Keyword Research is a very different field than it was just five years ago, and if we don't keep up with the times we might end up doing more harm than good. From the research itself to the selection and targeting process, in today's Whiteboard Friday Rand explains what has changed and what we all need to do to conduct effective keyword research today.
For reference, here's a still of this week's whiteboard. Click on it to open a high resolution image in a new tab!
What do we need to change to keep up with the changing world of keyword research?
Howdy, Moz fans, and welcome to another edition of Whiteboard Friday. This week we're going to chat a little bit about keyword research, why it's changed from the last five, six years and what we need to do differently now that things have changed. So I want to talk about changing up not just the research but also the selection and targeting process.
There are three big areas that I'll cover here. There's lots more in-depth stuff, but I think we should start with these three.
1) The Adwords keyword tool hides data!
This is where almost all of us in the SEO world start and oftentimes end with our keyword research. We go to AdWords Keyword Tool, what used to be the external keyword tool and now is inside AdWords Ad Planner. We go inside that tool, and we look at the volume that's reported and we sort of record that as, well, it's not good, but it's the best we're going to do.
However, I think there are a few things to consider here. First off, that tool is hiding data. What I mean by that is not that they're not telling the truth, but they're not telling the whole truth. They're not telling nothing but the truth, because those rounded off numbers that you always see, you know that those are inaccurate. Anytime you've bought keywords, you've seen that the impression count never matches the count that you see in the AdWords tool. It's not usually massively off, but it's often off by a good degree, and the only thing it's great for is telling relative volume from one from another.
But because AdWords hides data essentially by saying like, "Hey, you're going to type in . . ." Let's say I'm going to type in "college tuition," and Google knows that a lot of people search for how to reduce college tuition, but that doesn't come up in the suggestions because it's not a commercial term, or they don't think that an advertiser who bids on that is going to do particularly well and so they don't show it in there. I'm giving an example. They might indeed show that one.
But because that data is hidden, we need to go deeper. We need to go beyond and look at things like Google Suggest and related searches, which are down at the bottom. We need to start conducting customer interviews and staff interviews, which hopefully has always been part of your brainstorming process but really needs to be now. Then you can apply that to AdWords. You can apply that to suggest and related.
The beautiful thing is once you get these tools from places like visiting forums or communities, discussion boards and seeing what terms and phrases people are using, you can collect all this stuff up, plug it back into AdWords, and now they will tell you how much volume they've got. So you take that how to lower college tuition term, you plug it into AdWords, they will show you a number, a non-zero number. They were just hiding it in the suggestions because they thought, "Hey, you probably don't want to bid on that. That won't bring you a good ROI." So you've got to be careful with that, especially when it comes to SEO kinds of keyword research.
2) Building separate pages for each term or phrase doesn't make sense
It used to be the case that we built separate pages for every single term and phrase that was in there, because we wanted to have the maximum keyword targeting that we could. So it didn't matter to us that college scholarship and university scholarships were essentially people looking for exactly the same thing, just using different terminology. We would make one page for one and one page for the other. That's not the case anymore.
Today, we need to group by the same searcher intent. If two searchers are searching for two different terms or phrases but both of them have exactly the same intent, they want the same information, they're looking for the same answers, their query is going to be resolved by the same content, we want one page to serve those, and that's changed up a little bit of how we've done keyword research and how we do selection and targeting as well.
3) Build your keyword consideration and prioritization spreadsheet with the right metrics
Everybody's got an Excel version of this, because I think there's just no awesome tool out there that everyone loves yet that kind of solves this problem for us, and Excel is very, very flexible. So we go into Excel, we put in our keyword, the volume, and then a lot of times we almost stop there. We did keyword volume and then like value to the business and then we prioritize.
What are all these new columns you're showing me, Rand? Well, here I think is how sophisticated, modern SEOs that I'm seeing in the more advanced agencies, the more advanced in-house practitioners, this is what I'm seeing them add to the keyword process.
Difficulty
A lot of folks have done this, but difficulty helps us say, "Hey, this has a lot of volume, but it's going to be tremendously hard to rank."
The difficulty score that Moz uses and attempts to calculate is a weighted average of the top 10 domain authorities. It also uses page authority, so it's kind of a weighted stack out of the two. If you're seeing very, very challenging pages, very challenging domains to get in there, it's going to be super hard to rank against them. The difficulty is high. For all of these ones it's going to be high because college and university terms are just incredibly lucrative.
That difficulty can help bias you against chasing after terms and phrases for which you are very unlikely to rank for at least early on. If you feel like, "Hey, I already have a powerful domain. I can rank for everything I want. I am the thousand pound gorilla in my space," great. Go after the difficulty of your choice, but this helps prioritize.
Opportunity
This is actually very rarely used, but I think sophisticated marketers are using it extremely intelligently. Essentially what they're saying is, "Hey, if you look at a set of search results, sometimes there are two or three ads at the top instead of just the ones on the sidebar, and that's biasing some of the click-through rate curve." Sometimes there's an instant answer or a Knowledge Graph or a news box or images or video, or all these kinds of things that search results can be marked up with, that are not just the classic 10 web results. Unfortunately, if you're building a spreadsheet like this and treating every single search result like it's just 10 blue links, well you're going to lose out. You're missing the potential opportunity and the opportunity cost that comes with ads at the top or all of these kinds of features that will bias the click-through rate curve.
So what I've seen some really smart marketers do is essentially build some kind of a framework to say, "Hey, you know what? When we see that there's a top ad and an instant answer, we're saying the opportunity if I was ranking number 1 is not 10 out of 10. I don't expect to get whatever the average traffic for the number 1 position is. I expect to get something considerably less than that. Maybe something around 60% of that, because of this instant answer and these top ads." So I'm going to mark this opportunity as a 6 out of 10.
There are 2 top ads here, so I'm giving this a 7 out of 10. This has two top ads and then it has a news block below the first position. So again, I'm going to reduce that click-through rate. I think that's going down to a 6 out of 10.
You can get more and less scientific and specific with this. Click-through rate curves are imperfect by nature because we truly can't measure exactly how those things change. However, I think smart marketers can make some good assumptions from general click-through rate data, which there are several resources out there on that to build a model like this and then include it in their keyword research.
This does mean that you have to run a query for every keyword you're thinking about, but you should be doing that anyway. You want to get a good look at who's ranking in those search results and what kind of content they're building . If you're running a keyword difficulty tool, you are already getting something like that.
Business value
This is a classic one. Business value is essentially saying, "What's it worth to us if visitors come through with this search term?" You can get that from bidding through AdWords. That's the most sort of scientific, mathematically sound way to get it. Then, of course, you can also get it through your own intuition. It's better to start with your intuition than nothing if you don't already have AdWords data or you haven't started bidding, and then you can refine your sort of estimate over time as you see search visitors visit the pages that are ranking, as you potentially buy those ads, and those kinds of things.
You can get more sophisticated around this. I think a 10 point scale is just fine. You could also use a one, two, or three there, that's also fine.
Requirements or Options
Then I don't exactly know what to call this column. I can't remember the person who've showed me theirs that had it in there. I think they called it Optional Data or Additional SERPs Data, but I'm going to call it Requirements or Options. Requirements because this is essentially saying, "Hey, if I want to rank in these search results, am I seeing that the top two or three are all video? Oh, they're all video. They're all coming from YouTube. If I want to be in there, I've got to be video."
Or something like, "Hey, I'm seeing that most of the top results have been produced or updated in the last six months. Google appears to be biasing to very fresh information here." So, for example, if I were searching for "university scholarships Cambridge 2015," well, guess what? Google probably wants to bias to show results that have been either from the official page on Cambridge's website or articles from this year about getting into that university and the scholarships that are available or offered. I saw those in two of these search results, both the college and university scholarships had a significant number of the SERPs where a fresh bump appeared to be required. You can see that a lot because the date will be shown ahead of the description, and the date will be very fresh, sometime in the last six months or a year.
Prioritization
Then finally I can build my prioritization. So based on all the data I had here, I essentially said, "Hey, you know what? These are not 1 and 2. This is actually 1A and 1B, because these are the same concepts. I'm going to build a single page to target both of those keyword phrases." I think that makes good sense. Someone who is looking for college scholarships, university scholarships, same intent.
I am giving it a slight prioritization, 1A versus 1B, and the reason I do this is because I always have one keyword phrase that I'm leaning on a little more heavily. Because Google isn't perfect around this, the search results will be a little different. I want to bias to one versus the other. In this case, my title tag, since I more targeting university over college, I might say something like college and university scholarships so that university and scholarships are nicely together, near the front of the title, that kind of thing. Then 1B, 2, 3.
This is kind of the way that modern SEOs are building a more sophisticated process with better data, more inclusive data that helps them select the right kinds of keywords and prioritize to the right ones. I'm sure you guys have built some awesome stuff. The Moz community is filled with very advanced marketers, probably plenty of you who've done even more than this.
I look forward to hearing from you in the comments. I would love to chat more about this topic, and we'll see you again next week for another edition of Whiteboard Friday. Take care.
Since when Hummingbird rolled out, I tend to talk of Keyword AND Topical Research, not Keywords only.
As you say, creating hundreds of pages each one finely optimized around one specific keyword is not effective anymore. Moreover, it can even lead to problems (Panda being one of them).
Instead of thinking in "keywords", it is much better - and useful - thinking in Semantic Assets or Groups, which are keywords/queries that are semantically related one each other because of meaning and context.
The result is creating less pages, but richer ones and more able to answer to a greater set of queries.
I won't explain in details my process, because that's the topic of my talk at Mozcon within one month (one month!!), but I can offer a sort of teaser of it:
@Gianluca Fiorelli: Thanks for the tips. :)
I think you could check additionally to Google suggest synonyms with an online dictionary like thesaurus.com to see if there are similar words and check them later with the keyword planer since Google Keyword tool is omitting data as
i do that, but a better source (also for understanding the culture behind how words are used) is Wordreference.org, especially its forums
wordreference dot org is dead! Try com.
Amazing tips - for those of us who won't be at MozCon - be sure and share the slides! :)
Great Tips. It's an important part of the foundational work for a successful long term campaign. Often we just don't do enough research, and the campaign stalls part way, then we fluster attempting to bring it back to good health. Good SEO foundation steps like this equal greater results.
Thanks Gianluca for your ideas.
Hey Rand! Loved this post and implementing your suggestions. We created a spreadsheet of your whiteboard chart and wanted to share with the community. Just make a copy here: https://docs.google.com/spreadsheets/d/17xcSGXZCHPrXKkq0aSeDguAoEynNs4jDpWDafuEGWhw/edit?usp=sharing Thanks again!
Great Share Allison thanks!
Thanks for the tool! Beats me having to make one every time with Google Analytics export data.
Good Share Allison, Thanks
Thanks allison
Thank Allison! Very handy!
Thank you Allison ;)
Nice one Rand... Gave me a few things to think about...
One thing that I thought was missing from your presentation was the use of Webmaster Tools data. I believe that WMT data is the best source of key phrase data to draw from, as quite often there are a number of phrases that are sitting there on 2nd, 3rd or 4th page that show real signs of great traffic (through the impressions and clicks they are already receiving).
Agree Daniel
Its a very helpful resource for keyword research. But new website have to wait for few months or week because WMT report the data in search analytics after sometimes.
For old domain it's a very helpful to improve your keyword research process and I also suggest to revise keywords in your title and content according to the keyword terms shown in WTM for specific page. By using WMT's search analytics data and keyword suggestion by Google, I have achieved ranking in for a set of "ecommerce development" keyword group for my website www.3iinfo.co.uk
I’m happy you brought this up. For the last 4+ years, I’ve almost felt as though I don’t even do keyword research anymore (at least in the old sense of the phrase w/ AdWord KW Tool as the be all & end all). But how can an SEO not do KW research??!! is what I would think to myself. Am I missing something by looking more to related/suggested KWs, brand/consumer insights, on/offline conversations, trend analysis (w/ social listening & G. Trends), competitor data (rankings/est traffic drivers w/ SEMrush), and WMT search query reports? Apparently not. I just stopped thinking of it as KW research and more as consumer research.
And thank goodness for ‘gone are the days’ of a separate page for every KW group! I have a brand site where we’re cleaning house - a little nerve wracking since those old pages could still have some influence - but ultimately we’ve decided to do what’s best for the visitor.
Rand Fishshirt.
Hahaha. How did I miss that??!!
Thanks for this roundup on keyword research, rand.
I would like to suggest an additional column to your Sheet: "User intend"
Is the user intend navigational, transactional or informational? I use this column differently depending on the job at hand. For content marketing the kind of intend is important. If it is informational, a classic guide or howto could work. If it is more transactional a product test or product comparison could be a better way to go.
For online-shop optimization i use a score from 1 - 10 which shows how likely the user wants to buy a product. This is, in a way, the stage in the funnel he is in. For example for a keyword with a score of 10, the ranking page should probably be a product page. For a 7 it's a category. And for a 5 it's some kind of shopping-guide. For a 0 there is no intend in buying so don't bother ranking for this with a shop (e.g. "Why are bikes better than cars" would be no keyword for a car dealer, but a solid 5 for a bike dealer).
Navigational keywords can in some cases be used later on for brand campaigns.
So with this user intend in place the BIZ value is more a represantation if it matches the main range of products and if we for some reason have to push this area. Also, of course, how much money is to be made with this.
Do you use some kind of formula to calculate the Prio from the other values? Some kind of weighted average perhaps? That could be helpful for huge lists to have a column to sort by when reviewing the list.
This is a very important point. It is captured somewhat in the Biz Value column, but this explicit approach is much better.
Thanks dude, nice WBF.
I've been considering for a long time to write a blog post about keyword research...
...but I was anxious I am not doing a good job.
Now I will most definitely focus on a blog post for here on Moz.
About the WBF, I especially love the Req/Options part, and how simply you explained it (the video example).
Stay awesome,
P.S. I am writing this p.s. on every comment: Since you have video transcript, why don't you add an English subtitle, because reading while listening helps a lot to consume the information.
Hi Rand,
I agree that Google KW planner always hides some of the long tails and user specific search queries in their keyword lists. Being into SEO & AdWords, I have followed the approach that you & Gianluca outlined here. I just add another column to my spreadsheet as search queries derived from AdWords interface for the client and also the search queries from Google search console. I look into the possible permutations and combinations in those searches and build a custom target list based on all that data. This has helped me optimize the performance not only in AdWords but also in SEO. It also helped me a lot in devising content strategy based on some of the queries out there. Google Auto-Suggest is awesome and related searches give a lot of added in-depth if measured correctly and applied to keyword research. Another thing to add here is, figuring out the possible opportunities in the Google Answer box or Rich Snippets for content and optimization. This would help in optimizing for contextual searches. This is really a great landmark WBF and will give much insights for SEO's. Thanks!
I am JUST learning how to be an SEO master. Thank you for this.
Soooo, I think Moz should put together a page in Moz Analytics like this.
I know there are pages similar, but this would be super helpful to prioritize based off of easy/medium/hard opportunity for a keyword/phrase. For instance, a user would enter the following
Then Moz would find the keyword ranking and prioritize based off of the difficulty to increase ranking and the corresponding revenue impact (i.e. Keyword rank of 9 to position 6 is medium difficulty and would increase revenue by x amount). This would help prioritize for a company.
Just saying. :)
P.S. I suppose I could just put together a pivot table for something like this.
This is by far great and good for anyone who has not adapted to the evolution of keyword research. By taking the time to really dig in deep when you are starting a new project it will overall create better results and this WBF just verified that even more. Thanks to both Rand and the Moz team for keeping us informed below is the method that I utilze for keyword reserach and it matches up quite well.
My method is :
Great whiteboard as always Rand!
Question regarding the "Opportunity" tab when conducting the keyword research though.
When searching for the terms to get an idea of what opportunities we would be competing against e.g. Instant Answers etc. would you suggest using a private browsing tab? When I use a private browsing tab I tend to get allot less AdWords results than I would using a normal browsing tab...
Hi
very interesting point of have come across first of in your post and then very well experts opinion which has given me a really good understanding about the topic.
Thanks for the info Rand, I've always suspected that adwords keyword tool doesn't tell the whole truth and you've confirmed
Keyword Planner has changed what it means with "search volume" and keywords' case since August 2014, when it moved from "exact match" to "close variants":
Yes, I agree !
I'm still unclear on how to use Search Volume for long-tail KW with low volumes.
Say, for example, I have a batch of keywords with volumes of 50 or 100 or even less. What do I do with them? Should I think of them as a group, adding the volumes up as if they were one keyword (and preparing content accordingly)?
Or, should I just ignore these low volume keywords (even though I see competitors bidding on them?)
Hey Carlos
I am new here, no I wouldn't ignore them, they can be implemented into the on page content if they have the same user intent instead of building out numerous pages for what is essentially the same user intent.
The competitors bidding on those keywords may either have a great long tail PPC strategy or they have a poorly run PPC set up and are using broad match across the board.
I hope this helps you Carlos.
I usually go for the big money terms first because the little ones will follow. Plus today its about building authority anyway. It is harder today to optimize On page SEO as the number one priority. Back when EMD was king I was still beating them with Superior On page SEO.
AT Still Screen cap example https://adwebdesigns.com/college-lead-gen.html
I remember when the Google suggest was termed "the death of SEO" but it is a good tool. Google related and suggest are also great as a PPC tool especially for negative KW.
College and university terms are just incredibly lucrative. Yep that is too true I have spent a million dollars in a year and was just a tiny player in the space.
when it moved from "exact match" to "close variants" That is so annoying, just another way Google is lining its PPC pockets. Sometimes the slightest change in keywords can really hamper your success. Especially when your costs per click can be in the higher $10-$50 cpc Thanks G for making my -KW list even bigger.
Anyway for SEO the more KW tools you have the better picture of the over all keyword landscape you will have.
I always go for the big keyword money terms anyways because sometimes even when you miss your SEO first page ranking mark you can still get more traffic than hitting a number 1 spot for a KW with less volume.. make sense?
Happy Friday
Excellent fresh ideas for organized keywords research for seo and ppc. Thanks for sharing Rand!
Not to toot my own horn here, but I covered some of these keyword targeting concepts in my Hummingbird blog post in April of 2014, although I don't discuss keyword research in the post.
Anyone interested can find the post here. I've got comments off on the article to avoid having to clean up spammer links, but anyone with questions is welcome to shoot me a message via Moz.
Definitely agree on your points about more intelligent keyphrase research. Although in my opinion, click curves are still a useful dataset for organic search. Although not perfect, we can use them to segment CTR by intent, keyword type, device and size of the brand in order to create a valuable benchmark with which to compare our own data.
For me the emerging challenge will be over the loss of contextual meaning in search data. The growth of conversational search and Google’s recent announcement of location aware searches means that we no longer see what the searcher was enquiring about in many search strings. Perhaps a good subject for another whiteboard Friday?
The timing of this whiteboard Friday is awesome! Using it to help train a new employee. The question came up yesterday, and in addition to our answer, your's is more in depth. Thanks, Rand!
In addition to the basic keyword research mentioned above, we also add a column that shows how many of our competitors are listed in the top 10. If no competitors are showing, there are probably multiple ways to interpret the keyword phrase. In one of our focus industries (print), a lot of times the keyword phrase our client wants to use is also a software solution, therefore the software solution gets the nod. Or it could be a government tax form, so the .gov, Wiki, and .edu sites get higher placement. That's when you suggest longer tailed keyword phrases, so you have a higher chance of being found.
Usually I wouldn't post a product suggestion into the comment area, but this time, I believe we have a great addition to the topic itself. Rand mentioned it would be nice, if someone would aggregate all that. We are working on this and have come quite a bit. It's a free suggest scraper (it has paid optional features though) but it features a lot of additional features and metrics.
We aggregate suggests from Amazon, Bing, DuckDuckGo, eBay, Google, Google Play, Wikipedia, WolframAlpha, Qwant, Yahoo, Yandex and Youtube in more than 20 countries. But I don't want to make a big sales comment out of it now. Just check it out at https://www.sem-tool.com or get in touch at [email protected]
Especially during the research phase, going beserk on suggest and related can be a great resource. Now the problem is, that you often end up with huge amounts of keywords and need extra data to qualify the single keywords. CPC and Searchvolume can be interesting, but sometimes are not available. In this case, hints such as number of ads, distinct domains found or serp features (e.g. knowledge graph, PLAs, universal etc) can be an indicator for filtering.
By the way: suggest scraping can be a great source to find negative keywords for exclusion in your adwords campaigns.
So, I'm new... (you will be able to tell by my questions)
The whiteboard presentation is awesome... and after reading the comment from Gianluca, we have ton of keywords (including long tail phrases, etc.)
Q1: How many keywords should a site target ? 50 ? 100 ?
Q2: What is the actual mechanics of inserting those keywords into the site? Just follow the traditional "Best Web Design Practices" (Clean Urls, Unique Content, Sitemap, Internal links, etc.) ?
If the answer is Yes, per the white board (https://moz.com/blog/avoid-unrealistic-seo-expecta...), Rand clearly explains those "Best Practices" don't help ranking...
So... I am a little lost.
What do we do with the keywords once we identify them? (Other than following Best Web Design Practices)
Thanks in advance for any guidance.
Thank You Rand Fishkin for the wonderful article. Always like your WBF. As Daniel Laidler said Google Webmaster Tool search queries are really helpful for older website than google keyword planner, by getting to know what people are searching for and for which key phrases they are visiting your website. If Google Keyword Planner does hide the data, then we have to think alternate things. Well like to share what approach I use to get the keywords....
My way of approach is
Basically its not important to choose low volume keywords or high volume keyword, it depend upon how people are searching In Google Search Engine. Perfect way of finding this, is doing a lot of research on what people are looking for using Google Refined Search, Wikipedia.
Thank You
Thanks for the awesome post. To me, it seems that there are a lot of factors, you have described to take into consideration. Is it practical to go through this process doing a Keyword research for 100+ keywords? Is Keyword Effectiveness Index (KEI) no longer a metric to measure keyword effectiveness?
So yeah, when it all comes down to it, all the stats and analytics in the world can't beat good old-fashioned research with real people. If you want to know a great search term in its truest, rawest form, sit behind your Mum or Dad when they're using a search engine...and watch...
Haha.
Thanks Rand For Providing Nice information about Keyword research and this is very useful for specially for ecommerce sites because they sell many product and one category having many products.
Great article Rand,
Yet somehow an even better moustache! Thanks for both!
Q.What about having your 1A and 1B keywords as primary and secondary keywords and really splitting them?
For example: College Scholarships - University Scholarship | Site Name
Notice that although Scholarships contains Scholarship I made a point of adding both the singular and Plural versions to give the maximum spread of possibilities.
Some great Comments here from the MOZ community. You all make me feel so normal.
Tomas
You can also go to Key Word Tools like: https://ubersuggest.org/, add your key word and copie/paste the list it will show you in the Google Key Words Tool.
Give it a thumbs up if you find this helpful :)
I also like using the free "web grader" tool on hubshout.com. It tells me what keywords my competition is getting found for. As you mentioned the only real data needs to be paid for by running an adwords campaign. If you don't spend the money and time to do your initial shot gun campaign, you won't know what volume is available in the geographic area that you are interested in.
Rand, as usual great WBF! I too now include topic research in with my keyword research. I really dig the chart, some cool additions to it.
Great WBF as usual Rand, however I have a slight disagreement with the tough domains and pages thing. I have seen fresh and more entertaining content with a lot of social shares beating the best and the toughest domains in the SERPS. Recently I have worked with an organization dealing with beauty products and they did a PR Event and got a lot of coverage, so I cleverly crafted a post about this event which was about awareness and it beat the toughest ones in the SERPS within a week.
Keyword research is so important we hired a dedicated person to do nothing but keyword & competitor research. This way our research is extremely consistent across multiple clients & within verticals. We have someone who is an expert on the niches we handle most frequently. It's been a big boost (and yes, I just sent him this article. Thanks!)
Good WBF @Rand, thanks!
Regarding the creation of content around pretty similar keywords - you did a suggestion in terms of the title, but how would you use the related terms college scholarship and university scholarships within the content itself? Any pointers there to give?
Rand excellent, as always, I expected no less from you.
I love this post, I've written down all their suggestions, see if I improve the way I do the keyword research.
A greeting.
A great source of keyword ideas is in the initial correspondence we have with clients either in meetings or via email where they are trying to describe their products and services to you in a more informal way. They tend to explain thing in more basic levels of communication which transfers well online these days.
Thank you Rand for useful WBF, as per usual ; )
We'll definitely start adding your "Requirements or Options".
Agreed that Keyword Planner can only be used to determine search volume relative to other keyword ideas and to get estimate on long tail after being inputted.
Looking forward to my very first MOZCON this year!!
Sir, you are awesome.
Enlightening as ever! Thanks Rand.
Hello I am working with a customer who is ranking for only few keywords in the 100 range though he has good brand equity. I happened to check this manually since I haven't got a reliable tool. Is there a tool which tells you from which keywords is your traffic coming, What is the rank of your page for that particular keyword and your competitors as well. Please make note that GWT hasn't been set up for my customer and only analytics has been set up.
Great post Rand. I think that the problem with the new way of keyword research it's almost al the tools (even Moz) use keywords non topics to make a score.
Hey Rand
Great post, have been following you for a while now and it's the first time commenting. I love all the ideas and information provided in here and always find something useful to apply within my job role as head of e-comm. Often depending on the state of a site you take over and the level of black hat techniques used, sometimes it can be hard to rank for brand new keywords. I find using the webmaster search queries data and building meta and title data around those queries you already rank for a great tool for SEO purposes and although again the data is skewed in here, I have seen some great results from implementing this method as a starting point in various roles before a SEO site clean up has been implemented.
I wonder how many others have used this and found success.
Great points regarding adwords. I've been saying that it's not as user-friendly as it should be for some time. It's complicated and many times takes you to three different levels during your research. It's very frustrating! Then the platform makes the user select phrases along the way, and you end up with keyword topics that are hardly relevant to where you actually started your research. I also like the idea of using a chart that prioritizes the phrase, it's relevancy and the difficulty (and cost) it will take to get your keyphrase objectives.
Very useful information on keyword selection. Thanks for sharing Rand
Here's my method on Opportunity, and it really is rooted in being realistic with your data.
The first thing to know is that you can only expect a percentage of the search volume. If it's not a branded search, the most you may ever see is 40 - 50% of actual impression volume. In fact, CTR varies between 20 - 40% in my experience for #1 organic SERPs. The median is around 22 - 24%. If you add a full complement of top ads that decays further to around 15 - 18% ... really depends on the ad quality, quantity, do they show all day, etc. I've seen strong organic SERPs still pull 22 - 25% with ads above.
Ads on the right side tend to be negligible due to ad / right side blindness and the "F" style of readability.
So how do you determine this if I want to show opportunity for varying keywords? Visit this valuable tool: https://www.advancedwebranking.com/ctrstudy/
You can view with and without ads, and other valuable data sets, even within industry. Use this to define the max opportunity you can reasonable expect from a single organic SERP. If you have videos, shelf-space, or structured data you can add more, but being conservative is generally a safer bet.
So if you have 1000 potential searchers and competition from top level ads, you should forecast a maximum yield around 20 - 25% of that figure, or 200 - 250 visits, for a #1 ranking. Bare in mind, that is just that single term. The long tail should apply here as well. To figure the long tail, do a quick think on how many variations of the term are prevalent and lay that over how comprehensive your content is.
Also important, is understanding how Adwords reports this data. They are averages, not guaranteed figures! I've seen Adwords report 480 / mo. only to have 1800 / mo. show in empirical data. Another strategy I have used for keyword research is first writing the articles (for no-brainer topics that fit your product) and retroactively looking at impression data. I will then do a final pass optimization on the page.
For big, authoritative pieces (aka Skyscrapers) you can expect a larger long tail. In fact, it's nearly impossible to accurately predict the long tail for big content with high quality and excellent deployment. Spending a bunch of time trying to guess at it is not efficient. Pull potential for L1 and L2 keywords the article best covers and let the long tail weigh in empirically. Take the time you saved not playing weatherman (or weatherwoman!) and roll that into building more uniqueness.
Once you have that maximum visit data, you can apply it to your sales funnel to deduce ROI. Getting a page to rank well for "Mesothelioma Lawyer" is going to pull a higher ROI per visitor than say "Free Public Defenders". It's always a good idea to have qualitative and quantitative.
Hi Rand,
Just wanted to mention something around quantifying business value. As most of us are using Google Analytics a powerful metric to put business value on a given page/topic is to use the Page Value metric.
https://analytics.blogspot.com/2012/07/understandin...
Obviously, this is more for sites that have established pages vs starting from scratch, but we have found this metric to be invaluable in understanding how to prioritize keyword topics and is an important aspect of how we see how our current content is doing and how it could be improved.
Hi Rand.!
Very good information to complement the keyword search.
I use in my searches, the planner keywords Google Adwords, Google trends, synonyms and groups of words with similar meanings, along with related searches of google
And thanks Gianluca, very good information!
Such a great presentation this week. You covered alot and I'm glad to see that my thoughts about recent content was confirmed, I've seen this for several weeks and months. I have to say, from a local seo standpoint, I've seen evidence that reveals synonyms should still have separate pages (university and college). I've tested this over the past 6 months and have found when combining keyword synonyms into one page, rank drops from #1 to #4 on the next iteration. When changed back to splitting pages... back to #1. Overall, I give this Whiteboard Friday a 10/10. Loved it. Many changes over the last several months with instant answers, videos, splitting the SERP and organic rankings. Lots to consider.
Checking the landscape of the SERP is a must out side of SEO too. In paid search, observing that shopping comparison dominates the landscape could be the difference between success and failure.
Great WBF and totally agree! It's great to find ways to make our jobs easier and to help out clients with realistic expectations.
I have been using a tool that does a lot of the things you were discussing looking at difficulty and opportunity, but this pulls in information from your Google Analytics and Webmaster tools so you get search volume and not provided keywords also. For anyone interested there is a post here which explains more about it.
Hello Rand,
I was wondering if the keywords case (Broad, Exact or Phrase) in Google Adwords can influence the results in a better way. I mean choosing to exclude some situations in which the keywords don't have any connection with the search.
Anyway, great Whiteboard presentation!
Have a wonderful day,
Mirela.
Hi Rand,
Super WBF, I've been fighting on this topic for long time and I think it's time for everyone to consider all points you mentioned if you want to bid/create content for the right keywords. Even in Paid search we are seeing how AdWords suggestions are becoming too commercial and not comprehensive for niche search intents.
In the past few months I particularly worked on the "opportunity" considering even with Google Scholar search results in some instances. So far that process has been pretty manual, would you be able to recommend me any tool that could give an idea of the opportunity at scale?
Thanks for sharing Rand
We all know the numbers shown in Adwords are not exact and thus still we believe the data. These are some brilliant ideas that everyone should focus. But, I am confused that is there any role of Google Trends in making the choice of keywords?
Good video Rand. Interesting, I was doing some keyword research these past few days and now i feel I have to go back and redo some. Thank so much man.
When tools are taken away it is time to not be so dependent on them and start doing your research Rand...it has pushed all of us who want to be in the search results harder...you give us sound ideas...many thanks.
Thanks for the updates! Always looking forward to whiteboard Friday.
Hi Rand,
Great WBF and wonderful to see this being addressed as it amazes me how many people I see focusing too much on the keyword research of old!
With this initial research just looking at the keywords/topics, it would also be great to have a follow-up post linking this with content creation. When I do keyword research I tend to also make a note of what kind of content could be made, whether it would be an inforaphic, video, evergreen article, post. This helps me personally to prioritise the topics to create content for; yes there may be higher search volume for topic A, however if topic B has a lower search volume but the topic itself is potentially more shareable and more suited to promotion (whether through outreach or paid), I would choose topic B every time.
Love this. I think Business Opportunity has been one of the most overlooked aspects of keyword research and great that you bring that up!
Another place I like to look is SEMRush and look at what keywords are bringing top competitors the most traffic. A great way to spot some opportunities you wouldn't have considered or maybe even found in the GKP.
apart from forums and other online discussions, I usually also look at the competitors from my clients. it happens a lot that they might be using different wording for the same products or services, and it would provide me with more ideas!
thank you Rand for the WBF!
Competitor keyword is definitely a good tool, and it doesn't hurt that most clients like to see this in their strategy. It is important that we use these competitor keywords as way to explore other keywords and not the end all be all. I find that comeptitor keywords are a great place to start because they lead to so many keyword rabbit holes.
Great information on keyword research very helpful for future ! keep it up
Hi Rand,
As always,thanks for sharing great information.
One additional tool that I find very useful is Übersuggest. It helps with intent by loading the short tails and seeing how people build the extended tail. Working your way up, by adding the second level extended tails gives further clarity. It really is a case of wash, rinse and repeat.
I also find that it helps clarify the the nuances by country. For example, what works in the US may not work here in Australia. The AdWords tool that is offered up in Australia shows a lot of results that are clearly from the UK. I doubt this is a problem in the US but it is probably not a unique problem to Australia, that we see the results that are inclusive of results from other countries.
So, remember to always pick your own country when using Übersuggest.
Great post, Rand, but it's important to remember that some of us were doing keyword research like this in 2010. :)
In fact, here's a deck I presented at SMX West in early 2009, talking about focusing on user intent and business value while doing keyword research:
https://www.slideshare.net/brysonmeunier/meunier-br...
It's definitely more effective now that Google has cleaned the SERPs up, but it's problematic in any case to assume everyone was focusing more on search engines than users with keyword research five years ago. Some of us saw the futility of that event then.
BTW, great to finally meet you at SMX Advanced, and to present on the same panel. Definitely enjoyed your presentation and hope our paths cross again.
Good stuff. I like to add a second layer of detail to a keyword analysis project, and that is to break down the serps of each KW in more detail.
The things I look for are: which websites are the thought-leaders on each serp? E.g., is it Wikipedia, a whitepaper, Forbes? And also, which competitors show up in on the first page for each KW. The clients love to see this, although it tends to piss them off at first lol.
i think most of us who using adwords tool must be suspecting about data is correct or not? i would like to introduce one tool here to get more idea about keywords and than use plugin your liked keywords in to adwords, using https://ubersuggest.org/ tool you can have more idea of keywords.
Very Useful Content for me
Thank q
https://technicaleduhub.blogspot.in/
I agree with Allison. Her comment was very interesting!
Great WBF Rand!
Thank you very much it's the best of article.