It seems every day Google, and now Yahoo as well, adds more features to its search – first it was phone numbers , then tracking packages, then patents, now it’s whois, flights, UPC codes, VINs, and God knows what else. Read a few pages of Google Hacks, and you’ll realize, you never use even 2% of Google’s power, and, most likely, you never will.
This leads me to wonder, where is this all going? I mean, the fact is, most searchers simply don’t use advanced search features *at all* – not even simple operators like quotes (” black jaguar” cat) or negative inclusion (jaguar -cat). So why are these search sites loading up on features that, honestly, nearly all their users will never take advantage of? Do they think searcher’s habits are going to change? I doubt it. I’d be interested in why you these features are being added with such abandon. Just because they can? Maybe they think folks will be building applications on top of the search platform, or will they do it themselves? Are they expecting that a layer of expert searchers will develop who peddle intermediary services (ie Google Answers)? I mean, I can get as excited as the next guy about the addition of the tilde operator or the “*” function, but….it feels like there is something in aggregate I am missing. Must be the varathane on the floor in the next room, keeping me from grokking the grand plan in all this. Help me out!
Maybe the text box search interface is not the way these get used. Google as IVR-over-VOIP service in a few years could allow users to exploit more of these specialized queries.
The common thread behind these features (apart from the calculator) is that they do not require any cognitive effort from users, unlike boolean operators, . Google will volunteer smart searches when some search keywords match common patterns, the way Ask Jeeves does. After a while, users will learn and be progressively conditioned to use Google to search for books, phone numbers and the like, the whole user experience being almost subliminal.
Is it not obvious? Make yourself infinitely simple, and infinitely useful. One stop shopping for anything you might want to look up. It’s insidious and genius.
This works nifty until you have to “disambiguate” collisions between conflicting code lists. A postal code in one country might be a stock ticker code for another. How do you know which to return?
The popular way to solve this is to set an ID or cookie based preference. Downside is if add too many of those that’ll slow your transactions down.
Then there is a third way. Try and infer what the user is looking for from the context of their search history and any other clues that can be gathered to enhance the intelligence of the assumptions made by the system. Much harder but slick and magical if you can get it right
First: if it doesn’t cost them much, and doesn’t complicate matters for the average end user… why not?
Second: if all search engines had only the same basic functionality… what’s stopping me from switching to another engine? These extras, once I discover them and use them are likely to keep me hooked on their engine.
Third: geek word-of-mouth marketing? I started using Google shortly after it was availible and recommended it to everyone I knew. By appealing to geeks they got a ton of word of mouth advertising. This is just another way to recruit more geeks.
It’s the same thing that plagued Apple, then Netscape. The geeks run the joint, don’t particularly care about users, and like to show each other how brilliant they are. They’re gods, you’re not supposed to ask why, you’re just supposed to admire them and wish you were as smart as they are.
Geek cred fuels the google brand, so I agree with both Mark and Dave.
I’m also thinking that as they add features, the API gets more powerful, which leads developers to add google search to their programming toolkits. That’s not much diff than MSoft’s old strategy – control the OS, you control developers.
And as the API spreads into other programs, it feeds that general ‘beyond the browser’ strategy you mentioned in the previous post. This also is perhaps a defensive move, designed to prevent the possiblity of the company being “Netscaped” – market penetration within the browser having been proven in many cases to be difficult to defense.
Perhaps it
I think Dave’s wrong, in this case – the “geeks running the joint” aren’t acting out of hubris. Tim and Joe and Fazal, I think, all have pieces of it:
– keep the interface dead simple, and try to guess well for cooler features (like Fedex and phone numbers), and you service the unwashed masses who will simply never explicitly use more than a tiny fraction of the tool – they get what they expect, and sometimes more;
– Provide advanced, cool features to those who have an incentive to use the tool well, and you’ve hooked them – switching costs go up for serious users;
– Build a rich feature set and let other developers come up with ideas. Roll whatever you can back in to the core product, and be make money on enabling anything you can’t External developers do R&D for free, and dig your defenses for you.
It seems like a smart plan for attempting to please three exremely important segments of a market like Google’s.
Look at each of these features in isolation. It offers a significant advantages to a small(ish) segment of the population. Imagine, for example, Ebayers using UPS tracking; they don’t use it enough to want to learn UPS’s own interface, so they’ll just stick with Google. This puts more eyeballs over Google’s ads, and provides Google with more information to help them target those ads better. You just shipped a Dresden figurine? Here’s a broker that specializes in antiques insurance.
I think ian is on to something. People specialize and have specialized interessts. people who really need to keep tracck of UPS shipments are going to learn about this right quick, and most of them go to google for search anyhow. As microsoft integraates searcch more and more into their operating system, and of as you believe, the true value added of the net is searchh, and finally as more and more interaction with the web becomes browser free, thee brand name for finding whatever you are looking for will cease being http://www.google.com and become just Google.
Who knows whaat machine people will be looking ffor things from in ten years but the tool google wants them to be using to find anything will be based on google code.
Vegeetarian restaurants; loan approvals; credit reeports; criminal reccords; judicial deccisions; planning departmeent applications; as google figures out how to integrate that material into their interface the mor elikely i am to come back to them.
thee oracle of delphi could answer any question. google just wants to more accurate