Comments on: When do you need a Digital Asset Management System? http://blog.razuna.com/2009/01/02/when-do-you-need-a-digital-asset-management-system/ Razuna - Open Source Digital Asset Management (DAM) / Open Source Media Asset Management (MAM), Hosted SaaS DAM Fri, 03 Sep 2010 01:55:22 +0000 http://wordpress.org/?v=2.9.2 hourly 1 By: Roger Howardhttp://blog.razuna.com/2009/01/02/when-do-you-need-a-digital-asset-management-system/comment-page-1/#comment-526 Roger Howard Fri, 20 Feb 2009 20:13:56 +0000 http://blog.sixsigns.com/?p=679#comment-526 Yes and no. First, I can't pretend to know what Google does and doesn't index. However, generally they are all about indexing content and context, and not metadata - for instance, while they *may* use metadata (everything from keywords in web page headers to, potentially, IPTC/XMP) they are far, far more likely to give weight to the context (what page it's on, who links to it) and the content itself.PDF and Excel - they index the *content* of these files, which is largely text. You can find a PDF on Google not because of any keywords that may be in the XMP packet, but because of all the context encoded as text in the PDF. Same with XLS files.The reason this has been such a successful strategy for them in general - and is what set them apart early on - is that metadata is easy to fake... people try to game many search engines (Craigslist is notorious for this) with bad keywording. But it's much harder to actually fake content, or fake the context (for instance, if lot's of respected sites link to your content, that provides trustworthy context).I wouldn't be surprised if Google does start indexing - or already does - embedded metadata in media files at some point... but there's no evidence they do now, or that if they did it would be used as a trusted source of indexing. Metadata is just too easily gamed... I can easily place a bunch of irrelevant terms into my keywords, making it unreliable (and people make CAREERS off of trying to fool Google), but it's much harder to get a whole bunch of trusted sites to link to your content if it's not what you say it is. Yes and no. First, I can’t pretend to know what Google does and doesn’t index. However, generally they are all about indexing content and context, and not metadata – for instance, while they *may* use metadata (everything from keywords in web page headers to, potentially, IPTC/XMP) they are far, far more likely to give weight to the context (what page it’s on, who links to it) and the content itself.

PDF and Excel – they index the *content* of these files, which is largely text. You can find a PDF on Google not because of any keywords that may be in the XMP packet, but because of all the context encoded as text in the PDF. Same with XLS files.

The reason this has been such a successful strategy for them in general – and is what set them apart early on – is that metadata is easy to fake… people try to game many search engines (Craigslist is notorious for this) with bad keywording. But it’s much harder to actually fake content, or fake the context (for instance, if lot’s of respected sites link to your content, that provides trustworthy context).

I wouldn’t be surprised if Google does start indexing – or already does – embedded metadata in media files at some point… but there’s no evidence they do now, or that if they did it would be used as a trusted source of indexing. Metadata is just too easily gamed… I can easily place a bunch of irrelevant terms into my keywords, making it unreliable (and people make CAREERS off of trying to fool Google), but it’s much harder to get a whole bunch of trusted sites to link to your content if it’s not what you say it is.

]]> By: Crishttp://blog.razuna.com/2009/01/02/when-do-you-need-a-digital-asset-management-system/comment-page-1/#comment-525 Cris Fri, 20 Feb 2009 00:29:32 +0000 http://blog.sixsigns.com/?p=679#comment-525 But doesn't Google use the meta data in media files (pdfs, excel, etc.) to help them rank better? I have been using <a href="http://www.extensis.com/en/digital-asset-management/" rel="nofollow">digital asset management</a> to add this and it seems to help it get indexed/ranked better. But doesn’t Google use the meta data in media files (pdfs, excel, etc.) to help them rank better? I have been using digital asset management to add this and it seems to help it get indexed/ranked better.

]]> By: SixSignshttp://blog.razuna.com/2009/01/02/when-do-you-need-a-digital-asset-management-system/comment-page-1/#comment-505 SixSigns Wed, 07 Jan 2009 00:57:49 +0000 http://blog.sixsigns.com/?p=679#comment-505 Looks that you are correct. I could swear that I read on the Google blog that they make use of the embedded metadata, but now I can not find the related link anymore. Thank you for your comment. Looks that you are correct. I could swear that I read on the Google blog that they make use of the embedded metadata, but now I can not find the related link anymore. Thank you for your comment.

]]> By: Roger Howardhttp://blog.razuna.com/2009/01/02/when-do-you-need-a-digital-asset-management-system/comment-page-1/#comment-504 Roger Howard Wed, 07 Jan 2009 00:44:57 +0000 http://blog.sixsigns.com/?p=679#comment-504 Just one correction - to my knowledge, Google Image search still doesn't leverage any of the metadata embedded in media files.Cheers -R Just one correction – to my knowledge, Google Image search still doesn’t leverage any of the metadata embedded in media files.

Cheers -R

]]>