Article Summary and Keywords

TIME SAVINGS IDEA: What if we automated a way that you’d no longer have to create an article summary or enter 3-7 keywords with each article submission?

With our scale, we’ve been thinking a lot lately about how to make the article submission process even easier than it is and one solution we came up with last night was to remove 2 transactions from every article submission that could be accurately handled with some automation behind-the-scenes.

It may not seem like much… removing 2 things to do with each submissions, but at our scale, that would remove a couple MILLION human transactions annually… freeing up our members to do 2 less transactions with each submission and thus simplify the process a bit…giving you more time to write & submit even more articles.

Here’s how it’d work:

Article Summary Automation: The article summary would be automatically grafted from the first paragraph. If the first paragraph was only a single sentence, then we’d automatically create a 4-7 sentence paragraph from the following sentences until the summary was enough of a teaser to be valuable. Most article writers already load up the first 3-7 sentences with enticing overviews about what they’ll get with the rest of the article.

Keyword Automation: Using our own analysis of statistically improbably keywords and keyphrases in your article title, article body and potentially even your resource box; we could take a strong uncannily accurate prediction as to which keywords best describe your article’s intent. 3-9 months later, we could use the actual keywords that are used to find your article by our users to further tweak and refine the best keywords to aid all sites who still use our meta keywords tags to help them understand what the page is about.

Your thoughts? Good idea? Would this be a helpful enhancement to your member experience?


tyler writes:

Great idea — just make it editable. :)

Comment provided May 15, 2009 at 8:00 AM


Carl Pruitt writes:

The keyword adjustment sounds particularly awesome. The automatic summary could be a great feature as well so long as the option is available to override it and create a summary manually.

Comment provided May 15, 2009 at 8:01 AM


Darlene Norris writes:

Personally, I spend a lot of time researching keywords to find ones that are low competition, while getting a decent number of searches. It’s a nice idea, but I wouldn’t use this service. Same for the article summary. I prefer to write my own. Thanks anyway!

Comment provided May 15, 2009 at 8:01 AM



Great idea on the summary, saves time.

Even better idea on the keywords! Will there still be the option to add keywords manually? Also, can you increase the space so more keywords can be entered? As long they are relevant there shouldn’t be a problem with that? That would be great to get more long-tail keywords in there.

Thank you

Comment provided May 15, 2009 at 8:13 AM




It’s widely believed after 7-10 keywords, that you’re wasting your time as the search engines are ignoring them.

Heck, I’m not even certain meta-keywords matter any more… but we’ll continue to include them.


Do you really think the keywords you enter for the keyword field are really that important?

It’s my gut feeling it’s almost irrelevant to the quality of the keywords used in the article title and article body (assuming of course the article is high quality, unique, exclusive rights and doesn’t create bounce backs when referring sites send their visitors to it).


What specifically do you perceive you’d lose if you no longer could have any say in the article summary field?

Comment provided May 15, 2009 at 8:23 AM


terryweber writes:

I like your Summary and Keyword solution to the problem. Saves time for me and for you and results would be better then me trying to do it, especially like your idea about automatic keywords. I alway struggle to find the best keywords. In short, doing it automatically is far less labor intensive for all of us.

Comment provided May 15, 2009 at 8:23 AM


Meggin McIntosh writes:

I’m intrigued…and it’s true…it doesn’t seem like this would save us a bunch of time, but it WOULD!

I really like the keyword adjustment feature….

THANKS! Drinking out of my EzineArticles cup right now! Drinking coffee, that is!

Comment provided May 15, 2009 at 8:24 AM


Dick Pennington writes:

Great idea!

Comment provided May 15, 2009 at 8:45 AM


Ramon Greenwood writes:

Great idea!!!! I hope you will do it. The sooner the better.

I continue to be amazed by your service.

Many thanks…Ramon Greenwood

Comment provided May 15, 2009 at 8:51 AM


Deb Younkin writes:

I think it is a great idea. That’s the thing that always takes me the most time on submission. I do my keyword research before I write the article. So if the keywords are in the article, they will be in the summary and on the keyword list, right? I think this is an awesome idea!

Comment provided May 15, 2009 at 8:54 AM



This would be a great addition to the already awesome services you provide. I just hope you’re ready for all the extra articles that will be submitted!

Comment provided May 15, 2009 at 8:54 AM



I’m in.

Comment provided May 15, 2009 at 9:21 AM



Keyword automation sounds great. But summary automation should be subject to writers’ approval. First para summary may not be the universal rule.

Article writing automation should benefit all, I suppose! Just feed the keywords and summary and…

Comment provided May 15, 2009 at 9:21 AM



That seems a great idea. Should speed things up nicely.

Comment provided May 15, 2009 at 9:24 AM


Mel Menzies writes:

Love the idea, Chris, but – like others – I’d like the final say. For instance, longtail keywords are often phrases and there never seems to be quite enough room in your little box thingy to put them in.

Still – that said, I rarely find that the keywords I use are identical to those which your system picks up later as the ones being used. If you see what I mean? And that would be really clever, being able to use the actual keywords 3-9 months after publication.

Thanks for all your ideas. Haven’t been doing much lately because I’m preparing talks and workshops for a major conference where I’m to be keynote speaker next month, but I love writing for EzineArticles. Mel

Comment provided May 15, 2009 at 9:59 AM




If the keywords are in the article, that’s what we’d base the META KEYWORD data on.

The Article Summary is important for mainly RSS feeds, email alerts, & widgets; yet I’m not convinced that most members can make a statistical difference if they write a custom article summary vs. one that we’d build for them based on their first 5-7 sentences.

The current majority of article writers do not write custom article summaries. Instead, they either fail to do the job all together or they copy & paste their first paragraph into the article summary field.

But, I do get why someone would want to get the right to include whatever custom article summary they choose…and that’s why we didn’t just make the change…wanted some time to think about it, get member feedback, contemplate both viewpoints.


We’re ready. Bring it on! :)

Comment provided May 15, 2009 at 10:17 AM


Chuck Bonner writes:

I’m reminded of the recent Twitter flap over removing a little-used option. The majority don’t care, some don’t care until you take it away, but the minority who do use it can be _rabid_. (Yes, color me rabid.)

Add my voice to those saying, “okay, but give me the option to override the automatic summary and keyword list.”

But the automatic adjustment of keywords after 3-9 months? That sounds great, and I don’t think I’d care to override after 3-9 months.

Comment provided May 15, 2009 at 10:35 AM


Gail writes:

I love it! My summary is usually my first paragraph anyway. And I think your analysis of keywords would be right on target. BTW, I am loving my mug! Makes me want to write more. Thank you EzineArticles.

Comment provided May 15, 2009 at 10:52 AM


Sherry Harris writes:

I think it is a fabulous idea! Can’t wait for it to be a reality. Thanks for the continuous research and development.

Comment provided May 15, 2009 at 12:25 PM


Brian writes:

Great idea captain. Make it so!

Comment provided May 15, 2009 at 1:30 PM


Cherry writes:

I go with those who like the idea of you being able to provide and automated summary and keyword list as long as it is offered for us to consider and allows us to override it if we want to.

But adjusting keywords after 3-9 months to track the words that really have driven traffic – that’d be great!

Thanks for always looking for ways to improve your service yet more!

Your Career Change Guide

Comment provided May 15, 2009 at 1:35 PM




Some of my search engine friends have said my idea of optimizing the meta-keywords 6-9 months after the fact may be a dumb idea… almost like why mess up something that is working; and upon thinking further on it, I’d have to agree.

Our big challenge with this proposed change is to make it EASIER to submit articles. The process needs to become lighter, easier, faster, and with less stress..less repetitive motions so that our members can focus their core time on writing more high quality unique articles.

What I want our members to do… is to give us arguments where they are certain that either their article summary writing or their manual keywords selection …can statistically outperform our ability to automate this data based on their own existing data.

It’s all about trade offs. No one likes having options removed from their control; yet we must also balance off the greater good concept of does the improved simplicity cause article submissions to rise because it’s easier or not?

Yes, of course we could AJAX in to make it optional, but that’s exponentially more effort and still does add to the complexity of the submission process …something we’re focused on making easier.

Comment provided May 15, 2009 at 2:08 PM


Rajeev Batra writes:

Nice idea but howmuch time does it takes to prepare a summary and write key words. It should be at the tip of the tounge of the writer who has put his head,heart and soul in his writing.

Comment provided May 15, 2009 at 2:11 PM




You’d be surprised how many people struggle at both of these!

What comes easy for you and many of our members, does not come easy for everyone else… so is life. :)

Comment provided May 15, 2009 at 2:16 PM


Karrie Ross writes:

As always, here you are thinking about us… and believe me we, I, appreciate it. I like the idea that the summary is automatic and if you do plan to auto-ize the keyword…that this is a spot on code or whatever it takes to make them appear in the best possible locations.

Keep up the good work.
Karrie Ross
Book Cover Design, Web-design, Branding and Consulting

Comment provided May 15, 2009 at 2:31 PM



Add my vote to those who love the idea, but want the ability to override it.

Sometimes I do want a summary that’s different than my opening paragraph, though the keywords really aren’t that big a deal to me.

Comment provided May 15, 2009 at 2:59 PM


Angela Booth writes:

Sounds interesting, but put me with the folk who say “Thanks, but…”

I’ve seen the results when Google et al decide to create their own page description from a page’s content, and sometimes the results are woeful.

Here’s the thing: I look on the meta description as my article’s own little advertisement in the search engine results pages. Nine times out of ten I use the first paragraph as the description. The tenth time I don’t — I have a reason to create a description that I want to see in the SERPs.

If you make this change optional, that’s fine, but PLEASE DO NOT just change it because it’s “easier” for people who want to submit articles.

Why on earth should it be “easier” when an article can stay online for years to come, and send you traffic for years? It’s surely worth a few extra minutes on the keywords and page description.

The majority of your authors may not care, but serious article marketers (as I am) do care.

As an analogy I look on it as the difference between those who want an automatic transmission versus those who choose manual gears on the cars they drive — I’m in the manual camp. I always have been and always will be.

If you make this change mandatory, in my opinion you’re devaluing the service you provide. I for one want MORE control over what happens in the SERPs (search engine results pages), rather than less.

Comment provided May 15, 2009 at 3:53 PM


Lovie writes:

Hey Chris, Lovie here and I a extremely new to the article submission process. I currently have 3 published articles. I say that to say it was an easy process to load the articles into your submission site. I just copied and pasted the article from Word. When I wrote the article already knew my keywords (research) and used the first paragraph as summary.

The bottom line is SEO. Whatever method is good SEO let’s use that one.


Comment provided May 15, 2009 at 4:10 PM


Gisele writes:

Really in on the keyword automation. My sense is that it’ll probably be done better than what I do myself.

Not so sure about he article summary.

Comment provided May 15, 2009 at 6:59 PM



Hi Chris

Re keywords: I think it’s not only for the search engines but mainly for your own internal search system.

How about introducing tag pages?



Comment provided May 15, 2009 at 11:06 PM


Lance Winslow writes:

I like the idea of saving time.

When you state: “Most article writers already load up the first 3-7 sentences with enticing overviews about what they’ll get with the rest of the article.”

What is Most? % wise? Are we talking 95%? If so 5% might not like that? So, would you have a check off box for automated load, or would you just go with the 95%. BTW- I don’t care either way on the “Teaser” or article summary box.

About the key-words, this might upset some people, but it would not really bother me much. Would you retro-actively change all the past key words of all articles ever submitted or would this be pro-active?

One issue I’ve done lately is purposely not used keywords that I may have accidentally used to much, because I do not want to trip that “keyword” stuffing rule and get my article rejected by accident. Many of your subtle rules, you have so many now, that authors are artificially modifying their behavior and could be costing you money in click ad revenue, plus, hurting themselves, which could make them discouraged as to the results. Which means they might stop posting, post fewer articles or just not be happy. Some might voice thier opinions most will not it’s the 1 person per 20 “unhappy campers rule” so this is a serious issue.

Too you had better be ready for the paradigm shift from Google’s new “drill down” options for websurfers with all the new options that come down to the left of their searches. You will be finding older articles pulling again based on those searching 6-months ago, or 1-year ago option.

In which case you may very well wish to redo all your key words. Since I have over 16,000 articles, you may, if you wish use only my articles for a test case scenario, letting the computer choose all the key words, see what happens. If this works, then test a larger batch of articles and authors in several categories. Now I assume you have already tried this and also it stands to common sense that it would work to increase traffic for EzineArticles and The Authors too. So, this could be the biggest win-win ever concieved here.

Too, you might boost all the articles ever posted’s placements in the search engine. This could get interesting. Those are my prilimary thoughts, but I have not studied it and have little real data to base my hunches and opinions, but I thought I might share these observations anyway.

Comment provided May 16, 2009 at 12:35 AM


Sara writes:

This is 1001% gonna helps a lot! Make it happen!

Comment provided May 16, 2009 at 2:48 AM


Deb Kukal writes:


*is this a step toward the site becoming less interactive/individually responsive in the way you do business? That feels uneasy…

*have you done any sample studies to see how well your keyword selection actually works? Do your keywords generate as many hits as the author’s keywords? Are they as high quality hits, ie, as focused on the target audience?

*if the answer to #2 is yes, then in reality #1 doesn’t really matter!

*re summary: If we know that going in to the process, then it will be fine–we will make allowances and write with that in mind. I’m ok with it.

Comment provided May 16, 2009 at 5:42 AM


Gene Stagg writes:

I kind of like the idea but would like it to be optional. I have to agree with Angela Booth, where she says, “I’ve seen the results when Google et al decide to create their own page description from a page’s content, and sometimes the results are woeful”.

The thing that strikes me the most is the tone of your responses. They all seem to have the underlying tone that says to us, “We’re going to do what we want whether you like it or not. The only reason we even mentioned it is to make you feel like you have a say”. It really comes across sounding like you’re talking down to anyone who doesn’t like the idea.

Just my 2 cents.

Comment provided May 16, 2009 at 7:40 AM



Hi Chris,

I’m a newbie/2 article person so far, but admire and respect the way you handle your site – and your writers, and do appreciate the service you provide.

I add my thoughts into the box with several others; I appreciate the assistance offered by this new plan (good idea), but want to retain the final right-of-edit of my own work.

Love your site, Chris!


Comment provided May 16, 2009 at 2:17 PM


Kerry Kanemitsu writes:

This is a great idea – both the summary and the keywords. But – keep them editable because someone may have some ideas on how they want the article searched.

Your keywords and summary could tell them whether they’re on the right track or not and that would be great! Maybe they’ll even change their minds when they see what’s been pulled for them.


Comment provided May 16, 2009 at 2:41 PM


Steve Kimball writes:

I would greatly appreciate these time saving steps

Comment provided May 16, 2009 at 4:28 PM



//ALL: Challenging assumptions here…please don’t take my thoughts as personal//


Wanting to keep the keywords editable assumes that you think your keywords can out perform our ability to identify the highest value keywords in your existing article, title and resource box.


I’m hearing a lot of “Don’t take something I might need away from me” instead of “Here’s specifically why we human experts can outperform your computers ability to identify our keywords and article summary.


The tone isn’t meant to talk down,…my tone here is a challenging one though… Looking for someone to make an argument that is not based on “fear of loss” but rather logical reasoning as to why my suggestion of simplifying the submission process by eliminating 2 extra steps that can be better performed by our software – is a bad idea.


Uhm, we’ve never been highly individually interactive on purpose. We’re an operationally efficient machine with a purpose to be fast, accurate, and a strong attempt on consistency. If we were highly membership-intimacy-based, our ability to process the hundreds of thousands of transactions we already do monthly would be impossible without 10x’ing our team size.

Of course we’ll continue to be interactive. :) If we weren’t, we’d have just made this change without consulting anyone.

As far as being ‘responsive’… I’m the CEO of the company and still commenting and blogging on a Saturday afternoon.

We also doubled our member support team this year to increase our response speed; even though I admit we’re currently not delivering on our member email response targets but have a plan to correct that.

Did we do any tests on this yet? Nope. How can we? The only way that it would be a real test is to take away the options from everyone for a period that we’d deem statistically valid…perhaps 9 months to a year and then evaluate whether we’re attracting more or less traffic per article per day per month per year.


Good questions… I don’t think we’d retro-actively change all of the past meta-keywords, even though we’d have the ability to.

Search engines don’t give much weight to meta-keywords these days (my perception)… as there are too many other dozens of important variables that they must evaluate to rank each article/page properly.

If I thought that meta-keywords was a mission critical piece of data that would help our members get their article the best traffic attraction possible; I’d probably have invested the server and software resources to optimize it by now… I just don’t have any proof it’s really important… thus why upsetting a minority or ~5% (just a guess) of our members who are hyper-tuners (I’m in that category I suppose of wanting to play “advanced thinker”) for the greater good of shaving off 2 transactions PER article submission might aid the other 95% who may not even notice those fields are missing.

If we make it easier for 95% of our members to submit more articles with less hassle, less work… we’ve got to weigh that against the 5% that will most likely be upset (currently only because of “fear of loss” rather than proof it’s a bad decision).

Yeah, those drill down options will most likely have an impact similar to a year or two ago when Google started delivering personalized SERPS (ie: each result is customized to the user so that it’s rare for 2 identical searches to return the same ranking of result.).

I do agree that OLDER articles will most likely benefit from it… a benefit to those who have been here before newbies.

Thanks for letting us use your articles as a test. I’m not sure we could get a conclusion back fast enough as my current gut feel is that it’d take 9-12 months before we’d know if there was any result… not to mention how difficult it would be to isolate the variables that do impact traffic.

You bring an important win-win point to the forefront: That we’re selfishly motivated to maximize and optimize traffic per article…just as much as our members are. Statistically we are in a better data-driven-decision making position to figure this out because we have access to millions of datapoints to consider across all articles and years of traffic.


Not convinced that tagging will add value despite the obvious individual position that it would of course add-value to the users. I fear content pollution indexing… anything that makes it difficult for search engines to figure out what and where the unique articles are vs. other ways of displaying data may not be in our mutual self-interest.

Each year that goes by does make stronger cases for tagging that can be done without hurting search.


Ok, but we’re not talking about the META-DESCRIPTION field here… We’re talking about the META-KEYWORD field that currently is fed by the keywords given during article submission.

I like manual transmission better than automatic myself! …I have a 7 speed manual/automatic SMG transmission where I have a clutchless manual 7 gears along with 5-6 shift points or I can have the car automatically shift for me. What I really want is a CLUTCH so that I can choose when the car gets the gear vs. when it doesn’t…but my car’s maker thinks that they know better than I do about shift points that make sense in terms of engine RPM + most likely a dozen other data inputs to determine when I can or can’t engage different gears.

Your argument is that you want a manual with a clutch [instead of the clutchless manual that my car has] because you believe you can optimize better than our software can.

The reason you say you want this clutch or control is so that you can help aid your articles chances in the SERPS…but how do you know that you’re statistically adding value vs. what we could do with software?

The standoff here is your gut opinion vs. my gut opinion… because neither of us have done a statistically valid study on this yet; yet I’d hope you’d give me the benefit of the data-drive-decision doubt because of the history of data we have access to.

Said another way: We’re looking to grow incoming article submission volume by making it easier for everyone…and if it were easy to let you have the ability to edit the data that our software could soon recommend/use, we wouldn’t be debating this topic.

Of course we don’t want to make your life miserable or remove your access to choice…

For now, fear nothing as we’ll be working on mock ups of how this could work and will present our ideas in this blog for further comment.

Future commentors on this entry:

If you’re in the camp of not wanting to lose control of an option you believe is *critically* important… Please provide arguments for how & why you think you can do better than our software for these 2 fields of data in question?

Comment provided May 16, 2009 at 4:57 PM


June writes:

Personslly, I think this is an awesome idea, and I’m all for it. This will simplify submission and make it quicker too. I know that there are many that struggle with trying to figure out the right keywords to use, so this would be much easier for them. For the others that would rather have more control over their article and keyword election, having the ability to edit the keywords will solve that problem too.

It is human nature to resist change, even when change is for the better. I think that once the change has been made, and folks get use to it, they will appreciate how it has saved them time and simplified their lives if only for one small moment.

That was my 2 cents.

Comment provided May 16, 2009 at 5:25 PM


Frank A Hilario writes:

Don’t bother. You can simplify things by requiring each article submitted to be of enough tease in the very first paragraph to invite further exploring. That will take some doing for the authors, but it will DECREASE your workload and INCREASE the articles’ pulling power.

Comment provided May 16, 2009 at 7:14 PM


Lance Winslow writes:


Okay, I hear what you are saying, so, let’s go ahead and allow your Artificial Intelligent software to choose all the key words for my 16,000 articles. The sooner we start the test the sooner we’ll have relevant data to answer the question. Plus, it took me a good 5,000 articles before I became proficient at key-wording. Since, I have articles in “every category” at EzineArticles, I think this would help with a good over all test.

I see no downside for me personally, only and upside. And for all the authors here, I see an upside if it works, which I believe it will. Plus, it will provide perhaps a boost to EzineArticles of a fraction of a percent even if I get a boost of 20-50% in traffic.

I do believe that a software program, like the ones that visually display the most used words in a graphic, seem to be quite good at choosing key words and the over all topic of a piece of writing. As I understand it, those articles with 500 or more words do best with the AI software, but I also believe that 250 word minimum word articles also will get a boost over their human counterpart picks.

Additionally, being a fan of Alan Turings work and all the AI specialists at the top of the food chain today, it seems, I do trust the software in this case. No its not foolproof, but, I believe it is better than I could do.

Further, what if the AI program upgrades, continued to go thru the entire data base as key words changed, user choices changed and Internet Surfer searches evolve, which they sure do. I am game for you to do this project test with my articles, and subsequent on-going test, as time goes on. I’d even like to write the research paper based on our findings, that is if you’d want to share such valuable data with your competition, as whatever we find will be an intellectual asset of great measure. Think on this.

Comment provided May 16, 2009 at 8:05 PM


Lance Winslow writes:


Do you have the current criteria that Key Wording software is using? For instance; the title words other than “the,of,it,is,etc” would not be counted but nouns would be counted “X’s Three” while words in the last paragraph “X’s Two” and words in the summary, which would also be within the first paragraph would count as “X’s two-point-five” and the category that the article was placed would receive “X’a one-point-five” and of course all words in the work would count as one-point each.

You see, I wonder if we could design a BETTER set of rules than anyone else has devised, that would be perfect for this venue; online articles of 250 words to 1,000 words. Much better than other people have done the prior, but also taking into consideration the search engines’ guestimated ranking standards. It is my contention that all the software programs out there might not be perfect for this venue, but a lot of very smart people working with larger data sets have indeed come up with some EXCELLENT criteria.

I’d like to see ALL my articles floating with keywords that change as the world around us changes. Imagine, the increase in income to EzineArticles and the increase in traffic to the authors. NO ONE else would be able to compete with such leveraged technology. And there is nothing wrong with blowing by the competition when you are better than they are.

Comment provided May 16, 2009 at 8:32 PM


Jeff Herring writes:


First of all I appreciate two things:

1. The EzineArticles Team’s constant focus on improvement and making things better

2. You taking the length of time you did above to respond to so many on a Saturday afternoon

Having said that, here is my 2 cents worth…

It seems like the issue has come down to

Automation vs. Control

My question is why does it have to be either/or?

While it may be more difficult to set up, why not have the Description and Key Words set up like the Resource Box, which you have set up to be automated, but can be altered.

I realize they are not exactly the same, but it would be great if the idea could work. Course there may be some technical reasons why it would not work. That’s where you guys come in, cuz I just know what buttons to push to make money with this stuff…

In the final analysis, I think for anyone that submits lots of articles, just about anything that can make the process faster is a good thing.


Comment provided May 16, 2009 at 8:57 PM


Lance Winslow writes:

Jeff makes a good point. Because many authors know or admit that they are not up on the keywording strategies much, and/or are new to the scene and are not sure what all to make of it just yet. They would really appreciate the automated concepts as it saves them time and they’d most likely be better off.

An advanced user might like the manual, and might actually sometimes beat the AI software, or might have specific keywords they are trying to attack in the “long-tail” theory, so they definitely want to do it manually. These are the authors who probably wish to do it themselves and have the option.

Still, there are authors like me for instance that would most likely if given the choice choose the automated software most all the time, but every once in a while with a super speciality niche article, I might like to have the opportunity to do a little keyword tweaking.

Comment provided May 16, 2009 at 9:46 PM



My suggestion is, go ahead with the automatic mode as a default, but leave the classic mode as an option.

Like all good video camcorders, they come with the option to switch between auto mode and manual mode. Like auto-focus and manual focus. I use them both for different situations.

So have both.

Comment provided May 16, 2009 at 10:52 PM


Alwahsh writes:

Great idea!

Comment provided May 17, 2009 at 4:21 AM




I saved your blog comment to discuss with our developers this week. I’ll report back as soon as I have an update and more answers to your other questions.


I’m proposing this issue as a software instead of dual-control (instead of either/or) because it’s SIGNIFICANTLY more difficult/labor intensive and server resource intensive for us to deliver choice instead of handling it all behind the scenes.

If giving BOTH options were an easy choice, we wouldn’t even be discussing this because everyone likes choice (duh. :).

What I’m excited about today: Asking our developers to deliver some more reports on actual keywords used in relation to what they should be if the title, body and resource box were analyzed as one unit of data.

Comment provided May 17, 2009 at 12:31 PM


Linda McCarrin writes:

When I quiet my mind, there is “space” and time for other thoughts…my natural intuition to kick in.

Comment provided May 17, 2009 at 1:33 PM


Ian Patterson writes:

How can you add keywords after the article is completed? Keywords are an integral part of the whole article; they are not guesswork added on at the end.

Like Darlene, I research my keywords for profitability not for popularity. There is a difference, a big one. Keywords must be used with correct frequency and placement therefore it is essential to know the keyword that will be used. In addition, we use vertical keywords, lateral keywords and longtails.

You mention ‚¬“after 7-10 keywords‚¬! Only one keyword should ever be used per topic and be relevant to the topic, that is what makes is so important to know the keyword upfront.

You mention, ‚¬“I’m not even certain‚¬ which suggests that a decision was made without full understanding. I am not against innovation, it is what makes the world go round, but it does need research.

You wrote, ‚¬“Did we do any tests on this yet? Nope. How can we?‚¬
This question should have been asked: ‚¬“We want to automate and the only way that we can see to do it after eight hours of brainstorming is to take away the options from everyone for a period that we’d deem statistically valid. Does anyone have another or an improvement on this suggestion?‚¬

I do.

Q. Some people don’t care, others do, so how do we please all of them?
A. Give them the option to opt in or out at the time of submission for a period.
Q. How do we benchmark?
A. Use statistics against previously written articles by the same writer; For instance, take a given writers readership before H hour and the readership after H hour. Is there an improvement or not? If there is an improvement or if responses remain the same, then initiate the change but if there is a decline, then revisit the task.

I like the idea of ‚¬“being able to use the actual keywords 3-9 months after publication‚¬ (I think that came from Mel)

Changing keywords to those that actually found the article is neat. If you can do that, go for it. It would be nice for individual writers to be able to view that type of statistic and use it to write fresh content on the same topic.

Comment provided May 17, 2009 at 3:13 PM




Clarification: We’re NOT talking about keywords or keyphrases in your article title, body or resource box.

We’re talking about the KEYWORD field in the article submission process where you enter in a few keywords that we then automatically already insert into the META-KEYWORDS file that is near the header of each page.

Our proposal is to automate these 3-7 keywords based on your most dense keywords and key phrases found in your article title, body and resource box.

Essentially, we’d only be spitting out what you’ve already included in your article… but with a high degree of precision.

Keep in mind… I already think these META-keywords are being ignored as being only marginally of value (if any)…as other factors are more important today (vs 10+ years ago).

Your benchmark idea: It would take at least 9-12 months (primarily due to the fact that old articles don’t get re-indexed as often as fresh new articles…and without the re-evaluation of old articles, the point of the study is lost) and we’re not willing to wait that long to find out if this idea improves the speed of the submission process.

As for the changing of the keywords to the actual keywords being used to find the article: That idea has been explored and is now currently off the table and won’t be considered. It was my idea originally and I’ve been corrected offline that it was a dumb idea.

As for being able to write fresh content based on the words and phrases being used RIGHT NOW to find your articles, that EXISTS for FREE for everyone. It’s called our “Traffic Search Terms” feature. Basic thru Platinum members get access to the top 5 keywords being used and Premium members get access to the top 30 keywords/keyphrases used to find your articles. It takes about 3-6 months before this feature becomes strong in value after a new article is published…but it’s one of the best features we rolled out in the last year.

Comment provided May 17, 2009 at 3:57 PM


Ian Patterson writes:

I understand what you want to do but I, like many others, have reservations.

The keyword used in the meta-tag should match the page name, page title, page description, page heading and page content.

My interpretation of your wording “your most dense keywords and key phrases found in your article title, body and resource box” is the keyword used most often. That is not what should happen. There is a specific frequency with which to use the keywords and I think the wrong keywords will be picked up.

The statistics in this example were taken today.

I write an article on Silk Flowers therefore I research for a good keyword to use. My research shows the most PROFITABLE keywords NOT the most used.

Keyword Value Demand Real Supply Profitability
silk flowers for a bride 537 2 268.500
silk cemetery flowers 231 1 231.000
silk flowers calla lilies 220 1 220.000

That shows that I should write my article based on ONE of these keywords. Of the 537 searches for ‚¬“silk flowers for a bride‚¬, only two sites are using that particular keyword.

My concern is that your automation will take ‚¬“silk flowers‚¬ as the keyword. Look at the difference in how many other sites I am competing against.

Keyword Value Demand Real Supply Profitability
silk flowers 27387 24216 1.131

If ‚¬“silk‚¬ or ‚¬“flowers‚¬ became the keyword, we would not stand a chance. This is the reason we pick a niche and need to control the keywords.

You mention ‚¬“the speed of the submission process‚¬; is this more important than the returns?
Some writers may want to get as many articles written in a day as they can, and probably use software to do it, well that’s ok for them, but some of us make a serious income from the results of our keywords.

I think, to be fair, there should be an option at submission to use it or not. It should be the writers choice.

Btw, ‚¬“It was my idea originally and I’ve been corrected offline that it was a dumb idea‚¬. No idea is ever dumb, it just needs looking at from a different angle.

Comment provided May 18, 2009 at 7:34 AM




Your assumption is that the META-KEYWORD field is valuable data that has impact on rankings.

I’m not convinced it’s as valuable as you think. It was a decade ago, but that value has degraded significantly (and rightfully so).

BTW, “Silk flowers for a bride” is not a keyword. That’s a key phrase. If you inserted “Silk flowers for a bride” as a keyword in the keyword field, it’s my gut feeling that your keywords would be ignored all together.

Again, my assumption is that the META-KEYWORD field is not as important as it once was.

Next up… we’re doing analysis this week of what members are actually doing with the keyword field and I’ll report back once we have more data to analyze and draw some more conclusions.

Comment provided May 18, 2009 at 8:04 AM


Ian Patterson writes:

I apologize for the format of the table.
this is how it reads.

Keyword Value = silk flowers for a bride
Demand = 537 searchers using that specific keyword
Real Supply = 2 websites using that specific keyword
Profitability = 268.500 A ratio 268.5:1 in website/writers favor


Keyword Value = silk flowers
Demand = 27387 searchers using that specific keyword
Real Supply = 24216 websites using that specific keyword
Profitability = 1.131 A ratio 1.1:1 in searchers favor

Comment provided May 18, 2009 at 8:12 AM


David Quimby writes:

If we did not have to write are articles around key words that bit or bit not be used anymore by the search engines the article bit flow better for the reantder witch may lead to more articles being read and more traffic to your sites. Most authors write above their readers head trying to come up with keywords that can be pick out of the article if it is well wrote, one that flows good, not trying to make a few keywords pop out at you.

A Summary is only a small desciption of what a good article is about to lead a reader to want to read more, witch being automated would allow for the best summary being pick out of a good article. It may lead to more time to try and promote their product.

Comment provided May 18, 2009 at 11:31 AM



Preliminary study shows that 4 keywords is the magic number based on what the greater majority of our members have done over the years:

# of Keywords % of total articles
1 11.2%
2 9.1%
3 15.1%
4 17.7%
5 15.1%
6 11.2%
7 7.3%
8 4.7%
9 3.0%
10 2.0%
11 or more 3.6%

My gut feeling is that 5-7 should be the number of keywords included with each article as it doesn’t hurt to up the keyword assistance…and yet it’s not high enough were it looks like you’re a keyword stuffer (something we’d never allow or advocate).

Comment provided May 18, 2009 at 2:51 PM


Deb Kukal writes:

That is really helpful–thank you.

Comment provided May 19, 2009 at 5:01 AM


Roseanna writes:

That sounds like a terrific idea to me – especially one who is not an SEO expert. You know how google thinks better than I do.

Comment provided May 19, 2009 at 3:17 PM


Shirley Bass writes:

Actually, I pick and choose my teaser from my article sentences. I didn’t realize we could actually summarize it without using the exact words from the article itself.

I, also, didn’t know how you used our list of keywords.

But I’m very particular about them.

I agree that the title, body and the resource box is enough for the search engines to pick up what they will…

And I can’t see that EA’s ideas would hurt anything in my article marketing plan.

Although, I prefer having the choice of changing the summery if I so desire.

You know me, I LUV EzineArticles, because you have always taken me down the path of refined success, credibility and trust.

I’m trusting you now to do what’s best for your authors, as well as EzineArticles.

Comment provided May 21, 2009 at 8:34 AM




Today we pulled a random sample of 20,000 articles to see a side-by-side comparison of what the author selected for the keywords vs. what our software detected was the best keywords.

Our next step is to analyze this data and make some assumptions as to what the patterns are when we believe the author chose more wisely vs. when our software made a better decision.

Appears a good handful of members think that the keyword field is where you write a single sentence description of your article or repeat the article title. That would be “wrong” because that’s what the Meta-Description tag is for.

One area our software clearly excelled was being comprehensive enough to list all 5-7 high-value keywords as many members gave up after entering only 1, 2 or 3 to 4 keywords.

Our next step is to figure out the wide discrepancies that exist in a small percentage of the sample where the member chose a high value keyword that they then went on to not use very much in the article body. ie: It didn’t show up in our list of most used keywords or keyphrases.

Lastly, some of the matches (small percentage)were nearly identical… That suggests the member was already using some kind of keyword density software tool or they edited their article after-the-fact and used our own data from the EzinePublisher view (where we currently display keyword data publicly on every article).

Comment provided May 26, 2009 at 11:06 PM


Lance Winslow writes:

Good progress on this, it is of interest and it could be very interesting to see the results.

Comment provided May 26, 2009 at 11:18 PM


Roseanna writes:

I agree with Lance. The results will be extremely interesting and beneficial to us all. Thankyou.

Comment provided May 27, 2009 at 7:31 AM




Three months of research, testing, development and re-development (a few trial runs proved fruitless)…and:

The keyword field automation tool is now live!

I’m not convinced this will have any impact on the traffic your article will receive but it WILL speed up your article submission process.

Comment provided August 17, 2009 at 10:29 AM


Gail writes:

Chris, thanks for always trying to make things better and easier for us. The keyword automation did speed things up but I was glad I could edit it. In some cases I only got 2 words and there were several more that I added.

Comment provided August 17, 2009 at 10:57 AM


Roseanna writes:

I’m finding it really useful thankyou and easier for submissions. I’m not good at writing for keywords; just tend to write what I believe in, so its very helpful for you to effectively tell me what my key words are!
Thankyou for your help.

Comment provided August 17, 2009 at 12:20 PM


Manuela writes:

It is not my first time to visit this web site, i am visiting this web site dailly and get good information from here everyday.

Comment provided January 21, 2014 at 2:37 AM


RSS feed for comments on this post.

Leave a comment

Please read our comment policy before commenting.