A customer of ours sent us a question about citations. They have been building citations recently and cleaning up NAP inconsistencies. But they have seen a significant drop in rankings.
I’ve never heard of ‘good’ citation building having a negative effect on rankings – have you?
Also what’s your theory on speed of building citations and if you build too rapidly do you see Google treating this similarly to building links too quickly?
These are great questions because they touch on virtually every aspect of local ranking and logical thinking. Rather than just reply via email I decided to respond publicly so that all could join the discussion. Can a citation building campaign have a negative effect on rankings? Can you build citations too fast? The short answer is NO. For the longer answer read on.
As is often said in the SEO field (to the point of cliche) correlation is NOT causation. Certainly correlation of a good data set to certain events can lead to more understanding of a situation but a single instance is a particularly weak data set. Humans have a tendency to see patterns and relationships where none exist. Search involves a particularly complex set of variables many of which we don’t even know.
Firstly let’s look at citations in isolation. The path of a citation into the Google cluster for a local business is often long and circuitous. An entry is made at a local directory; Google must scrape that deep interior page of the local directory and add the changed data into their main index; Google then must rebuild their Maps index from the recently scraped data and then rerank the listings inside of the Google local index. Google’s local index build occurs roughly every 6 weeks although the schedule is not fixed and it could be from 4 to 8 weeks. Many local directories take time to update their listings once an edit has been made and the page that the listing is on is of such low page rank that it may take weeks or months for Google to scrape it into their main index.
Thus a direct entry to a local directory could take anywhere from 3 weeks to 3 months to make it into the cluster. In addition if you added the listing to a major list supplier like InfoUSA or Localeze you have to add the time for them to vet the listing and provide a feed to Google. In the case of InfoUSA this could be as long as 3 months. Add the time delay for submissions from a company like UBL and you could easily see a 8 month timeframe before all citations have made their way into Google’s index. Thus not only do they often take a long time, they would come into your business cluster at a very erractic rate.
Secondly, citations, like directories before them, have become less important in the Google’s Blended ranking algo.
Thirdly, Google’s local penalties have become more aggressive and often roll through the index in waves. Within the last quarter we have seen increased penalties for service area businesses that show addresses inappropriately, there have been massive review take downs affecting rankings and old penalties like excessive geo-cramming still exist. Without looking at the listing we have no idea if that is a possible cause for a drop in rankings.
Fourthly, ranking is dynamic process, it is not just you and Google. Competitors are changing as well.
Fifth, Google still has troubles with dupes and merges. Given that your client has NAP inconsistencies there could be a duplicate listing that is splitting the strength of the cluster.
And last but by no means least, Google’s local search ranking algo has become increasingly complex and the predominant blended results use both organic and location ranking factors in various combinations. All of the recent Google organic algo changes and updates including Panda, Penguin and unnatural links could be affecting the listing.
So to even to begin to answer the question we would need to know where the listing ranked prior to the drop, how much it dropped, whether the drop was local or organic, what competitors were doing and what the quality of the listing is.
Search results are a multifaceted and complex animal. Just because a person takes an action A and sees an outcome C doesn’t mean that there is a causal relation. They could be related but in this case it strains credulity to think that they are.