Why Facebook's 'Graph Search' is a Bigger Deal than You Think.

There's a reason Google's been force-feeding us Google+ for 2 years now, and it has little to do with competing with Facebook as a social network, at least on the surface. 

Facebook 'Graph Search' may just change the search game forever

With Facebook's unveiling of 'Graph Search' today, they've finally begun the process of monetizing - in earnest - the enormous amount of user data they've been hoarding over the past decade. Apparently all they needed was some former Googlers to show them how; namely, Facebook's Director of Product Management and Facebook's Director of Engineering both have histories with the search giant.

Most notably, however, is Marissa Mayer, who famously fled Google's ship to take the helm at the floundering Yahoo search engine as CEO.

There may not be an immediate link between Facebook, Google, and Yahoo, but it's there, and it makes the launch of 'Graph Search' not just interesting, but downright juicy.

Yahoo and Bing, despite appearing as "adversaries" in the natural search world, have been in overt cahoots since 2010 in paid search, ever-attempting (and just as often failing) to gain market share against Google who owns nearly 80% of the search pie. Since paid search is the proverbial "cash cow" in this equation (Google reportedly makes over $100 million every day from paid search), Bing and Yahoo (collectively known as Adcenter) need a competitive advantage when it comes to paid search to make any sort of movement; an edge it hasn't had.

Until now.

Uh, Google? You look desperate

As the primary search engine behind Facebook's 'Graph Search' tool, Bing is set up to reap long term benefits in a way that Google has been clamoring to achieve via Google+. And while Bing now has access to over 900 billion sets of data from Facebook's database, Google's social network might as well be throwing pebbles at the upstairs window, hoping to be invited into the party.

With Facebook's 'Graph Search', Bing, and likely Yahoo by relation, will have the inside track on providing real-time, relevant, social search results to users by accessing Facebook's data set. Google on the other hand must continue to rely on '+1's' or other tame dynamic sources. Surely nothing that is as robust, personal, or connected as what Facebook could provide.

While Facebook's stock was dropping, the fact stood - Facebook has an unfathomable number of users and their data. By virtue of that fact alone, Facebook is viable. Bing and Yahoo know it. And perhaps Google knows it now more than ever.

Ditch the Broad Keywords and Embrace the Results

Broad match keywords no longer make sense for paid search advertisers. It's no secret that Google (or any other search engine) uses the "broad" keyword match distinction to boost possible impression share and thus click totals and - spoiler alert! - the money that they make.

But many advertisers lately have noticed the further loosening of relationships between broad keywords and search queries (have you looked at a search query report lately?), leading to a degradation of click through rates, quality scores and valuable clicks, and resulting in wasted spend.

The case against broad match

Here is the definition of broad match, from Google:

Broad match: Like using a chainsaw for surgery

...broad match means that your ad may show if a search term contains your keyword terms in any order, and possibly along with other terms. Your ads can also show for singular or plural forms, synonyms, stemmings (such as floor and flooring), related searches, and other relevant variations.

With the way people search and respond to search results, broad match keywords are losing importance. Most studies show that searching behavior continues to trend towards specific, long tail terms, while users become more in tune to what they are looking for and how to get the best experience from their query, which might include the search terms themselves, which ads/links they decide to click on, and how they react to the landing page.

Unfortunately, broad keywords offer little in the way of specificity or relevancy, resulting in lower click through rates due to poorly matched queries or irrelevant ad messaging.

Furthermore, Google's insistence on using session-based broad matching makes results even less reliable. In a clear attempt at increasing impressions and spend, Google decided that it would be useful to show ads based on a user's previous search, even if their current search is not as relevant. This is somewhat akin to "remarketing" using search ads. The problem is, a search inherently represents a request to solve a specific problem, not a problem that they had before.

The Case for Modified Broad Match

Improve your toolkit with Modified Broad Match

Broad match may have been handed it's death certificate with the advent of modified broad match, which is basically a supercharged version of the original. (If you need to know exactly what this does, Google explains it well enough.)

If you're using broad match right now, you'd do well to take your head terms - the keywords that are most important - and "modify" them so they are required in searches to appear. For example, if you are targeting "car insurance", some variations of broad match keywords may include "best car insurance" or "cheap car insurance". At the very least you should find and replace "car insurance" with "+car +insurance", which tells Google that both keywords must appear in the search query in order for an ad to appear. Building keyword lists using this model allows greater control over what your ads show for, leading to better ad serving and relevancy.

How to Optimize Keyword Match Types

So you've been running broad match keywords for awhile now, seeing modest results but perhaps lower-than-desired CTRs, CPCs and conversion rates. Now is the time to optimize your campaign by removing broad match keywords in favor of more effective match types.

1. Change all important broad match keywords to modified broad match, noting the "head terms" and key qualifying terms as the "modified" terms (see above). Acquisio has a useful tool for doing this, or you can simply use the find and replace function in Adwords Editor or excel.
2. Fill out phrase match ad groups with top keywords. Scan top clicked keywords and run a diagnostics to make sure all of them are covered as phrase match, then scan to make sure different [logical] phrase variations are included as well.
3. Scour search term/query reports for new long tail keywords. Long tail keywords will be your highest converting, lowest cost keywords. They typically equal about 20% of all clicks, but why stop there? Maximize the number of detailed keywords in your account for more relevant matching.

Broad match type for paid search keywords has become increasingly irrelevant. By optimizing your account to remove this type, favoring modified broad match and phrase match, you are setting yourself up for higher CTRs, better quality scores, and improved conversion rates.

Effectively Building Fragmented Content in 2013

2013: The Year of Content