Keyword Search Is Just Fine, Thank You
Ian Lurie Apr 25 2008
Eric Shonfeld at Techcrunch just posted Is Keyword Search About To Hit Its Breaking Point?
So, are we headed for a search apocalypse?
Better hope not. And I don’t think so.
Eric cites Nova Spivack’s presentation. In it, Nova says that if the amount of content keeps growing keyword search will collapse, leaving the internet in a general state of higgledy piggeldy (We’re talking flying monkeys here, people). Therefore, we need something better than keyword search.
Any chance all we need is better keyword search? I sure hope so, because the semantic web will never happen. Not in my lifetime. Not in my kids’ lifetime. Not before the sun goes out.
Keyword Search Is Pretty Darned Good
OK, it’s true. I used to get some pretty annoying search results when I searched for information on the Bushtit.
But keyword search offers an ideal balance of simplicity and accuracy that satisfies 99% of internet users.
Nova’s dire prediction looks scary:
But really, if keyword search continues to improve over the next 10 years as it did over the last 10, I don’t think the graph above will come to pass.
Also, if they start using the phrase Web 3.0 in 2010 I’m going to drown myself. Can we finish Web 1.0 first?
The Semantic Web Won’t Work
Eric and Nova point to the semantic web as the answer. But even Eric points out that “it essentially requires every single Webpage to be re-written”.
To support the internet as a global search infrastructure, the semantic web would have to become a standard. And everyone would have to comply with that standard.
Our standards-compliance track record to date sucks. Build a web page in standards-compliant XHTML and CSS. If you want it to look the same in Internet Explorer 6 and 7, plus Firefox, you’ll have to go through all sorts of ridiculous gyrations. Why? Because no one can agree on a standard. After 15 years.
Bring all web content ahead a generation? We can’t even get 25% of web sites to use the best the year 2000 had to offer.
It all sounds a bit like the XML craze that swept through soon-to-be-defunct startups around 1999.
In short, the semantic web is the greatest thing that will never happen.
Even If It Does Work, It Won’t Help
Oh, and what drives semantic structuring of data?
You assign and tag information, thereby assigning it to categories.
You still have to rank that information. Somewhere, someone will still have to decide what shows up first.
Which means someone, somewhere will try to find ways to be that #1 listing. And that will still come down to picking and optimizing for the right keywords.
We’ll come full circle: Right back to keywords.
The Answer: Improve
Better keyword search is the answer. “Teaching” search engines to understand that, in some contexts, “auto” is the same as “car” is the next big advance.
That’s why Google and the other major players are investing (OK, a little conjecture here, but I think it’s a safe bet) in latent semantic indexing.
If internet search is going to move into using highly structured data, it’ll do so by working with what’s out there now. Not through a bottom-up application of standards.
But, I’m sure someone will take this theme and run with it, all the way to an IPO.
By the way, semantic markup, microformats and similar concepts are pretty damned cool. They work brilliantly in environments where you can control the content. They just won’t work worth spit when you apply them to the internet’s huge, growing blob of content.
Ian Lurie is CEO and founder of Portent Inc. He's recorded training for Lynda.com, writes regularly for the Portent Blog and has been published on AllThingsD, Forbes.com and TechCrunch. Ian speaks at conferences around the world, including SearchLove, MozCon, SIC and ad:Tech. Follow him on Twitter at portentint. He also just published a book about strategy for services businesses: One Trick Ponies Get Shot, available on Kindle. Read More