Google chief economist, Hal Varian, has done some ‘back of the envelope’ calculations to estimate their value to US users and advertisers.

The value to advertisers was estimated from the value of click traffic Google delivered to advertisers minus the cost of that traffic. I’ve blogged before at Search Engine Journal and in this post about what I think are some deficiencies with their logic regarding their value to advertisers.

But now they’ve added their value to users to their analysis, particularly, the value of the time that using Google saves people in finding answers to their questions. A study they cited claimed students who searched in the library ended up averaging 22 minutes to find what they needed whereas students using Google took an average of 7 minutes, saving 15 minutes.’

Google estimated that their search engine saves each person 3.75 minutes per day, 365 days in a year, which multiplied by the average US hourly earnings (\$22) times 130 million, the number of people employed in the US, yields a \$65 billion annual value in savings for users.

There are a couple of big things I find wrong with this analysis. One is that Google makes a common mistake that lots of people do: assuming that one’s time is worth their hourly wage. The marginal value of your time is not \$22 – it’s whatever value you could have created in that hour. If your dayjob pays \$22/hour and in an hour after work you could have written a freelance blog article and gotten paid \$5 for it, but instead watched the hockey game, you’ve missed an opportunity cost of \$5, not \$22. You essentially gave up \$5 to see the game.

If working your dayjob makes you \$22/hour and your next best alternative is freelance writing for \$5/hour, then the value of taking a half-day unpaid off work to do some freelance writing is 4*\$5 (the value you create writing) minus 4*\$22 (the value you would have created at your regular job), for a total of negative \$68, which is why no one takes a half-day unpaid off their regular job to do sub-minimum-wage drudgery.

Here’s another problem: Google assumes that the next-best-alternative to doing a Google search is a trip to the library. It’s not. The next best alternative is a search on Bing, or, before that, AltaVista or Lycos. If using Google is your best option because its results allow you to find an answer in 7 minutes, and Bing is your next best option because it requires visiting a few more links to find what you want, so it takes 10 minutes and going to the library takes 22 minutes and not solving the problem at all means not being able to include an important factoid in a report that you could otherwise sell for \$200 but now could only ask \$195, then:

• The library is worth \$5 for 22 minutes (\$0.23 per minute)
• Bing is worth \$5 for 10 minutes (\$0.50 per minute)
• and Google is worth \$5 for 7 minutes (\$0.71 per minute)

If you are paid \$22 per hour, then you must generate at least \$22 in value per hour for the person employing you, or else it is not worth it for them to give you \$22 for an hour of your time. So, you must generate at least \$0.37 per minute.

Regardless of whether or not search engines exist, it is simply not worth the 22 minute trip to the library to add only \$5 in value to the report. In a world with search engines, both Google and Bing are worth using, and Google more-so because it gives you what you need faster (through better search results and features like Instant). This is why Google is making such a push to develop more real-time search capabilities and voice-driven search – They generate more value per heartbeat.

Stepping back, this presentation and Google’s previous presentation that I blogged about seem to be part of a very interesting trend: Google trying to quantify the value of their existence. Why is Google’s chief economist, Hal Varian, playing the role of ‘Baghdad Bob’ to justify Google’s existence? For a company that was the fresh-faced darling of Wall Street and Silicon Valley just a few years ago, it’s surprising that they seem to think they need this sort of PR effort now.

I agree on the second point, but as far as the marginal value of people’s time, back when I was studying in economics in college (yes, I know, that probably means it’s all wrong), they talked about how people’s time at the margin is worth more than their wages, not less. That’s why people take vacations, and go home and go to sleep and so on.

For instance, if I was working 120 hours a week, and had bloodshot eyes and could barely see straight, you’d have to pay me a LOT for 1 more hour.

I think rather than assuming people have to find some other lower-paying job to do than their main job to work incremental hours, it’s easier for analysis purposes to assume one job where the hours can vary, and that people work as long as it makes sense for them to work economically (i.e. pretend everyone is working hourly, with potentially unlimited hours).

– Ted

• Ted,

What we’ve said agrees with each other. If you were sitting around on the couch all day watching ‘Tears of Our Lives’, you’d gladly trade an hour of your time for \$22 in cash. Twenty-two bucks is worth more to you than that hour. But, over time the marginal value you place on selling additional hours rises. So, by the time you hit 40 or 50 or 60 hours sold, it would take more than \$22 for me to get you to stay in that chair.

Likewise, I know you’re not as productive in that 61st hour as you were in the 60th. The marginal value of you Working drops over time. The marginal value of you Not-Working rises over time.