Organic search: hazardous to your analytics?

Interesting informal gathering yesterday evening, on the topic of webAnalytics. At some point, someone says something like "X% of our visits come from organic search."

Organic search is what people in the field of Search Engine Optimization (SEO) use to describe what many users would call "links," returned by a keyword search.  It is contrasted with paid search, that would refer to the small text advertisements found sometimes at the top of the page, but usually displayed on the right hand side of the page returned by Google. As far as I can tell, organic search has been coined sometimes around 2001-2002 at a time when engines experimented with blended links (paid and organic), a practice that backfired and has been replaced with well segregated links and advertisements. (there were pragmatic concerns, i.e. consumers would start to distrust the whole notion of search; and legal ones, as blended paid search could be deceptive).

I point this out because for people like me, trained in marketing, referring to organic vs paid search is problematic. I tend to refer to organic search as "links" (I am not a gifted linguist) and paid search as contextual advertising. Consumers search for (relevant) links. They do not search for contextual advertising.

As I do not have access to first hand data, I can only rely on figures that are widely circulated on the Internet, which suggest that contextual advertisements generate a click thru rate (CTR) in the vicinity of 3%. A back of the envelope calculation suggests that consumers are more than three times more likely to click on a link compared to a contextual advertisement (assuming two ads and ten links per page). If this is the case, organizations should invest more resources in SEO than in the placement of contextual advertisements, as both types of leads appear to generate an equivalent number of conversions.

However, since webAnalytics is focused on tracking consumers once they enter the premises, I suspect that analysts will be tempted to focus on paid search, as it is easy to compute the ROI of additional advertisements, whereas improving the visibility of links in order to generate more organic search, is demanding.

The situation is in fact more complex as I would expect that consumers searching in order to make a purchase decision are much more likely to trust information provided by an independent third party, such that the best strategy is to try to make favorable independent opinions rise to the surface of links. Once a decision is made, consumers will either type a tentative url in their browser or will search for a specific entity, in which case a link or a contextual advertisement should have similar appeals. If this logic holds, the best course of action would then be to push favorable independent sites when category-related keywords are used, and contextual advertising when product-specific keywords are used. Maybe even if keywords suggest that the choice was in favor of a competitor.

But this is way beyond the point -- speaking about organic vs paid search is an unfortunate historical accident that blurs our understanding of the consumer.

(thanks to Jacques Warren -- see comments below -- who pointed out that "hits" have a specific meaning in the analytics community, in a nutshell, the number of files sent from servers to clients -- see here. Unfortunate because "hits" also has a well established usage in the context of search, in a nutshell, the number of occurences of keywords in a search space -- see here. I've replaced hits by links in my post, as a sign of respect for the analytics community ;)

Comments

[Jacques Warren]

I beg to differ.

True, knowing how to blend organic and paid results is an important success factor, and organic is still what drives the majority of traffic. I am not so sure however that it is because users distrust paid links. I believe people are after relevancy, and are happy to find it when they *shop*. As long as those sites are excatly what I am looking for, what do I care whether they paid to be there or not. I want to check 5, 6 sites and hopefully be able to make up my mind (NOTE: isn't it fantastic to see how web search disprooved the Rational Buyer model? I mean, who checks the thousands of results in order to make the best/rational purchase decision?).

I have seen, however, organic search performs much better than the paid one in terms of conversion (sales), but it was hard to really compare, since the organic and paid were not *competing* on the same keyphrases (using paid was rather a mean to rake larger than the organically well-positioned site, hence the less qualified/receptive traffic, I believe).

Finally, I like how you categorize both search types, but I would try to find a better word than "hits"; we web analysts tend to be allergic to it!

[Stephane Gauvin]

Thanks for your comment.

1) YES YES YES, I wish the superior being had given me the gift of words! :) I'll replace hits with "links". If you have a better idea, let me know.

2) Hmmm.... tremendous amount of literature on what is called "source credibility". If there is the perception of conflict of interest, impact goes down. If Microsoft says Zune is the best, you will discount the information. If ZDNet says so, you'll tend to give more credit (still, they sell advertising...). If some Warren guy says so, you tend to give more credibility. If someone you know first-hand and trust says Zune is cool, well, you either tend to give much credibility or revise your opinion about this person :)