Google was a pioneer in the search world for ranking websites based on incoming links as well as content. But which other measures could the search giant use to evaluate relevance?
Apart from the top level 'links' and 'content', which we know affect placement, online marketers use experience and testing to work out the other things Google looks at when ascertaining the relevancy of web pages for search terms.
What other top level techniques could Google potentially use (if it isn't already) to determine web page relevancy?
1. Video Content – Listening for brand mentions in video.
With Google owning YouTube, it is in a prime position to build technology which can analyse words used within user generated videos.
The technology could recognise the use of brand and product names, and rank the websites of regularly mentioned brands higher.
On another level, the technology could be intelligent enough to recognise brand logos in videos. One might argue it's a step too far, and gives an unfair advantage to established brands, but then again, aren't established brands the most likely to be talked about anyway?
2. Gmail – Rating websites by their mentions in email.
If any of these techniques happen already it's got to be this one.
We all trust our family, friends, colleagues and associates (the people who email us), so if we assume they would only send us relevant and resourceful links, wouldn't this be an invaluable measure of a web page's quality?
3. Conversation tracking – Rating brands that generate positive online conversations online higher.
I once wrote on Debra Mastaler's blog about how brands that are suffering bad publicity end up acquiring SEO friendly links.
Companies such as Market Sentinel have software that understands the sentiment of online conversations.
Imagine if Google could understand the difference between positive and negative conversation and discount links which are born out of negativity accordingly.
Those are a few of my suggestions, and I welcome any others in the comments box below.
Read more from Leon Bailey Green on his website .
The views of the author don't necessarily represent those of the publisher.
Related research:
Online Reputation and Buzz Monitoring Buyer's Guide 2008
Search Engine Optimisation (SEO): A Beginner's Guide
Related articles:
Is the role of the SEO dead and should PRs own natural search?
Can anyone compete with Google?
Reader comments (4)
Developer at Forward Internet Group
10:10AM on 12th August 2008
1. This sounds like a few big technical hurdles would need to be jumped before anything this clever could produce reliable results, a crowd sourced/mechanical turk approach might be better.
2 & 3. Both of these approaches could land google in a lot of privacy related issues and would likely require users/companies to opt-in and so would get lower takeup rates.
3:06PM on 12th August 2008
Out of all of these I think youtube results in the SERPS is the one which will take off. Maybe number of views and ratings will be part of the algo...anyone concur?
Chief Digital Officer at Mindshare
6:07PM on 12th August 2008
I tend to think that a continuation of the things they're already doing to some extent or another are more likley than the suggestions here: tracking users' web journeys, click data, social bookmarking etc..
All of these could be used to increase personalisation, whereas the suggestions here (other than the email one) would be aimed at impacting relevancy on generic results (something I'm pretty sure Google is keen to move away from)
SEO at Cheapflights Media
9:39AM on 13th August 2008
Mentions in video/audio I guess is no different in principal to counting the number of text mentions of a brand or keyword, so not massively insightful in most cases.
Sentiment analysis on audio will be an interesting development, and no doubt something that will come into play at some point.
Tend to agree with Ciaran that there are easier pickings amongst Google's extended portfolio of tools, services and user data collection points.
Log in to post a comment