  • The enhanced data set.
    For richer data sets you will have additional data like opening hours, price range and availability of parking areas.

  • User generated content
    UGC ratings and reviews of the business.

The only comprehensive set of data being the core listings there is very little information to base the answer on.

Compare that with all the factors that are taken into account for ranking websites and you see where I’m coming from.

2.    The lack of data is more visible than in web search

In web search the user assumption is “if it is not in there it doesn’t exist”. Which is fair enough cause that’s what we teach them.

In local search a user will often realize that Bobs Pizza place is not listed – even so it definitely exists.

So users will get that bad feeling of not getting all the information – and without the full story how are they to be convinced to find the right answer?

So how would the ideal website answer my question?

Well the same way a friend would do.

In a short conversation we would define the criteria the business would need to match and then I would get a maximum of 3 suggestions to choose from which would be ranked in a subjective manner.

DONE and I’m happy and good to go.

This post was written by...

– who has written 268 posts on LocallyType*.

Consultant at Valopex, Co-Founder Ypeas.com, Author & Speaker, ex Microsoft Bing Search Guy , ex Yahoo! Local & Maps / Answers, C64 Fanboy

Contact the author

{ 6 comments… read them below or add one }

Matt McGee December 9, 2007 at 9:16 pm

I do agree with you, Frank — but have you tried your same search at OpenList.com? The ability to refine a huge set of results is very much like the common way we all refine choices. You can refine by price, by user rating, by neighborhood, by cuisine, etc.

David Jaeger - Los Angeles Search Marketing Company December 10, 2007 at 5:49 pm

It sounds like you are saying that the problem with local search is lack of comprehensive data and the ability to get a quick answer.

However, I don’t think it is possible to generate this “utopia”, without having the search engine become more equipped to know your habits and preferences. With search, people have become acclimated to getting a consistent level of qaulity – relevant websites, with relevant content – that isn’t personalized.

Local search has to go beyond that, and learn personal preferences. Otherwise, we are hitting a scenario where you are getting the 3 places that the search engine chooses for everyone – without understanding the criteria(the highest ranked restaurant for customer service, when I really just want a quick affordable lunch).

Currently, if search engines would only pick the 3 most relevant restaurats, it would be like a lead gen company, that gives you three – of their choice of service providers, that match their standards.

Even if the company builds their brand around having quality service providers in their database, and provide service guarantees, would you prefer that over say, yelp or google maps -with their thousands of results? I may be in the minority, but I like to have as many options as possible. Those companies still rake it in (think servicemagic), but I don’t think that it’s the model of the future.

I agree that local search has to go beyond just the listings, enhanced data and UGC; and break it down into more specific measurements. E.G. for restaurants, they would have to rate it based on quality of food, style of food, speed of service, budget etc.

They would also have to extrapolate that data from current UGC – or train UGC reviewers to rate all of those specifics. Having a detailed database is critical.

But more than that, I think in your utopia, we would need personalized search… any company that wants to be like my buddy giving me referrals on food places better:

a) know my taste
b) have my trust
c) able to give me the answer in an easy format.

Translated to computer talk, they need to have built a database of my preferences in the specific area, and be able to give me the answer without asking me too many annoying questions. (I think that having to refine your search with check boxes etc. is lousy…it doesn’t reflect how you talk to your friend. You really do need comprehensive data on me.)

That means that the local search provider would have to have earned my trust for their search results as well… it also means that I would probably come back to it a heck of alot (meaning a bunch of recurring revenue for that company). So the first provider that comes up with an effective search methodology – and has it catch on, will probably gain the largest market share.

Maybe think general social network (has tons of personal preferences there – not so hard to build. Contrast that to yelp, which just has reviews, and not much of people’s personal preferences overall.) + local search engine… and the utopia might not be just a dream in the near future.

A great opportunity would be for facebook to integrate their business listings and personal preferences… They seem very intent on integrating users preferences into search – like with their recent fiasco with beacon. (I think it was a gutsy move into personal and social search, and although it was a PR hell-scene, it’s a start.)

Yelp and the other social review sites just don’t have a means to organically generate my prefernces in random areas… the main thing one does there is review businesses… and often, you are looking for advice on something you haven’t done before… so you haven’t reviewed other places.

Bill December 10, 2007 at 11:03 pm

You raise a lot of good points Frank, and a lot of issues involving both local search and local social networks.

Should local search be a recommendation system, or a complete directory listing of businesses, or both?

Should I be able to sort local businesses by:

1) Location and distance
2) Completeness of information about the business (such as parking availability, hours and days of operation, accessibility for people with handicaps, range of prices, contact information, etc.)
3) Broad recommendations that I can browse
4) (Trusted) recommendations by reputable reviewers as voted by other reviewers
5) Recommendations by reviewers who share common interests with me in specific categories, based upon factors that might be implicit or explicit or both
6) Recency of recommendations (I see too many old reviews that no longer apply to businesses)
7) Criteria about different aspects of a business that I can select, and give different weights to

Would local social networks benefit by allowing users to rate businesses by a number of criteria, including customer service, quality of goods, cost and value, and other criteria that might be appropriate for a specific category? And let other people rate those reviews?

I’d definitely like to be able to bounce back and forth between directory/search system and recommendation system, and be given a number of selectable options that involve specific criteria of my choosing, reputation of reviewers, similarity of interests, and some level of trust.

Acquiring enough data for a local search/social local network to work well is one of those things that I’m watching from the sidelines, to see how well it can be done. Sometimes asking the concierge and maybe a couple of other people is the best option.

Frank Fuchs December 10, 2007 at 11:33 pm

@Matt: I do agree that refinement is one step towards the right direction. And I’ll have a look at openlist.com but that is just one but right step closer to solve the problem. As said Yelp does a better job already cause they have a more adequate way to refine the results.

@David: Thanks for your comprehensive comment and for sharing your point of view. I’m totally with you regarding the fact that is a dream to think local search data could ever become comprehensive. Know my taste and have my trust is part of what Mr Sullivan would call part of search 4.0. This is clearly the way to go and based on gathering information like this and using them to refine the results is whilst still a dream one way to tackle the “problem”.

@Bill: I think you nail todays dilemma. It is very difficult to gather really comprehensive reliable data to start with. Then you want reviews and rightly you point out that you want recent reviews. As a 2 year old review is just not as reliable as one gathered 2 weeks ago. And yes ultimately going for a website to get information one time and talking with the concierge in another occasion is undoubtedly the best strategy today – regardless which site you would use to get a proper advice.

Overall the question is how is that going to change – if at all. And what would we need to do to really really change the game. As launching one review site after another seems to only provide a slight alternation to the sites we have available already.
I’m looking for the next big thing the change in the game…

sebastien December 11, 2007 at 3:09 pm

Really intersting article! Here are two remarks

1-In real life i choose my friends based on common experiences that i have shared with them over a certain period of time.
I trust them because i know them. So if i want to have a nice dinner in a new restaurant in Paris, i would rather ask my friends for an advice instead of screaming out loud in the street for it. It’d be less ridiculous and more efficient!
But among my friends, i would not ask all of them, i would choose the one(s) that i know share common cooking tastes with me.

In the end i might get 3 advices ranked in a subjective manner BUT those come from a long way. They are the results of the ability i had to choose my friends and grow our friendship.
Being able to reflect this real-life friendship scenario in the internet world would be a good way to filter out results and enhance any new local products.

2-In real life scenario, when searching for a nice place to hang out, it is not a matter of finding all the local restaurants BUT discovering the restaurants known by people i do trust which as i told can be friends, cooking specialist, neighbours, experts, cityguides…or people having my profile (family with a little kid), the tripadvisor feature enabling to filter out results based on your profile though being simple is quite efficient.

vinod August 24, 2008 at 5:30 pm

super points frank. I think you are bang on. Only thing in defence of local sites i will say is they are juts about 2 yrs old. ground swell movement has not resulted yet. Once thetipping point is reached u will see a lot more traction

Leave a Reply

{ 3 trackbacks }

Previous post:

Next post: