The conversation will take place with or without you

conversation I don’t often make “reactive” posts, but I did on Friday when I posted about local radio station Big Earl 96.3 switching formats and becoming Capital FM. I had noticed a lot of incoming traffic to an old post of mine about Big Earl, and decided I’d figure out why. I learned that the radio station “flipped the switch” on the new format that afternoon, and as a result, hundreds of listeners took the web to find out what happened. I posted about it so that others would learn the answer as well.

Now both posts are getting lots of traffic, and the new one has received a bunch of comments too. There are two questions to be asked here: why are people coming to these posts, and once they arrive, why do they comment?

My old post is the #1 result for the “what happened to big earl” search query, and my new post is #2 when you search for “96.3 capital fm“. Until today, it was actually #1, ahead of the radio station’s own website. The top five search queries that people used to find the posts yesterday were “capital fm edmonton”, “96.3 capital fm”, “big earl 96.3”, “big earl fm”, and “big earl”. In the last 24 hours alone, those two posts have been viewed more than 300 times.

So the reason that my two posts are getting lots of traffic is that they are ranked very highly in Google, and the reason people are searching is that they were given no notice about the switch. I guess that’s the way the radio industry works, you can’t really prepare people for a complete 180. As a result, lots of people were curious.

Once they arrived, why did they comment? I think the answer is very simple – Newcap Broadcasting simply isn’t participating in the conversation. Some listeners are happy about the switch, and they want to let the station know. Others are very unhappy, and they too want to voice their opinions. Aside from a very cumbersome “Members Club” section of their website, Capital FM doesn’t make it easy for their listeners to communicate. I think it’s a shame, really.

Like newspapers, radio stations are on the decline. Listeners are abandoning the airwaves for the web and iPods. And companies like Newcap aren’t doing much to reverse the trend. Which would you prefer – a radio station that suddenly starts playing completely different music than what you’re used to and basically says “tough luck”, or a radio station that changes its tune and also tells you to “have your say on our Facebook page?” It’s a no-brainer (even if your opinion won’t change anything, you’ll feel better about being able to share it).

CKRA has changed formats so many times now that you’d think they’d be better at it than they are.

It’s a different world than it used to be. Fifteen years ago, if a radio station switched formats, an article in the local paper would probably be about the only coverage it would get. Today, the web makes it easy for anyone to chime in.

As the comments on my post illustrate, the conversation will happen anyway. Newcap would be wise to join in.

UPDATE (4/1/2008): They’ve created a Facebook group! You can check it out here.

Searching Wikipedia Sucks!

Post Image Have you tried searching Wikipedia lately? Don’t bother, because you probably won’t find what you’re looking for! I am continually amazed at how terrible the Wikipedia search results are. Here’s an example of what I mean. Go to Wikipedia, type “al gor” in the search box, and click the search button. You should see something like this. That’s right, the top results are Al-Merrikh, Cy-Gor, Firouzabad, and Kagame Inter-Club Cup.

Absolutely terrible! If you type the same thing in the search box at Google, not only do you get accurate results, but Google prompts you with “Did you mean: al gore”. Why yes, I did! So why is searching Wikipedia so bad?

Part of the problem is that Wikipedia actually has two search modes: “Go” and “Search”. If you type “Al Gore” (spelled correctly) in the box and click Go, you’re taken right to the entry about Al Gore. If you instead click Search, you’re taken to a list of articles that contain or reference “Al Gore”. You can read more about searching Wikipedia here. So they’ve sort of complicated things by including two buttons instead of just one. The Go button is useful when you know the name of the article you want, but useless otherwise.

The other part of the problem is that the search algorithm just plain sucks. I know they don’t have a lot of resources, but you’d think that one of the most popular websites on the web could have a decent search feature. Matching “al gor” with “al gore” is a problem that has been solved for years, yet Wikipedia doesn’t even come close to accomplishing it!

Wikipedia itself mentions external search engines as a way to find what you’re looking for, but they aren’t really much better. For instance, if you type “al gor” at the special Google search for Wikipedia page, you do get the correct Al Gore entry as the first result, but the rest are not relevant at all.

So here’s where we’re at. Google knows that if you type “al gor” you probably mean “Al Gore”. Wikipedia knows about all of the entries that reference “Al Gore”. What we need is a way to combine the two! Is that really so much to ask?

If you know of a better way to search Wikipedia, please let me know!

How Google names products

Post ImageYesterday Google announced that they have renamed Froogle to Google Product Search. The change is explained on the official Google Blog:

Froogle offers a lot of great functionality and has helped many users find things to buy over the years, but the name caused confusion for some because it doesn’t clearly describe what the product does.

I don’t think that’s why they renamed it. I think Owen Thomas is right to point out that Google’s marketing is run by engineers. Froogle was/is simply a subset of search in general, so why not name it as such?

Think about it. They have Google, Google Image Search, Google Book Search, and Google Blog Search, so why not Google Product Search?

You could almost use the following rule for the way Google names products:

Is search the core feature of this product?
If yes then call it Google _____ Search
If no then call it Google _____

Obviously not all Google products fit into this rule, but most do.

Read: Google Blog

Powerset secures rights to search tech

Post ImagePowerset is back in the news again (you may recall they were ‘discovered’ back in October), this time for winning the exclusive rights to search technology developed at the famed Palo Alto Research Center in Silicon Valley. The technology essentially allows Powerset to understand the meaning of your search query (you know, “natural language” as they call it). Apparently Google is developing something similar. Here’s what VentureBeat says:

Clearly, Powerset faces challenges. Even if its technology does prove to be useful, it isn’t clear how long it will keep any lead (in natural language) in the face of an onslaught from Google. Another challenge is changing peoples’ search behavior, which is used to keyword searches.

Maybe I am being naïve, but I don’t think changing peoples’ search behavior will be all that hard. We still think in natural language before deciding what keywords to enter into the box. And some people don’t even bother to pick out keywords, they just type a sentence or question.

I think their biggest problem will be proving that their technology works and is useful. For now I’ve got Powerset filed under the “believe it when I see it” category. And assuming they really can do natural language search, will it be that useful? The keyword based search we use today works fairly well for general queries. I think natural language search definitely has value, but I don’t think it will replace Google overnight, if ever. There are certain types of queries that are probably better suited to keyword-based search.

And let’s not forget that millions of people (myself included) use Google and other search engines as navigational tools almost as much as information-finding tools. A quick glance at the 2006 year end Google Zeitgeist will show you that – half of the top ten queries were the names of websites.

Read: VentureBeat

Midomi worked for me!

Post ImageMichael Arrington wrote about a new startup called Midomi today over at TechCrunch, and he reported that he couldn’t get it to work. Midomi is a voice-based music search engine, which means you can sing or hum part of a tune and it will tell you the artist and name of the song.

I first saw a service like this back in 2003 at the Imagine Cup finals in Spain. I forget where they were from, but there was a team with precisely this kind of search engine. It worked pretty good too, if I remember correctly.

I just tried Midomi out for myself, and it worked great! I hummed the chorus of Madonna’s “Like a Virgin”, and it came back as the first result (the reason I chose this song is that I had to hum it on New Years Eve during a game of Cranium…and I did very well that time). Pretty impressive. Give it a shot, see if it works for you!

The only thing I don’t like about Midomi so far is the social networking features they evidently felt compelled to add. I think it is largely unnecessary for the site, even if it is the feature du jour.

Read: Midomi

Google Video searches YouTube

Post ImageGoogle announced today that it has integrated YouTube results into Google Video. This is the beginning of a transition for Google Video from hosting provider to search. Liz over at NewTeeVee wonders if this is necessary:

In thinking about video search, we’ve been concerned that with the huge number of videos coming into and and video streams coming out of YouTube, there would be little need for — well — video search.

I think there’s a huge need for video search. Just because most of the videos are in one place doesn’t make the search good or effective. There’s lots of things Google can do with it’s video search product to make it the destination. I’m thinking about speech and visual recognition to improve accuracy, and other really complex things.

When it acquired YouTube, Google got more than just a video hosting site. It got unfettered access to one of the largest test beds for video search around. That’s a big asset to have when you’re trying to build an excellent search engine.

Read: NewTeeVee

Wikiasari Search Engine

Post ImageThe Times of London is reporting that Jimmy Wales, founder of Wikipedia, is planning to launch a search engine next year in collaboration with Amazon.com. Dubbed Wikiasari, the search engine will allow users to rank web pages in an effort to create more accurate results (via Techmeme):

“Essentially, if you consider one of the basic tasks of a search engine, it is to make a decision: ‘this page is good, this page sucks’,” Mr Wales said. “Computers are notoriously bad at making such judgments, so algorithmic search has to go about it in a roundabout way.

It appears the big selling point of the search engine will be that it harnesses the wisdom of crowds. Google already does this, with PageRank, but in a less direct way. I am not sure if the new idea is going to fly – how many people really want to rank pages when they search? Usually you just want the results immediately. I’d bet most people won’t want to invest an extra few minutes to visit and rank the results.

I really have no idea what Amazon.com has to do with this project, but recall they too have their own search engine, A9.

Read: Times of London

Pluggd is cool, but not yet perfect!

Post ImageExaggerations make for good headlines, but often are less than accurate. VentureBeat ran a post last night about audio and video search startup Pluggd, announcing that the company has raised $1.65 million in funding. The post also says that Pluggd declared it has “perfected the user experience” for audio and visual search. This morning, I came across this NewTeeVee post that sets the record straight:

While we like what Pluggd is doing, that’s a bit of an overstatement.

Video search is often attempted by analysis of the soundtrack, rather than the picture, and we expect that’s what’s going on here. But mainly, we take issue with the claim that anyone has “perfected the user experience” in this area, because a big part of user experience is having a product that works.

Writer Liz Gannes goes on to explain the obvious – that speech recognition technology is far from perfect (though it is getting better). I’m fairly certain that audio and video search will be perfected eventually, but not we’re not there yet.

To their credit, Pluggd commented on the NewTeeVee post:

Matt’s coverage of our technology on venturebeat.com is one of the most thoughtful and complete descriptions of our technology that I’ve seen, but I was also a little startled when I saw the word “perfected”.

While they have lots of work ahead of them, I am sure the new funding will help Pluggd improve their offering! I am looking forward to their technology going live.

Read: NewTeeVee

What's a Yahoo! to do?

Post ImageAlmost every day now I read something about Yahoo! and its “problems” and/or “options”. Those are in quotes because it seems people are very divided on Yahoo! – some think it’s in trouble, others don’t. I’ve been a Yahoo! user since the days of the grey page-background, and if you count sites like Flickr and del.icio.us, I’m still a pretty active user. Allow me to put on my Yahoo! pundit hat for a moment.

I guess Yahoo!’s main problem is Google. Now that there’s a search-media company consistently outperforming Yahoo!, it makes them look old and stagnant. It’s actually pretty unfair, because let’s be honest, no one has the kind of growth that Google does. Yahoo! actually does pretty well in terms of search traffic, advertising dollars, and all that other stuff, but where they seem to be lacking is in respect.

So what’s a Yahoo! to do? Here are the most commonly suggested strategies I have come across:

Replace CEO Terry Semel
This suggestion is actually fairly new, and if you read Eric Jackson’s open letter to Yahoo!’s founders, it starts to make sense. Seems to me this is a relatively short-term fix though.

Buy AOL
Apparently Yahoo! has approached Time Warner about purchasing AOL. I think this would be a good deal for Time Warner, and a not so good one for Yahoo!. It would bring the failed AOL Time Warner merger to a complete end, but it would only provide a minor increase in Yahoo’s traffic and advertising, all things considered.

Buy Facebook
This rumor has been floating around for months actually. It might bring some more eyeballs to Yahoo!, but it would do nothing to help transform or improve the company. And besides, from everything I’ve read, Mark Zuckerberg (Facebook founder) is no Caterina Fake (Flickr founder).

Merge with eBay
The two companies might seem complimentary because of their completely different focuses, but that might present a problem rather than a solution. I agree with Fortune: I think this one is unlikely, because I think integrating eBay and Yahoo! would prove extremely difficult.

Sell to Microsoft
This one is my favorite, and it has a long history too, first appearing in June. Microsoft certainly has the cash, and it turns out that the two companies are fairly well-aligned – Yahoo! has made heavy investments into IE7, is a PlaysForSure supporter, and has hooked up with Microsoft on a number of initiatives ranging from Sitemaps to Instant Messaging. According to the latest comScore data (released today), a combined Microsoft-Yahoo would have around 40% of the search market compared with Google’s 45%. Of course, there are some easy to spot problems with this deal – mainly that Microsoft has invested heavily in Live Search and adCenter already. That’s not a total deal-breaker though.

Stay the course
The people that don’t view Yahoo! as floundering like this suggestion. Sure Google is #1 for now, but it can’t stay that way forever, right? Seems like this is Yahoo!’s currently preferred course of action. If they could somehow turn around their disappointing sales and profit numbers, this one might be the best option after all.

The Microsoft option is especially appealing to me, because it would have extremely broad ramifications for the industry. It also seems somewhat unlikely, given Microsoft’s huge investments in their online properties (MSN, Live.com, etc). That said, purchasing Yahoo! would instantly make them the leader on the web, a position they have long sought after. I wouldn’t be surprised if Yahoo! ended up staying the course though, and in the end, maybe that’s better for everyone – Yahoo! included.

Update: Here is more excellent commentary on Yahoo’s current situation.

Coming together to support Sitemaps

Post ImageAs much as I enjoy reading about how Microsoft plans to defeat Google and how Google has trumped Yahoo and started on their way to ruling the world, it always gives me a good feeling when I read about the three giants working together. Sitemaps are the latest technology that Microsoft, Google, and Yahoo have come together to support:

The goal of this effort is to improve search results for customers around the world. This protocol enables site owners everywhere to tell search engines about the content on their site instead of having to rely solely on crawl algorithms to find it.

Interested in the gritty details? Read more about the Sitemaps protocol at the official website: http://www.sitemaps.org.

As I understand it, Sitemaps do not replace they very common crawling algorithms, but instead augment that data and help improve the crawlers. Seems like something that should have been developed a long time ago! It’s amazing what can happen when you work together isn’t it?

Oh, and the coolest part of all – Sitemap 0.90 has been released under a Creative Commons license.

Read: Live Search Blog