First off, apologies for my absence from the blog these past few days. It's been an incredibly busy time, trying to wrap things up before I leave for San Diego over the holidays. So much for a December lull... In this post, I'm going to try tackling a lot of the recent trends we've been observing from the engines and talk about my personal perception of what's to come over the next 12 months.
#1 - This Real-Time Search Thing is Outta Here
Microsoft initially beat Google to the punch in announcing their integration with Twitter data in their SERPs. And in response, last Monday, Google released what is, in my opinion, an early test version of Twitter integration that's nowhere near ready for prime-time. Google has a history of jumping the gun to prevent other companies from stealing the press narrative, but in this case, I think it's seriously damaging (and nearly everyone, consumer or search enthusiast, agrees) their usability and relevance.
As Danny Sullivan notes, it's like we're back to Infoseek in 1997. If you want to rank #1, don't worry about quality content, relevance or popularity, just be the last person to Tweet about a topic and you'll come out on top (at least, for a few seconds).
This is, in my estimation (and many others), the worst implementation of new results Google's ever implemented. I imagine the clickthrough and abandonment stats have their usability folks up in arms already, and it's only to preserve face from a PR perspective (as well as an increasingly prideful attitude of "Don't like it? So what are you gonna do about it?" that Aaron Wall describes in a gutting fashion here) that this has stayed in place as long as it has (1.5 weeks).
In 2010, I think this fades away. Perhaps not entirely, but we won't be seeing it for nearly as many queries with the prevalence we do today. Google may love real time, and it's certainly gotten them a lot of press (though very little of it is entirely positive), but they can't continue sacrificing quality for PR in this fashion. I think the engineers still run things over there, and the stats data is already making them balk. Although I don't have numbers, my impression is that we're already way down in the quantity of queries showing real time results compared to last week.
#2 - Twitter's "Link Graph" is the Real Deal
All that real-time integration bashing aside, I'm a firm believer in my original hypothesis that Twitter is cannibalizing the web's link graph. In fact, I think a rough history of "recommendation sources" looks something like:
Google has always strived to keep up with the latest ways that content is being recommended and suggested. It's how they determined popularity and relevance with PageRank and I think Twitter's data is merely the next evolution. Just yesterday, they launched their own URL shortening service (I think this was more to get data, but it's also possible it was a pre-emptive PR strike against bit.ly, who launched their PRO service just a day later).
Google's not going to just take raw number of tweets or re-tweets. I think we're already seeing the relevance and reputation calculations in their decisions of which tweets and sources to show in the real-time results, and I expect that algorithms/metrics like PageRank, TrustRank, etc. will find their way into how Google uses the real-time data. Today, SEOs want to turn tweets into links so they can get SEO benefit. My feeling is that tweets are going to carry their own weight in helping pages rank in the not-too-distant future.
#3 - Personalized Search is Here to Stay
Unlike real-time's temporal nature in the results, I think personalized search is here for the long haul. Google released their "permanent" personalization of results last week, and Bing released their own just this week. As usual, SearchEngineLand's coverage is impeccable, though one big question remains in my mind:
What metrics impact personalization?
Is it merely clickthroughs from the organic results? Does visit history play a role? Or clicks from other vertical search services Google offers? What about clicks from paid search ads - either in the SERPs or from AdSense/DoubleClick?
I'd love to see experimentation done on this front so marketers have a better idea what they're dealing with. If it's proven that you can get organic benefits by attracting PPC clickthrough, this may be the new "paid inclusion" for 2010, and could drive bid prices up massively as companies compete not only for paid listing clicks, but for the chance to earn "organic" positioning as well.
Personalization means a few things for SEOs, but it doesn't fundamentally change the game, IMO:
- The Rich Get Richer - It's now truer than ever. If you rank well, and earn solid traffic, you're going to be even harder to unseat. Startups and upstarts are going to have an even greater uphill battle to climb than before.
- Branding is More Important - you want your loyal visitors and fans scouring the SERPs for your listings, and clicking them more so than anything else. I expect some clever spammers are going to be manipulating this with everything from Mechanical Turk to virus infections that make their browser search for their brand and click those results. We'll see if Google has good protections in place to defend against this.
- There is No Normal Ranking - Or, at least, there's no "normal" ranking that's "average" in a personalized SERPs world. Rank tracking may still carry some value to understand how non-personalized searchers see your pages, but that data is going to be less useful in comparison to what your analytics report about search traffic and the trends. Win the "personalization" battle, and you may start to care less about the classic "rankings" battle.
Whenever we encounter these "paradigm changing" events in the SEO world, I like to go back to myphilosophy about SEO fundamentals. From what I can see, it looks like things haven't changed enough yet to warrant panic. It's been a massively dynamic 3 months, but we're not on the precipice of anything that's going to shift SEO in the ways some previous "game-changers" have.
#4 - It's Going to Be a Two-Engine, 80/20 World
The latest figures suggest that Google continues to slowly gain market share in the US, while Bing & Yahoo! compete for share that will eventually belong to them both (once the regulatory hurdles clear, which I think they will). I believe that a year from now, most webmasters will be looking at a scenario where Comscore/Hitwise reports Binghoo! has ~25-28% market share, but those engines combine to send a little under 20% of all search traffic (remember that they count searches on all Microsoft and Yahoo! properties - even internal searches - while Google tends to send the vast majority of their search traffic externally to other sites).
#5 - Site Explorer & Linkdomain will Disappear
Tragically, everything I hear out of Yahoo! and Bing is that Site Explorer is off to the great beyond. The expense of maintaining a web index isn't something Yahoo!'s willing to invest in once they don't have to, and Bing's given no indication that they're going to re-open the portal to link information. The best we can hope for is an acceleration in the functionality offered by Bing Webmaster Tools, but even that's unlikely to offer competitive link intelligence.
I'm guessing other services will rise up to try to take Site Explorer's place, as the service had millions of monthly queries run against it.
#6 - SEO Spending Will Rise Dramatically
Forrester put out a great report on US Interactive Marketing Spend (a little pricey at $1749, but interesting). Two graphics struck me as particularly compelling:
SEO trails only social media and online video as places where marketers (not just search marketers, but ALL marketers) will be shifting dollars.
Meanwhile, SEO continues to outpace PPC in terms of CAGR. We've still got a long way to go before balance is established between the share of clicks SEO commands and the fraction of spend it receives, but the gap is slowly closing.
#7 - 2010 is the Year of Conversion Rate Optimization
If I were doing another startup today, it would focus on software for conversion rate optimization. I think this is still the most under-utilized and highest ROI activities in the marketing department, but more awareness is on its way. CRO isn't just about testing; it's about building a process for improving converion over time. Online businesses can generate so much revenue from this, yet few invest. I think 2010 is the year, simply because it's an inflexion point for companies to assess their spend and where they derive value. These guys are likely in for a blockbuster year; I wish I could invest :-)
This graphic comes via my post on choosing which Internet Marketing Channel to Pursue.
#8 - More Queries will Send Less Traffic
Google & Bing are both doing more to make their visitors stickier and get their queries answered without ever having to leave the engine. This is a good product practice for both companies, and I'm surprised Google's taken so long to move away from their "get people off Google" point-of-view, but it's definitely happening. Check out some recent examples:
Everything I need to know is right there - the last game score, the record, the opponent, their next match day and time. The only thing missing? What channel it's playing on in my area.
I don't even have to complete my query! Google's got that weather report sitting in the suggest box. Theywrote about this feature here which launched last week. Google O/S had another good post on the topic.
Thankfully, I'm not actually headed to Kodiak, but those results are pretty spiffy, and are likely to prevent me from needing to visit Alaskaair.com and get that flight info.
The customer service number is something Bing's started to provide more and more (though there's one company even they don't have that data on). With Fedex, you don't even need to leave Bing to track a package (Google also offers similar functionality).
My perception is that the more the engines can apply "instant answers" to search queries, the more they will, and the less any other sites will see traffic from those queries. It's a better user experience this way, and I'm certain it's one of the biggest things that engenders loyalty and return queries - something both engines are desperately competing for.
This post isn't intended to be one-sided, and I'd love to hear from you - do you agree? Disagree? Think I'm out of my head? Let everyone know :-)
