WebProWorld - View Single Post

**crankydave** · #12 (**permalink**) 12-27-2005, 07:22 PM

Quote:

Originally Posted by pemburung

Maybe I wasn't clear. Rumblepup's original question was " How much am I worth to Google?" Not, "how much is the average visit to Google's website worth to Google?" The former refers to a person, the latter to a visitor to a website, from a computer address. To get to the 90 million or so each month each address is counted only once regardless of how many actual visit there are from it, and the clock is started again each month, I assume. As I said, if it is interpreted as not starting again, then in 2005 at these figures approximately 1/6 of the world's entire population would click on Google; the year before, about the same; the year before, a little less. I don't believe that over the last three years half of the entire world's population has clicked on Google.

Rumblepup mentioned several time frames; I think what he is trying to get at (correct me if I am wrong, R), is the concept referred to as "the life of the client." DBurdon also took the meaning this way. This treats a customer as a unique entity, and the business looks at the profit from the entity, rather than the profit from a sale. Most catalog companies operate this way, depending on repeat business for their profitability; this is why they work so hard to retain customers. In this case we have to look at Google "entity customers", not Google website visits. And, as we don't know how many of a given month's visitors also return the following month, in this particular exercise we need to make a best guess. Mine would be that most people who use google do so regularly (many on this forum now apparently excepted!), and so the monthly unique use would not be a particularly small proportion of the total unique use. So, I used a range, going from "one month's unique visitors closely reflects the total unique visitors", to Dave's original "every month's unique visitors are truly unique, and so the monthly figure is 1/12 of the annual figure". It's not a wild guess, just not very accurate - factor of nearly 10. But then, withoug true unique visitors, we can't be accurate, and the dreaded average doesn't mean much either. Don't forget, if 12 women are nursing, and each is married, those 24 parents have an average of one lactating breast each. I think I'd go broke creating a company based on a nursing uni-bra.

A few problems with your reasononing...

1. You have no idea how Comsat and Nielson//NetRankings come up with unique visitors count. You have no idea what they factor in and factor out, or allow for in their count, if anything. You are only guessing.

2. The number of people using multiple locations (increasing unique visitors count) could be exactly offset by the number of different people clicking from the same location (decreasing unique visitors). Factoring anything plus or minus is a wild guess.

3. Your correllations to the averages and with the averages are flawed because you overlook that they are averages and treat them as exact counts as well as equating "unique" with "new". Two different concepts. As dburdon pointed out, people don't sign up so you can't know exactly how many different people. You extrapolate the averages from the knowns not the unknowns.

4. My use of 90M unique visitors was used as the average and I did not reason there were 90M new visitors every month for an entire year. This is part of your misuse of the averages. You can't simply add the numbers when the count is restarted every month. If you'll note, my math broke down profit to a 1 month average then divided it by one months unique visitors not the other way around.

I don't know how or if Google looks at a value per user. If they happen to be, my bet is on them looking at the value per known and not the unknown. P&L's don't make assumptions.

Dave