• by Rob Garner , August 19, 2009

Last week Google informally gave a heads-up that we should all be expecting a change in its main Web search results, based on a new update to search technology that mostly affects its indexing process.  Dubbed Google Caffeine, it is a "secret project" considered to be next-generation architecture for Google Web search. And in addition to shaking up the results a bit, it may also pave new roads toward the goal of real-time search results.  Early reviews in  preview SERPs indicate that there will be noticeable differences in the results pages for many terms and brands, so those paying close attention to their SEO campaigns are advised to watch this update closely.  Overall, here are some observations as Caffeine prepares to roll out:

Significant displacement is revealed among big brands for some broad terms.  After reviewing a number of high-volume category level terms, we found many big brands have swapped places in the results for individual key terms, while others have risen to new heights, and others have gone down to new lows.  This is no small news to either the winners or losers, as some of these category-level terms may be generating millions of dollars per year in revenue, per term, alone.     

Less emphasis on universal results.  While it appears that universal results like video and images may be displaced by more blue text links in this update, don't expect them to go away, and don't stop optimizing various digital assets types.  In fact, they may eventually come back at a greater frequency in future updates, as Google continues to test the right balance of universal in paid placements.  Google has reported previously in interviews that organic universal results were cannibalizing paid results, and a future rollout of universal paid placements may help offset this issue until video and images can reappear more prominently without as much impact to revenue.

Some social media sites may gain in favor, while others decline.  If there is any consistency in search updates like this, it's that positive and negative favor toward certain sites changes constantly.  While big brands will eventually feel the algorithmic impact of the shift, sites like Facebook and Technorati generally seemed to have benefited, while Wikipedia may be slightly less emphasized throughout the results.  Twitter may also get dialed up a notch or two in this update. 

For enterprise marketers, the same SEO principles still apply.  At the end of the day, what enterprise marketers get from natural search is impacted by 1) how you maintain your digital assets (things marketers control on- and off-site); 2) user behavior; and 3) the level of algorithmic bias toward your site.  Number three can't be controlled, and I would advise readers to stay away from any SEO who says it can be; two is also not fully in our control, as evidenced by the major massive changes in search behavior as a result of the economy.  But number one is in our control, and at the end of the day, best practices still apply.  Staying focused on this area keeps your program focused on gaining sustainable long-term natural search benefits, and not taking a reactionary approach that might jeopardize future returns for the sake of trying to spare last week's positive algorithmic bias.  This means staying focused on core elements like content, link structure, site architecture and structure, promotion in the right offsite networks, and all other general best practices. 

So for those who are already taking the best-practices approach, step back and prepare to observe the change.  For those shifting with the wind, I would strongly recommend moving to best practices, and also taking this update somewhat in stride.  While the best-practices strategy is always subject to good days and bad, it is still the best and most sustainable long-term approach.  Again, note that the observations above come from previewing results, and may be subject to change upon final rollout.  Just be ready to analyze the impact when this happens -- which, for your business, could end up meaning nothing, a little, or a lot.