Debunking Long-Tail PPC Search (The Google Instant Connection)

How long is too long (into the search keyword tail)? Very few people seem to take the time to think about this question before embarking on a long, arduous, and potentially wasteful keyword expansion task. More keywords for the sake of more keywords is not a valid PPC search strategy, nor is it a good use of your time (or the time of your internal or agency staff). Every advertiser is different, but you need to put some serious thought into when to expand keywords vs. when to leave them alone. Plus, when you do decide to expand your keyword variations further into the tail, you need to have a plan that tells you how far into the tail to go.

Google Instant and the “Tail”

Some “experts,” pundits, and analysts believe that Google Instant will shorten the tail through the curation of the SERP. However, early data hasn’t yet validated that hypothesis. For example, looking at the log files of the PowerProfiles business directory, the long-tail search traffic which comprises the vast majority of visits has not changed appreciably since the launch of Google Instant. Client data (taking into account standard seasonality and volatility) does not reveal any systemic change in keyword query length or traffic behavior/makeup, but the data sets so far are still small, so it’s too early to be sure. Plus, consumers’ adoption of Google Instant may not be uniform or even at a high level among power searchers. Regardless of whether or not the long tail gets shorter, there is a point of diminishing marginal returns with regard to extending a campaign into the long tail. Finding that point of diminishing returns for your business is the key to a balanced, effective campaign.

When I ask search marketing practitioners when or how often they expand keywords, you’d be amazed at the diversity of answers I get. Similarly, for product-based advertisers, the number of keywords pointing to a specific product landing page can vary from one to hundreds. The important thing is to figure out the right way to think about and calculate the best campaign structure and keyword mix for your business.

How many hours have you spent in the last year expanding your keyword list further and further only to find that:

1. Ad quality score didn’t go up
2. You are now stuck with a ton more match types on long queries to consider (do you extend your campaign in exact, or broad and, if broad, why not simply capture them with a shorter broad match listing?)
3. Your campaign just got more complicated
4. You get just as many clicks as before (because they were captured before using broad match or phrase match with negatives)
5. Conversion rate is the same as before
6. Landing page management got much more complex

Optimizing Campaign Breadth Through Behavior and Intent

There are two ways to get to the optimal breadth of a campaign and they are not mutually exclusive. The first is to simply work off the data. Use your Web analytics, campaign management technology, or the “Search Query Report” in Google to get an understanding of your current makeup of queries. Pay particular attention to the number and percentage of “other unique queries” for an ad group. A large percentage of “other unique queries” warrants at least a closer look. This leads us to the less quantitative way to evaluate the breadth of your campaign, using behavior and intent. If searchers have the same or highly similar intent and are only articulating their intent differently as a search, then we would expect that the same landing page would have similar conversion rates. So, breaking up an ad group may not make sense here. However, when there is a high volume of repeating phrases that are occurring as part of a broad match listing, the opportunity exists to create an ad specifically tuned to the searcher’s expressed intent, even if you use the same landing page.

You’ll have to decide the cutoff that’s right for you. Some search marketers would prefer not to have to worry about another ad group for 10,000 impressions a month of differentiated queries. For other small businesses, even 1,000 repeating unique queries is worth splitting off an ad group in an attempt to get a higher quality score. Very large marketers might find it better to take a very high volume “head” keyword (what I like to call a power keyword due to the opportunity locked in that keyword) and segment the campaign and ads around that search to treat geographies differently. That’s a completely different “tail” than the one commonly pursued. Custom creative and offers by geography, plus the ability to manage bids separately, may be just the ticket.

When exploring the long tail in either keywords or other targeting options (yes, Microsoft adCenter gives you demographics), take a scientific and rational approach and use your head (pun intended).