Google Keyword Unplanner – Clickstream Data to the Rescue

Let's start with the happy ending, which is actually a happy start, too. Moz Keyword Explorer used data keyword clickstream derived in a new way since day 1, allowing us to provide keyword volumes consistent despite dramatic changes Google Planner keywords in data availability and reports. You have probably not noticed changes in our volume of keywords, and you probably will not notice any move forward, which is just the way we built it to start: resilient, changing, and trustworthy.

That said, the truth is that the keyword data was recently on land sliding as the foundation on which most keyword tools are built - Google Keyword Planner - has been severely disrupted. This single point of failure put a lot of risk tools, so let me explain how we have addressed this concern preemptive and subsequently has not lost a step.

Problem 1: Scheduler keywords began aggressively grouping keywords

You've probably seen this story floating around for some time. Google Keyword Planner has always combined a few words, especially the spelling mistakes, so when we built Moz Explorer, we have already planned a strategy to address these whenever possible. It is that same technology volume disambiguation works for other types of mixed terms. For example, groups of keywords Google Planner "SEO" and "Search Engine Optimization" together, recognizing that it is an acronym of the other.

seo-exampleAs you can see, the key words of the planning reports "SEO" and "Search Engine Optimization" as having the same average monthly searches and suggested bid price. Worse, because Google has grouped the words when making forecasts of volume, but not grouped the words in the construction of the graph, it appears that if you were to advertise on these two words, you get more than 200,000 impressions per month (at least, from the graph). Well, you do not have to worry about it if you are a user Moz Keyword Explorer, because we get on the right, showing the two sentences as having different volumes in the correct proportions.

seo-example-moz-02seo-example-moz-01Another classic example of the keyword group that we see in the Planner keywords is bound to contain. Take for example the word "game", which is also the stem of "play" and "play." Google groups these three words together in Keyword Planner and presents them as having the same average monthly searches and suggested the offer. Again, we see the same graphic problem as well, where it seems that someone rankings for these terms could benefit almost 1 million searches per month. This is actually a misrepresentation already grouped keywords.

Digital Marketing Course in Hyderabad

play-exampleSometimes you can get lucky and if the keywords are commercial enough, you can see their actual proportion in relation Keyword Forecaster. It's not always the case. Forecaster has a very particular behavior when it receives a keyword grouped as a spelling rather than just a similar term. This differential treatment of semantically related terms lexically vs Forecaster makes a reliable replacement for Keyword Planner alone, but in this case it is a decent figure. If we were to define identical results in Google for these terms, the key word "play" would be much more impressions and clicks "play" or "parts".

play-example-forecasterWe can confirm this with our clickstream data, which gives similar performances. We can marry clickstream data with historical data, data forecaster and planner of data to build our own volume forecasts.

play-example-clickstreamWhich when everything worked, looks something like this:

play-play-moz moz moz-game

Problem 2: Keyword Planner began limiting access to raw data for non-active campaigns users.

In perhaps a bigger bomb announcement, Google began obscuring data to users who do not spend enough money on Adwords. The beaches are very large and, frankly, unattainable for those seeking to do the keyword research (for Adwords or SEO). But, again, the mixed technology Moz Keyword Explorer keeps us ahead of the curve. Although we were never able to get the keyword volume again from keywords Google Calendar, we would be able to continue to provide users with a stable set of volume measurements that closely models actual Google search volume.

How we do it:

1. How can we determine when the words are grouped?

This is an area where size really matters. Moz has a huge keyword corpus of over 2 billion keywords, and we raised the volume of Google for hundreds of millions of them. Because of this, we can identify the rare occasion when two words have the same historical research data (same CPC, competition, volume, etc.). Sometimes two words share the same history by chance, so that we then use a variety of NLP and chain-similarity measures, including a learning model in incredible depth constructed by Dr. Matt Peters to determine whether the keywords are linked to each other. It is important to use multiple methods because the string-similarity methods are notoriously fickle. Once we apply these various parameters string similarity to all keywords with the same metric, we can identify those who are grouped by keywords Planner.

2. Once we know what the words are grouped, how the volume of each?

Once we have a group of related terms, we apply a predictive model based on data from both Google and our clickstream sources to determine the appropriate percentage of traffic that should be allocated to each word or phrase. Again, this is where having a broad set really shines data. Without detailed information on the constituent sentences, we would have to make unwarranted assumptions about how to divide the bulk volume. Fortunately, this is rarely the case, and we choose to be explicit with our customers and the state "no data" when we do not have enough data to make a prediction.

3. How can we determine the volume of keywords when we do not have data Planner Google Keyword?

Fortunately, we can rely on our vast clickstream data for these calculations. clickstream data is inherently noisy and biased, so our models are complete enough to eliminate random occurrences, strip bias in the sample data, and the model projected against the general corpus Google traffic. There is a chicken / egg problem here, to a degree, because we can not model against Google's data if it has grouped-keyword problems, but we can not solve all the problems of keywords grouped without clickstream data. However, as long as we are reasonably certain that clickstream data is proportional internally, then we can count on it to solve the first problem grouping, then use the data of keywords separated planner to model against data General clickstream. It is a complex procedure, but in the end, we can reasonably predict the monthly search volume without ever having data from Google.

Let me give you an example. Khizr Khan, father Purple Heart recipient Captain Humayun Khan, caused a political uproar following his speech at the DNC convention. His story is a common problem in the keyword data in that, before his speech, nobody ever searched his name. After his speech, his name shot at Google Trends but even then the Google Keyword Planner was delayed in reporting its numbers due to long months of delays in data dissemination. Because our clickstream data can pick up on trends, we can predict the volume of Google without the need for data Planner Google Keyword.

Khizr Khan-example

This is also the case for keywords that are not a trend. If we see a term that is regularly searched in our clickstream data, but are not represented in our set of Google data, we can make predictions without having to rely on misleading potential (volumes grouped) or inaccessible data sources that Google Keyword Planner has become.

A long story short

If you are a user password Moz Explorer, you can be sure we will continue to deliver metric state-of-the-art, regardless of the difficulty of Google provides keyword planning data. This is just another way that the keywords Moz Explorer continues to lead the way in the search keyword. If you need keyword data, come get it.