1. Perhaps I am retarded, but is this an index? What does the "Internet Presence" measure mean?

If anyone knows the answer, that would be great.

Posted at 8:04PM on Dec 18th 2005 by Gina Bianchini

2. Hi -

Can you please publish a key of what these number represent -- and how they were actually derived?

Until then, I'm afraid these graphs don't equate.

"subjects are carefully defined to key in on the right subjects" ?
"Look in many places" ?
"Various statistics to smooth, combine and otherwise present" ?

Please explain so we can understand where you're coming from....

Thanks.

/m

Posted at 8:04PM on Dec 18th 2005 by Myles

3. The numbers are an index showing the CHANGE in presence - therefore the actual size presence is netted out. The observation that Friendster has more users than another site - or more content (presence) on the internet - is different than the growth rate in presence. A simple way to think about presence is the number of pages on the internet discussing a subject - in this case the subjects are the various social networking sites.

In answer to the questions on the exact methodology - that is proprietary - but a very simple way to think about presence and trends is do a search on Google every day on say "Friendster" and store the hits coming back - then plot the change in hits (pages on the subject) over time.

TrendIQ has found in working with its customers in various subject areas that the presence trend date provides unique insights into what is happening in "the real world." E.g. see the democratic candidate information: http://www.trendiq.com/info/Pres-graph-all.htm

Posted at 8:04PM on Dec 18th 2005 by Paul Feldman

4. Thanks for the explanation. I appreciate it.

Posted at 8:04PM on Dec 18th 2005 by Gina Bianchini

5. Paul, I 'm afraid I still don't understand your explanation. While in principle you are trying to define a Brand Development Index and thus perhaps a change in relative performance levels this looks pretty suspect. Guess it starts with starting with 100 whatever that number means.

While intuitively these numbers (with the exception of LinkedIn) seem to have the right sort of movement it is based on data that has nothing to do with actual usage. In fact in the period Orkut would have had enormous impact on any numbers (whatever they were) you crunched and Orkut is not even in the sample.

Presence is also a misnomer. I believe you mean it like visibility or mana in the search engines. However as these services also have different access and pricing models we are even further from understanding whether this "presence" you describe actually leads to a purchase intent.

Your distinctions above also fail to address differences between heavy and light users. So is this presense number you state suggesting that interest is increasing or decreasing, or is it merely a number that says... TOM top of mind awareness is X. TOM may not matter if they won't use it anyways.

I'm more than skeptical and I can't find any data desciption on your site that provides a more detailed explanation. There was a time where people rushed to get their hits up.... just to get funding. I hope to hell there is some "presence" rush for investors will be sorely disappointed.

Posted at 8:04PM on Dec 18th 2005 by Stuart Henshall

6. Questions for Paul - What does Presence mean, Paul ? I'm not clear how you define it. Does Presence only mean "the number of pages on the internet discussing a subject" ? What trends then are u talking about ? What implications ? What learnings beyond 'Bush bombing Iraq' or 'Vajpayee vs Musharraf' being popular subjects of discussion ? Does it take into account how users are behaving - eg. you might have 10 active members flooding the boards / entries while another 90 simply 'lurk' ?

Posted at 8:04PM on Dec 18th 2005 by Dina

7. I agree with Stuart that the term 'presence' is misleading. What is really intended is a measure of buzz -- how often the service is measured. It doesn't really equate to product uptake, or use, or over all numbers of users. Its just a measure of PR results. Which is perhaps a leading indicator of other things, either good or bad. Perhaps its best thought of as a measure of volatility.

Posted at 8:04PM on Dec 18th 2005 by Stowe Boyd

8. if they are doing more than measuring the number of times that something is mentioned on the net, than that would be cool, but his explanation and the ensuing discussion seems to be more hooey than actual meat.

Stuart - I think you are asking for artificial intelligence trend iq. i don't think anyone is there yet, except for maybe the folks at ibm working on websphere.

from the explanation above, it looks like trendiq is doing the same thing as waypath's buzzmeter. you can do that yourself and don't need to pay anyone to run the query or create the pretty graph. you can do it all through a simple form. . also, once you put the code on your site, it generates the graph for you over time and is always current to the date, instead of pasting an image file.

i created a graph at waypath one and I am gonna attempt to add the code here. not sure if it'll work.

Posted at 8:04PM on Dec 18th 2005 by Peter

9. yeah. that didn't work. you should just allow html in your comments. we are all bloggers, we know how to write html.

Posted at 8:04PM on Dec 18th 2005 by Peter

10. On the comments about the index and 100... - its like the Consumer Price index which also starts at 100 by convention - the index is a measurement of percent change in X - like the CPI is the percent change in Price.
On the meaning of presence - it does measure buzz - or pages of content containing the target subjects if you like. We have found and have data sets that show very high correlation between what is happening in the "real world" versus teh internet presence indexes - that shold not be a surprise - but its nice to have the data. Our trend data is more interesting when there is not a separate "real world" description of a subject - like there is for example in presidential elactions.
Predictions - no it doesn't predict except that trends obviously point in directions - but it doesn't predict any more than any other trend found by any other method.
Presence (buzz) vs members, etc. - this data only measures changes - not levels - so if one company had 100 users and went to 110 (and their buzz followed linearly) their trend presence is up 10% - if another company had 10 users and went to 20 (and the buzz followed linearly) the trend is up 100% - its up to the trend interpreter to say which is more interesting and where to do more research.
Can you do it yourself - yes anyone can type a phrase into Google every day and look at the hits coming back and make an index. We do far more than that in many ways but the result similar in intent - we just get a more accurate measure. The cost of us doing it is also less than than the cost of anyone doinng it themselves (assuming you value your time).
we do offer "sentiment indexes" that measure how people feel about subjects - and that far more complicated and involves cross sectional regression models, etc.
Glad to answer more questions ...

Posted at 8:04PM on Dec 18th 2005 by Paul Feldman

11. "Can you do it yourself - yes anyone can type a phrase into Google every day and look at the hits coming back and make an index. We do far more than that in many ways but the result similar in intent - we just get a more accurate measure. The cost of us doing it is also less than than the cost of anyone doinng it themselves (assuming you value your time)."

I am not convinced of this statement. You mentioned regression models for advanced trend analysis and prediction and that is definitely a value added service. However, I fail to see anything more complex in your analysis for the graph created above on YASNS than counting and reporting how many times a term was mentioned on the internet. (possibly using google's index?)

I am not trying to say that your service is not a good one. I am just saying that there may be a free self-service alternative. I am certainly not going to type terms into google or overture's inventory tool everyday or month and record that for all these terms. You are right, that would take a lot of time. However, waypath.com counts the number of mention in weblogs of terms and allows anyone to chart and compare it. Have you used waypath's buzzmeter? How does it compare to what you are doing?

I understand your solution is proprietary. However, If there is more than the "# of mentions" that goes into your "calculation" of internet presence, I'd think you could atleast mention those. This way, I will be able to determine the value of the graph above vs waypath's free self-service tool.

http://www.waypath.com/buzzmaker

I've create a waypath graph and posted it on my blog for the top 5 services according to your graph. (since I can't post html here)
http://worcester.typepad.com/pc4media/2004/04/waypath_buzzmet.html

While, we are talking about measuring buzz, also of use is
1. www.alexa.com which measures browsing habits of people with their toolbar (stats also available and collected via amazon's new search engine, a9)
2. http://inventory.overture.com
3. Using google's adwords program to see predicted costs of keywords based on amount of searches a term gets.
4. of course, the google zeitgeist, lycos top 50 and all the other top # reporting services are worthwhile for general "what's popular now?" .
5. eurekster (if more people used it, it'd be worthwhile), would allow us to see what specific groups of people think are buzzworthy.

I think that the whole "proprietary for a fee vs free" trend analysis and internet data mining should be something we are all concerned about. So far, information on the web is free and searchable. Even though we enter info into proprietary YASN systems, we and anyone we want to allow access to the info, can get to it. However, with projects like ibm's webfountain and trend iq's analysis, we are putting a price on mining and analyzing this data, which puts the info and knowledge out of reach of most people and only into the people's hands that can afford to buy it. This raises concerns about information equality. I am not sure how I feel about it. Everybody's got to get paid for what they do. And access to information and knowledge certainly isn't equally provided to everyone on earth now. But, on the other hand, this is a step away from the "information is free" ethic that the internet has championed.

Posted at 8:04PM on Dec 18th 2005 by Peter