The Proper Way to do Keyword Research in 2016

[JoyHawkins]

I read this article by Patrick Stox and had to share it with everyone.

It's extremely common for people to use the Keyword Planner the wrong way and just look at the difference between what I get when I use it the quick (wrong) way and what I get when I do it Patrick's way. HUGE DIFFERENCE!

Wrong Way



Right Way

 

[James Watt]

Thanks for the share Joy, I didn't know you could get better data that way, good to know. The adwords keyword tool has been pretty unreliable all along, and it's been not worth using their volume data after they started grouping. I'll definitely be using that as part of my research stack now.

 

[Eric Rohrback]

I'll share two tools that are in my arsenal these days (disclosure: they're paid subscriptions); Moz Keyword Explorer and SERPwoo's Keyword tool.

I was a beta tester for Moz's tool and it looks like a lot of the feedback for changes I had got implemented in the final version. There were things like explanations, different filtering options, etc. I guess enough people had the same concerns and they definitely did a great job with the final product.

SERPwoo's is a different, but pulls in a lot of great ideas to start working off of. The big upside I see with this tool is that it lists which sites (within the top 10 SERP results) are targeting each keyword. Very cool feature. This has search volume and CPC data, so it is a good alternative to KW planner.

I like Patrick's method Joy mentioned, that seems a little more reliable than KW planner's default settings. I think to really make keyword research work you need a couple of sources/datapoints to compare. Each tool will be different, so it might make sense to work with a few data sets. If you're handy with excel you can use macros or referencing tables to consolidate and pull the most important keywords.

Stepping away from only looking at the tools used, what is your typical process for keyword research? How do you find the gems in the mountains of data?

 

[JoshuaMackens]

Why do the impressions keep evening out? It's in your example and in Patrick's? Suspicious on Google's part...

Am I missing something?

 

[JoyHawkins]

Sorry Joshua not sure I follow?

 

[Patrick Stox]

Why do the impressions keep evening out? It's in your example and in Patrick's? Suspicious on Google's part...

Am I missing something?

They seem to take the middle of their Monthly Impressions forecast. In Joy's example 3.6-4.4k, middle is 4k so that's what they end up totaling for the impressions.

 

[JoshuaMackens]

They seem to take the middle of their Monthly Impressions forecast. In Joy's example 3.6-4.4k, middle is 4k so that's what they end up totaling for the impressions.

Sorry Joy, that's what I meant, the impressions count.

It seems like Google loves rounding. Can we really use this as an accurate estimate? Or should we use it as we did in the Keyword Planner as less of a traffic estimate but more of a relational model between keywords ie "toronto plumbers" gets 4,000 impressions, "toronto plumber" gets 8,000 impressions. We should use this data not to dedeuce that "toronto plumbers" literally gets 4,000 impressions and that "toronto plumber" literally gets 8,000 impressions but more so to say that we should focus more on "toronto plumber" as it gets more impressions than "toronto plumbers".

Does that make sense?

Do you think that's an accurate take or do you think the impression counts are somewhat accurate?

 

[JoyHawkins]

Josh,

I think the impressions are "somewhat" accurate in the sense that it gives you an idea if there are 10s, hundreds, or thousands of searches for that keyword. Like you said, I generally use it to give me an idea of which keywords to target or which ones might be the ones everyone is trying to target. I don't take it as an exact science. The only thing I really rely on for real numbers is ACTUAL AdWords data from clients.

 

[JoshuaMackens]

Josh,

I think the impressions are "somewhat" accurate in the sense that it gives you an idea if there are 10s, hundreds, or thousands of searches for that keyword. Like you said, I generally use it to give me an idea of which keywords to target or which ones might be the ones everyone is trying to target. I don't take it as an exact science. The only thing I really rely on for real numbers is ACTUAL AdWords data from clients.

Agreed. Adwords impressions seem to be the best way to measure.

But even then your client has to not have an ad schedule and max budget so the ad always shows up.

Anything else you can think of that I'm leaving out?

 

[mborgelt]

I've always found it interesting that Google isn't particularly fond of helping you use their tools the right way. Perhaps they have bigger fish to fry but it seems like apps like Facebook, Mailchimp and Leadpages are a lot more prone help you use their tools effectively.