TechTalk at Google tomorrow

by Eszter Hargittai on January 29, 2007

I will be presenting in the TechTalk series at Google tomorrow.

    Google TechTalks are designed to disseminate a wide spectrum of views on topics including Current Affairs, Science, Medicine, Engineering, Business, Humanities, Law, Entertainment, and the Arts.

Interesting list, where do I fit in?

The title of my talk is “Beyond Gigs of Log Data: The Social Aspects of Internet Use”. I will be talking about the importance of social science research in gaining a better understanding of how and why people use digital media. That is, while companies like Google may have unbelievable amounts of information about their users based on logs of their online actions, I argue that there are other factors difficult to capture in logs that are also important to understanding how and why people use various online services the way that they do.

From Google’s perspective, I think one puzzle concerns the following. Despite being a media darling and getting a ton of positive press coverage over the years, other than search and ads, the company hasn’t gained significant market share in any realm. Even in search, how is it that they are only used by about half of all searchers with the kind of attention they get? (I actually have answers to this, my point here is that some people don’t seem to take a sufficiently nuanced approach to how the company’s products are doing.) And of course, search and ads are very important areas, but if Google thought that was enough, the company wouldn’t be expanding to other realms. It is expanding, but not very successfully.

Google Maps* and GMail** may be great products – I’ll be the first to admit it -, but again, the company’s market share is small compared to some of its big competitors. Sure, these are relatively recent entrants, but is there any evidence of significant diffusion to new users? Of course, if we really want to hit the bottom of the barrel, we can look at Google Checkout or the now defunct Google Answers.

My point is that simply having automated data about your own users’ actions isn’t going to tell you that much about why others are not your users, and why users of some of your services aren’t embracing others of your products. Doing so is like estimating public opinion about a Republican political candidate by going to a Young Republicans meeting or estimating public opinion about global warming by observing an environmental meeting.

Hopefully Google understands all this and works with people in this realm. I know for sure that they do some interesting work in user experience. But a bit more attention in this area than is apparent could be valuable.

[*] Based on some data I collected last year about a diverse group of college students’ Internet uses (N=1,336) here are some figures: Mapquest: 85% use it sometimes or often; Google Maps: 39% use it sometimes or often (an additional 33% have tried it, but don’t use it); Yahoo! Maps: 34% use it sometimes or often. This population is much more wired (more time online to explore things, easy access) than the general user population so figures here are likely to be higher than what one would find with a more representative sample.

[**] Based on some data I collected last year about a diverse group of college students’ Internet uses (N=1,336) here are some figures: Yahoo! Mail: 54%; Hotmail: 31%; AOL Mail: 19%; GMail: 12%. This population is much more wired (more time online to explore things, easy access) than the general user population so figures here are likely to be higher than what one would find with a more representative sample.

{ 9 comments }

1

Matt Austern 01.30.07 at 12:02 am

I’ll be there!

2

Daniel 01.30.07 at 9:00 am

My point is that simply having automated data about your own users’ actions isn’t going to tell you that much about why others are not your users, and why users of some of your services aren’t embracing others of your products. Doing so is like estimating public opinion about a Republican political candidate by going to a Young Republicans meeting or estimating public opinion about global warming by observing an environmental meeting.

This is an incredibly important point, and it took the marketing industry nearly ten years to make it. Back in the days of CPM and click-tracking, there were loads and loads of people who really didn’t understand advertising but who had natty monitoring and data-mining programs which they proclaimed were going to allow them to target their advertising scientifically and make all of those girly “marketeers” obsolete. I am frightened to think how influential Dilbert cartoons were on the development of the e-commerce industry in the 1990s.

Ten years later, people are gradually learning that the old fashioned problems of branding, consumer identification and market research haven’t gone away, and we’re all creeping back to something that looks very like marketing textbooks circa 1965.

3

John Quiggin 01.30.07 at 11:10 am

The fact that a generic portal like Yahoo can do so well indicates that inertia must play a big role.

4

eszter 01.30.07 at 12:23 pm

John, I suspect it’s partly inertia and partly that they offer a bunch of services that are relevant to users. Then there are the default portal pages that come automatic with certain ISPs, ones that users never change. My observations suggest that those continue to be used quite a bit in some groups.

5

Michael E. Sullivan 01.30.07 at 1:03 pm

It’s not clear that domination of a market is required for a product to be successful. I’m pretty impressed that 39% of internet users use Google Maps sometimes or often given that Mapquest was around as *the* brand for a few years before google maps was launched. That’s a *huge* client base. In what other industry would going from 0% market recognition to 39% in just a couple years be considered a failure?

People want everything Google does to be as successful as search, when all that really matters is that it provides a significant user base with a solution and makes them money.

So are they making money on these things? Or not? That’s what matters to shareholders, and the quality of the offering and the ability to maintain it is what matters to users. I greatly prefer google maps to mapquest for many reasons. I’m intrigued that 33% tried it once and now don’tt use it, and would love to hear why not.

Michael

6

Jake 01.30.07 at 1:57 pm

Google won’t tell you if they’re making money on Google Maps, although to some extent it’s hard to tell, because there’s definitely a brand value to “Google is the place that has all the cool and useful shit on the internet” that will/may keep people coming back to their more profitable properties.

I use Yahoo Yellow Pages rather than Google Maps/Local because it works how I want. I type in yp.yahoo.com, then type “bearings” into the search box, and it remembers where I live and shows me the bearing houses nearby. Google Maps doesn’t, and I can’t be bothered to try to figure out how to make it do what I want.

If you look at the areas in which Google got stomped (Orkut v. Friendster, Hotmail/Yahoo! Mail v. GMail, YouTube v. Google Video) vs. the areas in which they do the stomping (Search, AdWords, AdSense), I think that some fairly clear patterns develop.

Solving complex technical problems: excellent! Understanding what engineers/technical people/businesses want: pretty good! Understanding what the average person wants, when it does not overlap with the previous category: not so good!

7

Colin 01.30.07 at 5:07 pm

Yes to 6. As an occasionally-frustrated user of gmail, the thing should be much simpler — you get the sense of engineers figuring out solutions that appeal to them and deciding you should like them too.

Google Earth is great and fits what I think of google as doing well, which is bringing out structure that’s already in information and showing it to me.

8

ArC 01.31.07 at 7:24 am

If you look at the areas in which Google got stomped (Orkut v. Friendster […]), I think that some fairly clear patterns develop.

Solving complex technical problems: excellent! Understanding what engineers/technical people/businesses want: pretty good!

I’m just an Orkut end user, but I would never say they did even the technical side of it well.

9

ArC 01.31.07 at 7:25 am

Hey, the preview box for the comments led me astray — it formatted the two paragraph blockquote correctly, whereas the real comment ended up putting the second half of the quote outside the blockquote tags.

Comments on this entry are closed.