A visit to Google London Options

[Scratch]

I was fortunate enough to get a couple of hours presenting to the head of European marketing for Google yesterday. I'd emailed the VP of product development to ask for the opportunity to discuss some ideas I've had in the networking and search engine space.

He put me on to this contact in the London office, who told me, "You can imagine how many emails like this we get. 99% of people will be told to go away." Luckily, she thought I sounded nice and agreed to set up a chat :-)

I presented 3 concepts for Internet applications, with the intention of exploring opportunities for working with the company. I'll say a little bit about what they are, and what I learned from the meeting.

1) People's postcodes.

This was a loosener to illustrate the way I think and work. PP is based on the idea that people's addresses and postcodes are already in the public domain, but Royal Mail (a private company now) charge anything from £1500 per year to subscribe to it. Also, the database updates are supplied on CD media quarterly! Not great service for something that's kind of ours anyway.

My idea, which I half-prototyped, was to use spiders to run SE searches on parts of postcodes, scrape the results pages, and try to find addresses in the source code. They'd get put into a database of candidate results, compared and promoted, and then made available back to the community via useful means, like web services.

2) ID+

You can read about this at www.idplus.org. They were really keen on this idea, and are willing to forward some stuff from me to the team working on the Orkut product. They've asked for a few points summarizing the level of synergy between Orkut and ID+.

From what I can tell, Orkut is a fairly immature social networking product. Does anyone have experience of using it more extensively?

3) Google Mart

This is the biggie, a new concept that I've developed in the last 3 months. It's based on the idea that search engines are quite ineffective at matching supply and demand of products & services. They're fine for raw information: words and phrases. But when it comes to recommending the best place to go for a certain product or service for me in my particular context, they're rubbish.

I've argued before that, to do this effectively, SEs need to ask for, and compare, more specific data:
[list]

[*]Price (max/min) (Froogle is starting to get hold of this)

[*]Distance (Google Local is doing a good job of this now)

[*]Trust / quality.. worked out from people's ratings.

[*]Context (being able to predict what I'm most likely to prefer, based on my preferences and prejudices, my socio-economic profile etc.)
[list]

Basically, the straight list the Google gives you today when you search for "babysitter Birmingham" is very likely not to contain the right answer. And if it does contain the best service for you, that could be waaaaaay down the list and never get seen.

That's because the algorithms that the big G uses today just aren't broad enough to deal with the task. However, they don't need to be much more complex.

My Google contacts listened with interest, and told me that basically the company's strategy really excludes saving people's data, and any appearance of push-marketing. I guess they think people won't trust them if they ask for personal info. For me, if you could give me better matches for stuff, hell I'll trade data with you! Sure!

I learned that Google has an internal "ideas pool", that all employees can contribute to. What they're really short on is development resource (nearly all is done at Mountain View, with some in Geneva, Switzerland). Obtaining development resource for a project was described as a "bloodbath".

So, any opportunity to work with the Great G may be weeks or months away, and I'm not holding my breath. But I've come away more passionate about the rightness of this supply-demand project than ever before...

I'm off to Barcelona for a stag weekend now...

Peace

[travis]

Mate,

Your an ideas man, no question.

Thats good positive stuff. Its always the smaller business that comes up with the new ideas.

Good luck with it.

[randfish]

Scratch -

Brilliant stuff, simply genius. I hope they take you up on your presentations; I think you've got some real potential. Enjoy Barcelona!

[cre8pc]

Ben,

Most impressive. I hope you remember us "little people" when you become rich and famous :wink:

You wrote:

My Google contacts listened with interest, and told me that basically the company's strategy really excludes saving people's data, and any appearance of push-marketing. I guess they think people won't trust them if they ask for personal info.

That caught my attention because of this article from June 3.

Google's long memory stirs privacy concerns

Google officials say their extensive log files help them improve service, fight fraud and develop new products, and unlike many other online companies, it seems willing to pay for the enormous storage capacity needed to save the data.

\"If it's useful, we'll hold on to it,\" said Nicole Wong, a Google associate general counsel.

My impression from the article was that enough personal data is stored, from emails to registration info to tracking searches. Seems to me this is enough information for developing refined, personalized searches - which from what I gathered from the NYC Search Engine Strategies conference 2005 is exactly what all the top search engines are striving for.

[send2paul]

Of all the webforums I've recently started to frequent to learn more about Google - this is the first time that I've read anything about them being, well, human - and open to ideas.

And those ideas are good! Nice one Scratch

[Scratch]

Thanks guys! I´m in easyInternet in Barcelona at 1.25am, fairly drunk, but I´ll try to give some sensible response to your comments.

@Paul.. Well, they were pretty human. It was interesting to hear Google talking openly about where they see themselves relative to Yahoo, Amazon, and eBay. They did not come across as arrogant in the least, much more like a real business competing in an open space with other worthwhile and capable enterprises. They have their strengths & weakensses & limitations like anyone else. They don´t have the means to put a fraction of their ideas into practice, for example... One guy spoke highly about Amazon´s customer profiling data / methods. He said "Amazon have the best profiling.." and that they didn´t make the most of it. I think that is with full strategic intent. Amazon is a strategically-led company IMO.

At the same time, they were pretty tight-lipped when it comes to talking about their direction or policy. Well, perhaps "not forthcoming" would be more accurate..

@Kim.. I ain´t rich and famous yet! And there´s no sign of suddenly springing up there, because it takes a lot more than great ideas to make it. I´ve learnt this recently in my interactions with the venture capital sphere. Good ideas are ten-a-penny (as we say in England).. what matters is the leverage or ability to put stuff into practice.

This is the reason why, when Google asked me if I wanted to send a mutual NDA down for them to sign in advance of the meeting, I said "No". I´ve learnt recently that nobody is interested in one golden egg: it´s the goose that counts. VCs don´t care a fig for an idea unless there´s a capable team behind it that they think can take the idea and make a business out of it. Larry and Sergei would back this up: they had a real, working prototype when they went for funding, and they got the funding.

Google have the same experience internally, it´s clear. They have lots of ideas, but dev resource is scarce. Again, ideas aren´t worth the paper they´re printed on.

So, no-one signed anything prior to this meeting. I knew that if they really wanted any of my ideas, they´d go for the goose, rather than the golden egg. Plus, my ideas are such that, if they exist, the world will be a better place, and if Google see fit to steal any of them, so be it, and good luck to them!

They actually recommended that I speak to eBay about the supply-demand concept, because eBay have the most sophisticated trust network already. I didn´t go for that idea, as I "don´t love eBay".. Google is a lovemark to me, warts and all. I realise now that they´re just a normal company, doing the best they can with the resources they have. They don´t feel invincible or all-powerful, that´s clear. They´re just normal guys.. I like that.

[Scratch]

As for how Google use our personal info, this doesn´t sit right.

To me, Google´s biggest challenge is not to become the next Microsoft: domainant but mis-trusted.. I totally appreciate their desire not to "be seen to" collect and hold personally-identifiable information about their users, but you know what, in my view they have to do this if they´re going to reach the next level of search performance.

The argument is very simple: I want a single place to which I can turn to find everything I want. I don´t want to remember a bunch of car sites, job sites, information search engines etc. I want one search engine that can make sense of whatever I throw at it.

So, if I´m searching for a house, I want to be able to say how much I´m willing to spend and the minimum number of bedrooms. If I´m searching for a car, I want to be able to specify make and model, and mileage or features. That´s not a lot to ask.

Let´s discount "natural language" search at this point: I don´t think the likes of Ask.com are capable of interpreting this kind of data in a useful way. Sorry. It´s not that complicated, doesn´t need to be. This is about a small number of discrete, relevant parameters.

If I do a search for "babysitter" or "plumber", why the hell shouldn´t Google (or the one that comes after Google) know:
a) Where I live, and who´s within range of me for the service I´m searching on..
Which service providers are most trusted..
c) Which babysitters are most trusted by people like me!

This isn´t magic, and isn´t rocket science.. I´m running out of time now on the ticker, and have to go and drink more cerveza, so I´ll leave this thought here...

More to come, I´m sure.. Maybe we´ll see Google Mart or something similar in due course...

Buenas Noches!

[Scratch]

One more thing, to finish off... I´m convinced that Google should trade us for a bit more personal info. I´d be willing to trade, wouldn´t you??

If they don´t, they´re capped. And someone will, take my word. Google got dominant in a matter of months. It can happen again.. Watch this space.

[travis]

Mate,

Good to see where your passions lie. 1:30am in the morning, smashed in Barcelona and you found an internet cafe to say hello.

Thats dedicated forum membership.