Accuracy of our search volume data

We have spent many months of research, experimentation, AI training and adaptation to ensure the best possible precision.

Overview

Getting search volume has become a subject for a scientist, that's not what we want our customers to suffer from.

Google has changed the Adwords Keyword Analyzer more than one time in the past years, each change was a blow against SEO services.

• Google removed the option of receiving 'broad match' search volume, an essential information for IT specialists.
• Google removed accurate results for normal customers, only high paying customers receive meaningful data at all.
• Google introduced a total page-view limit for their keyword analyzer, crippling even their largest customers.
• Google stopped reporting accurate keyword information, they started to group keywords together making it impossible to judge many keywords.

All these changes are a major problem to SEO providers just as well as to website owners and marketing specialists.

How other services work

Most services seem to rely on a database they scraped from adwords, they have up to millions of keywords scraped and use those.

Some services use Google trends to get an estimation about the search volume.

The values they obtain these ways are outdated, inaccurate and error prone.

Now they use simple linear algebra combined with randomization to generate "real looking" numbers.

The result is nothing you could rely on, it's actually random data with some outdated old information mixed in.

How we tackle the situation

We obtain a large amount of different datasets about each keyword.

All these many different datapoints are normalized and filtered.

Now self learning neural networks are used to fuse all these datapoints together and adapt the search volume to highest precision.

The result is a highly accurate "broad match" search volume and in the large majority of cases a very accurate "exact match" volume.

We have compared the results with Google adwords and reached accuracy values up to 99.8%.

We do not use pre-collected databases at all and we do not spike our results.

Every single keyword is fresh calculated at the moment of scraping.

From our knowledge we are unique, no other service can currently compare with us.