I had lunch today with a young, ambitious entrepreneur. He’s fresh out of a prestigious tech university, has a fledgling company and visions of shaking up a big industry. He wanted to talk about search.
Him: “I’m pretty psyched, we just launched about 4 months ago and we’re already ranking number one for <insert obscure phrase>.”
Me: “So, how much traffic is that generating?”
Him: “Well, none.”
Me: “So how useful is that ranking?”
Him: “Well, if we can rank for that, I just need to get us ranking for terms that will get us traffic.”
Me: “errr. . .”
I hate ranking reports. They are used by agencies to validate retainers. They are cited by businesses to suggest success. In reality, they cloud the issue.
They convey progress while hiding failure. They distract from business goals and promote the misallocation of resources. Talk to me about traffic, but don’t show me another damn ranking report. Here’s why:
An increasing number of searches return personalized results; this can be based on things like previous user behavior (i.e. my search for “news” returns CNN, because I visit CNN so frequently) or geo specific searches (i.e. my search for “restaurant” returns Seattle restaurants) or searches based on a social graph.
Worst case, a ranking report can reflect atypically high results from an employee’s frequent visitations of her own site. How widespread is personalization and for what types of searches? No idea – so why report on it?
Local & Other Stuff
Search results are no longer a simple list of results with ads sprinkled around.
How do you report rankings for searches that include local results (accounting for some 30% of all searches)? What about when video shows up? News? Images? How do we count these? How do they impact user click behavior? No idea – so why report on it?
It’s not like results are static – experience shows that as you go higher in results (20+), ranking results can fluctuate wildly.
Additionally, the engines switch up the format all the time as they optimize and test. For example, today’s local results may look totally different that tomorrow’s. How can you have a consistent metric? No idea – so why report on it?
Ranking reports only show a score for a specific term, they do nothing to demonstrate success or failure for multiple variations of a similar query that should drive traffic to a given page.
For example, how successful is our Seattle DUI Lawyer page at getting traffic for “seattle dui lawyer”, “drunk driving attorney seattle, wa” and multiple variations thereof. Does your page capture traffic for query variations or not? Can’t think of all the variations – so why report on it?
Ranking Reports & Agencies
Agencies that are evaluated (and compensated) on their ability to move those ranking reports can be very dangerous.
Moving specific terms can be most easily accomplished by seo optimized links, which opens the door for agencies to engage in short term spammy linkbuilding tactics that can have site-wide, long term consequences.
Remember The Long Tail?
Remember how excited we got about the long tail? A focus on ranking reports can easily channel your efforts to a finite set of terms way up on the tail.
Now consider your given resources, how many terms can you realistically impact? 20? 100? 200? Now how many pages does your site have? Why are you ignoring 99% of them with your silly ranking reports? Why aren’t you creating more long tail content instead?
Ranking reports aren’t completely useless – every site should have a finite set of what we call ubercompetitive terms – those terms that really define your business that you are competing aggressively for. And by finite, I mean a 1-2 non-branded terms max for most of you, 3-6 for some and more than 6 for a very select few sites.
Additionally, generating ranking reports for online reputation management is absolutely appropriate. But for the most part, stop reading ranking reports. Start reading traffic reports.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.