Amazon invents the affiliate

If you recall the history of e-commerce affiliation you will remember that it all began with Amazon. In 1996 Amazon created what we now know of as the affiliate scheme and did so on the back of their own technology development (well they had to, didn’t they, because the third party systems followed later). Amazon’s motivation was simple, let’s encourage our customers to link to the products they love and/or find useful by rewarding them if a visit they refer to us leads to a sale, because by doing so, we’ll get more visitors. This encouragement worked, and there was an explosion of affiliate activity, motivated by, we’d argue, love and usefulness first with money following. While I’m no affiliate expert, I suspect Amazon have now and probably have always had the largest affiliate base (estimated at 2 million in the summer of 2008).

For those of you who were paying attention to Google’s search results at the end of the nineties and in the early noughties you may remember watching Amazon begin to dominate huge tracts of unique product searches, and as our understanding of the importance of Reputation to Google’s algorithm grew, it became clear that their affiliate scheme was behind much of this performance, with thousands of unique links, often with their own unique content, supporting products. We have referred to this phenomenon as Amazon’s Happy Accident; it just so happened that Google’s growing sophistication was taking more and more account of the contextual relationship between linking pages and where the linking page had its own reputation, as well as context, the target URL got a terrific boost in relevance, and this played right into the hands of Amazon’s scheme and, crucially, its implementation method, which was to link to an almost identical URL as the product URL, but with an affiliate parameter on the end. Amazon’s implementation method has had a further advance in recent times where they conditionally 301 redirect Google (not users) to a permanent, keyword-rich, version of the product URL (see Chris Smith’s Natural Search Blog analysis of this. We don’t agree with Chris’ cries of ‘unfair’, we think this is very smart).

Google: Affiliate Links as Paid Links?

So how does Google view affiliate links in the context of their purpose? If we park for a moment the existence of industrial-strength spammy affiliate behaviour and look to what an affiliate is actually telling Google by linking to a retailer/product combination. On the whole we’d argue that this is going to be tantamount to a vote of confidence; an endorsement of both the retailer and the product, and if the linking site has any sort of credibility or authority or Reputation of its own, then Google should be allowing it the influence the Reputation of the retailer/product target. Really, it should. Naturally there is a big ‘if’ in that last sentence, and here we believe is the crux of the matter. Google’s algorithm puts considerable store in a site’s ability to influence the performance of others by its linking, and its assessment of that ‘ability’ is at the core of why its results are better. Google is capable of identifying whether or not an affiliate site should be influencing the performance of its target retailer by analysing its own Reputation. Without any Reputation the affiliate simply will not have any effect on the performance of the target.

When an affiliate produces a site worth visiting, it begins to accrue Reputation. The consequence of this is that the affiliate site then starts to get its own traffic from search and begins to acquire the ability to improve the performance of its retailer through its links (it is a separate issue whether the affiliate would actually welcome this outcome). Google’s view of this, we believe, is that it is as it should be. A site that gains its own Reputation should be able to pass it on, even if it earns a payment as a result of the visitor it passes on buying something.  So if this is Google’s view, why is it not saying so?

“An Affiliate Link is Not a Paid Link” unspoken by Google

We believe Google is dancing round this issue because it doesn’t want to invite trouble for itself. The uncertainty and ambiguity in the marketplace at the moment around Google’s determination to stamp out paid linking (for influencing its algorithm) and the ‘terms of service’ policy of nofollowing  commercially-motivated links (again, for influencing the algorithm) suits it. Why? Because actually we suspect that Google isn’t doing as good a job as it would like in the detection of paid link behaviour, particularly at the industrial-strenght spammer heights and the last thing it wants to encourage is more affiliate linking (for influencing the algorithm) by giving the green light to it as a way of improving natural search performance. Of course, an affiliate gains very little from improving the performance of its retailer target, but the unethical use of affiliate sites by retailers themselves is a worry Google would have.

We believe that Google says to itself that an affiliate link is not a paid link, because it has the power to improve results for Google’s users and that is a good thing. And in particularly competitive markets, such as hotels and finance, good affiliates can really help Google sort the chaff from the wheat.

The Search Johnston Advice on Affiliate Links

In the past, we have championed affiliation as a smart way of building contextual links into sites to improve their natural search performance. More recently we have interpreted Google’s intentions around paid links to include affiliate link behaviour, so have discouraged our clients to look at them as a strategic option. As a result of our view on Google’s current position, we are moving back to including affiliate schemes as a potential weapon in our clients’ SEO armoury - providing the links actually point at the site in question and not a third party affiliate engine, which is a complete waste of time. When and how to deploy such a weapon is a subject for another post entirely.

• Keyword density The number of occurrences of a term in its containing document², divided by the total number of terms in its containing document.

• Global weight. The number of documents within the whole document collection (in Google’s case, the entire web!) with one or more occurrences of the term.

The Global weight is usually a log value, i.e.

From the above table, we can see that a term that appears in every document will have a zero term weight. A term that appears in 1 document in a collection of a thousand is 50% ‘weightier’ than one that appears in 10 documents, and 300% times ‘weightier’ than one that appears in 100 documents in the collection.

The term weight (’importance’) is a function of the local weight divided by the global weight. A few conclusions can be drawn from this.

• A term repeated n times in a document is not necessarily n times more relevant or meaningful. Google knows this.

• Global weight is a measure of the specificity of a term over a document collection. Global weight, then, is not a relevancy measure, it is a rarity measure.

• High search volume terms, or terms with a good conversion ratio, are important regardless of whether these are rare in terms of Global weight.

• We can use vector calculations to measure the level of similarity between a search query and indexed documents.

The real question, however, is whether this is a good way of deriving a relevance measure.

• Ranking by similarity in such a way cannot incorporate semantics – the meaningful relationships between words.

• Ranking by similarity in such a way cannot incorporate information content (’entropy’).

• Ranking by similarity in such a way cannot incorporate authority.

More successful ways to calculate relevance

In reality, a search engine considers the following things about the content:

• Existence: the presence of a keyword or key phrase in the text

• Proximity: the relative distance between key words/phrases in the text

• Positioning: the location of the key word/phrase in the text

• Co-occurrence: the frequency with which terms occur with other terms

• Subject: the main topic and sub-topics of the text (subject), also

  • Synonyms: similar words e.g., book → tome

  • Hypernyms: more general words e.g., book → publication

  • Hyponyms: more specific words e.g., book → paperback

  • Meronyms: parts e.g., book → page

• Entropy: the amount of information carried in the text

• Authority: the trustworthiness of the source

We have taken a lot of time and trouble over the promise that we are expressing here for the first time: SEO, Redefined. We feel SEO needs redefining because there is almost never a occasion where we talk to a client for the first time and they aren’t full of stuff and nonsense fed to them by SEO amateurs. For those of you who know us well, this statement will not be a surprise, and for those of you we are yet to meet, you will find a lot more where that came from when we do. We welcome any feedback or comments about how we are doing and look forward to working with you (further). Onwards and upwards.

Hear me speak at SES London.

Advanced Track - Organic Listings Forum - 19th Feb 2008 - 9.30am

“Pose questions to our panel of experts about free “organic” listing issues, plus participate in this session that allows the audience to share tips, tools and techniques. There’s no set agenda, so this is an ideal session to discuss any major recent changes with organic listings.

Moderator:

• Mike Grehan, Co-chair, SES London and Founder and CEO, Searchvisible Ltd.

Speakers:

• David Naylor, SEO, Bronco
• Ralph Tegtmeier, aka Fantomaster
• Edward Cowell, Technical Director, Neutralize
• Steve Johnston, Google Consultant”

Sorry about the short notice!

1. wiltshire county council
2. wiltshire times
3. wiltshire college
4. jobs in wiltshire
5. schools in wiltshire
6. wiltshire police
7. wiltshire farm foods
8. wiltshire gazette & herald
9. bbc wiltshire
10. wiltshire news
11. wiltshire buses
12. wiltshire map
13. cottages in wiltshire
14. wiltshire libraries
15. holidays in wiltshire
16. wiltshire football league
17. this is wiltshire
18. wiltshire hotels
19. wiltshire estate agents
20. weddings in wiltshire

The event price is a low £45 and there are only 110 tickets available. All the profits go to the school’s major projects, in particular its new library. All attendees get a free copy of the new edition of my ebook (22,000 words, 56 pages), worth £19.95.

For further details and to buy tickets, please visit the Ebay UK page for the event (event now in the past).

While maintaining the overall structure, I have expanded the content extensively, in fact by almost 100%, going from 12,500 words to 22,000 words. There are around 10 new sections.

For further details and the purchase page: Google SEO eBook