Comments on: Online Political Market Research http://danzarrella.com/online-political-market-research.html DanZarrella.com, Social & Viral Marketing Scientist Sun, 07 Sep 2008 06:29:38 +0000 http://wordpress.org/?v=2.6.1 By: Market Research Guy http://danzarrella.com/online-political-market-research.html#comment-29136 Market Research Guy Thu, 30 Aug 2007 21:32:26 +0000 http://danzarrella.com/online-political-market-research.html#comment-29136 Hi, You might be interested in this market research study about how social networking sites may or may not have an effect on the 2008 election: <a href="http://www.gmi-mr.com/gmipoll/release.php?p=20070821" rel="nofollow">http://www.gmi-mr.com/gmipoll/release.php?p=20070821</a> Hi,
You might be interested in this market research study about how social networking sites may or may not have an effect on the 2008 election: http://www.gmi-mr.com/gmipoll/release.php?p=20070821

]]> By: Dave Morgan http://danzarrella.com/online-political-market-research.html#comment-724 Dave Morgan Tue, 28 Nov 2006 19:14:01 +0000 http://danzarrella.com/online-political-market-research.html#comment-724 Dan, pricing starts at $1,800/mo. for unlimited testing. (Unlimited means across any page(s), for any number of success metrics, for any number of segments, any number of concurrent tests, A/B and/or multivariate, etc.) http://www.sitespect.com DM Dan, pricing starts at $1,800/mo. for unlimited testing. (Unlimited means across any page(s), for any number of success metrics, for any number of segments, any number of concurrent tests, A/B and/or multivariate, etc.)

http://www.sitespect.com

DM

]]> By: Dan Zarrella http://danzarrella.com/online-political-market-research.html#comment-720 Dan Zarrella Tue, 28 Nov 2006 11:56:47 +0000 http://danzarrella.com/online-political-market-research.html#comment-720 me? nothing, not my style. But a potential client is. I see you're from a testing company, what kind of pricing do you guys offer? me? nothing, not my style. But a potential client is.
I see you’re from a testing company, what kind of pricing do you guys offer?

]]> By: Dave Morgan http://danzarrella.com/online-political-market-research.html#comment-714 Dave Morgan Mon, 27 Nov 2006 23:58:05 +0000 http://danzarrella.com/online-political-market-research.html#comment-714 > I can state that I will be using them in 2007. what office are you running for? :) > I can state that I will be using them in 2007.

what office are you running for? :)

]]> By: Dan Zarrella http://danzarrella.com/online-political-market-research.html#comment-712 Dan Zarrella Mon, 27 Nov 2006 15:12:34 +0000 http://danzarrella.com/online-political-market-research.html#comment-712 It seems like this is an area ripe for using behavioral targeting. pre-targetting and all that. I'm be surprised if these kinds of tools and tests haven't been used already in past elections, and I can state that I will be using them in 2007. It seems like this is an area ripe for using behavioral targeting. pre-targetting and all that. I’m be surprised if these kinds of tools and tests haven’t been used already in past elections, and I can state that I will be using them in 2007.

]]> By: Dave Morgan http://danzarrella.com/online-political-market-research.html#comment-704 Dave Morgan Mon, 27 Nov 2006 03:02:29 +0000 http://danzarrella.com/online-political-market-research.html#comment-704 Addressing the issue of multivariate web testing as a tool for honing political message(s): the challenge comes in knowing enough about the users to be able to segment them into targetable groups. Whereas a simple multivariate test may reveal which message is "best overall", the ideal approach is to further segment the data into discrete groups so that the "best per group" messages may be identified. In theory this is no different from pollsters who conduct phone surveys to identify hot-buttons for specific groups of voters based on geography, gender, special interest group, societal class, etc. On the web, we can run multivariate tests and do the same thing, but as of now we're somewhat limited in terms of reliably identifying the segments (e.g. how to identify the gender of a user?) But of course on the web, we can test 1,000,000's of people in days - far more than telephone polls. Looking forward to 2008, I won't be surprised if these sorts of tools are used ... will you? Addressing the issue of multivariate web testing as a tool for honing political message(s): the challenge comes in knowing enough about the users to be able to segment them into targetable groups.

Whereas a simple multivariate test may reveal which message is “best overall”, the ideal approach is to further segment the data into discrete groups so that the “best per group” messages may be identified.

In theory this is no different from pollsters who conduct phone surveys to identify hot-buttons for specific groups of voters based on geography, gender, special interest group, societal class, etc.

On the web, we can run multivariate tests and do the same thing, but as of now we’re somewhat limited in terms of reliably identifying the segments (e.g. how to identify the gender of a user?) But of course on the web, we can test 1,000,000’s of people in days - far more than telephone polls.

Looking forward to 2008, I won’t be surprised if these sorts of tools are used … will you?

]]>