New Wordtracker keyword tool is coming 20 February 2009

The new Wordtracker tool is on its way and as well as giving you a preview of our thinking, this blog will allow you to have an important say as we debate some difficult questions.

The new tool will allow you to look at keywords through three important metrics - frequency, competitiveness and commerciality. We believe these three metrics give you the most useful picture of the potential of a keyword, and allow you to uncover profitable keyword niches for your PPC and SEO campaigns.

Snapshot of the New Wordtracker Keyword Tool

The new tool will allow you to look at keywords through three important metrics - frequency, competitiveness and commerciality. We believe these three metrics give you the most useful picture of the potential of a keyword, and allow you to uncover profitable keyword niches for your PPC and SEO campaigns.

We've spent many months responding to the comments you've sent us, researching and analyzing the SEO & PPC workflows that you use and running usability tests so we can create the most user friendly and useful keyword research tool on the planet.

Our new tool will fit easily with the work you do, and add to your experience as you find what people are really searching for online.

We've given access to the new tool to roughly one third of our customer base and will be contacting everyone else in the next few weeks.

But of course, since we launched Wordtracker 10 years ago, many things on the web have changed and this throws up many questions:

  • What metrics do you really need?
  • Do you need prediction of the daily traffic a keyword may attract or do you just want to know the frequencies in our own database?
  • How useful and accurate is the keyword effectiveness index (KEI)?
  • What alternative ways to rank keywords would you like to see?
  • How can we best report the competitiveness of a keyword?
  • How do you uncover commercial intent?

We'd like you to be part of this discussion through this blog.

We're very much looking forward to the year ahead and have a number of milestones planned in 2009. Expect to see:

More keywords

Starting with our 365 day metacrawler dataset which will now include single searches. Did you know that adding single searches increases the number of unique keywords from 21,658,982 to 197,374,808 unique keywords? That's 9 times as many.

A better measurement of competition

We will be moving away from Google exact match website count towards the use of allinanchor. Serious search engine optimizers need to examine incoming links to a given page for a search term. We're going to tell you how many incoming links appear per keyword for the internet as a whole!

Online commercial intent

In a landmark paper on OCI, Microsoft describes a query as commercial "...if the user submits the query because he/she intends to purchase / sell something (product, house, car, etc) or commit some commercial service (such as stock trade, auction, dining, mortgage, insurance, business consultancy, real estate, loan, restaurant, lottery, retirement planning, or an obvious paid service etc)."

Please subscribe to the RSS feed to keep up to date with our changes as we're looking forward to bringing you into the discussion.

Over the next few days and weeks, you'll be hearing from both myself & Andy (the founders of Wordtracker), Ken McGaffin (our chief marketing officer) and our in-house SEO experts.

-Mike Mindel

73 comments

  1. Looking good Mike.. look forward to checking out the new features. Look forward to trying out the commercial intent tool...Regards what would we like,,,Some kind of "query analysis" tool would be great.

    1. User 1A searches for "Digital Camera"
    2. User 1A refines search for " Best Digital Camera"
    3. User 1A finds what they require and searches for "Canon Model No 123" etc etc

    can this data ever be obtained? The Search Journey of a customer?

  2. Will you please make a wordpress plugin

  3. Dunno about commercial intent - ex being info marketers. rank for montreal and sell people the best tourist guide to montreal, ever.

  4. "We will be moving away from Google exact match website count towards the use of allinanchor."

    For competition results we really need allintitle, at least as an option.

    Thanks, Linda Buquet

  5. The biggest improvement I would like to see is accuracy in data. Or the relationship between keyword numeric value. I know you will never be able to tell me how many hits I will get from a keyword on Google, but I would hope you can get somewhere closer than you are now.

    Using Google Keyword Tool, their numbers are light years away from what you project as keyword volume. If you took a list a keyword from Google Tool and compared that with the traffic volume from your tool. The two dont look "relatively" the same at all. In the WT case you would have over half the keywords as getting zero clicks where in the Google example you could see 100% projected clicks for all the terms.

    They might as well be looking at two completely different data sets.

  6. If at all possible, provide a result display of the amount of back links to the domain/page from the the top ten results for a keyword, that gives you a good idea of the competition.

  7. I agree with Jaan Kanellis. It is important that results tally between Google and Wordtracker and for that matter any other word tracking tool. If they don't match then at least one of them is wrong - but which? Customers need to be confident that the results are a true picture of what people are actually searching for. No one will subscribe to unreliable data. I would also like to see extra emphasis on UK data.

  8. From your point of view this is great. It will improve your competitiveness. From my point of view, I already use 2 other tools that have been around for quit some time and provide this data. They beat you to the punch by a couple of years. WT is one of several tools I use to evaluate data and ideas.I currently use some aspects of Word Tracker as an idea generator, not as a real research tool.

    Show me... Bill

  9. I agree with Linda B. The allintitle for a better evaluation of potential keyword competition would be a big plus. I've never used WTs KEI as it's based on No. of web pages indexed for a given keyword/search vol. Perfectly good keywords could be discarded as too competitive if relying on this measure - I always have to manually evaluate using allintitle which is a pain.

    Jaan also makes a good point regarding data. Most clients I work with are only concerned with Google traffic. As such, is metacrawler data really as relevant and representative? I understand the issues with data skewing on Google search estimates but isn't there a way to take this into account to provide an estimate? Beyond me but right up your street!

  10. Lipstick on a pig? Why not step out of the Wordtracker box and aggregate your numbers to work with Google and MSN using APIs? Give us sortable column listings using all the matrix's. Yeah, sure your profit margins will lessen, but the tool would become the absolute most powerful combo in the world.

  11. I agree with Jaan,

    With such a sophisticated keyword tool as Wordtracker I hate trying to explain to clients what the 'COUNT" column really means.

    For example the keyword 'dog training'.

    In Wordtracker the Count is 1029 In Google the Feb Search Volume us 135,000

    So a client thinks there are 1029 searches for 'dog training' when looking through Wordtracker (over a period of 160 days to boot!) What is more skewed here Google or Wordtracker?

    Am I missing something???

    Here is the description Wordtracker gives to their data:

    This shows the number of times a particular keyword has appeared in our database.

    E.g. Our database currently holds 269,015,367 words. A count of 147 tells us that this particular word has appeared 147 times in (this is over 160 days).

    Our keywords are taken from major metacrawlers (a service that queries all the main search engines simultaneously).

    Our main sources are Metacrawler and Dogpile, the two largest Metacrawlers on the net. Metacrawlers have the major advantage of matching the search profile of the search engines very closely. But are not subject to the same kind of skew from software robots that continually check web site and pay per bid positions.

  12. As above - it's puzzling when comparing Google Keywords with your own -exactly as if you 2 are looking at different populations...different planets.

    I'd like you to focus on real markets rather than fuss about work flow or Lord save us, kei.

    The sooner you drop that nonsense the better

  13. Looks good, can't wait to try out the new tool!

  14. What we expect from new tool is accuracy of data. We have been using word tracker and many other tools such as Google Adwords etc. But still can not see any tool giving accurate data about Searches Per day for a phrase. We don't expect 100% accuracy or even 90%. If anything close to 70%-80% can be achieved that would be great.

  15. Can you address the issue of database size? Yours is now 197 million keywords (is that searches?) and Keyword Discovery claims theirs is 36 billion.

    Are these apples & oranges? Does their larger database make their results more accurate than WordTracker's?

  16. IO second the request for competitiveness measure based on Google AllInTitle operator. We're adding that info manually now.

  17. Google's Keyword Tool will NEVER give you accurate numbers. Did you ever notice that Google's numbers are always the same? 49,000 or 12,100 or 680, etc? They will never tell you the right numbers. I think WT has always done a great job of giving more realistic numbers without a ulterior motive of selling ad space.

  18. If its a good as your other products then it will be something special. Traktori

  19. How is the OCI index calculated? Is it based on real data (how?) or a semantic score for the keyword?

  20. I do a measure I call "toughness of competition" which includes the PageRank of the competition sites in the top 10 spots. Often, but just using KEI, I get discouraged by low numbers, when, if you consider the expected PageRank of the site I'm working on, I really shouldn't worry, because even though there's a lot of competition for those words (thereby lowering the KEI), that competition is soft (sporting low PRs).

    Is this what you mean by "commerciality"?

  21. Agree, would love to see allintitle alongside and average PPC for 1-3. Add domain name availability and you can name your own price for membership... thanks, Allen

  22. I would like to see more local capabilities to limit the results to a specific region of a country. Results from other parts of the world or other parts of the country are not so relevant to my clients who are marketing locally.

  23. It would be good to be able to see seasonal results. So if you are researching something to do with Halloween, you can go back to October/November the previous year. At the moment the results for the month we are in are not relative to past or future search numbers for seasonal topics.

  24. Good stuff, yet still not what I am waiting for from Wordtracker:

    1. Keyword source and database size. Metacrawlers provide 1-2% of total searched keyword volume, and the people who use them are not exactly my audience (far from it, actually). I am far more interested in a type of database that KD have, for example.

    2. Keyword Search volume trends. These are very important, and there is only one service out there that shows them in presentable graphs for analysis purposes. Google Trends is a good example, but it lacks the search numbers and the other search engines.

    3. Country/region/city filters. Again, Google Trends is a good example.

    When these 3 happen, I will be a loyal Wordtracker customer again.

  25. Sound great. How soon is soon ?

  26. Nine times the keywords?! Great. WordTracker's ignorance of the fine end of the long tail bugged me. Will I be able to ditch disclaimers to clients looking for regional niche geotargeting of SEO campaign keywords?

  27. Here's a simple fix we need.

    Allow the following type of search

    boarding houses hotels -bed and breakfast

    Allow us to do a search and to eliminate keyphrases from the search before it's done, rather than using the SELECT + DELETE method afterwards.

  28. Sounds good but please, don't get so 'enhancement' focused it starts having more bells and whistles than non 'internet marketer experts' can use it.

    That was one of the reasons I stopped using Nichebot, it got more complicated to use than was needed.

  29. Thanks for the tool... I look forward to using!

  30. While many subscribers, including me, are using your data as a competitive advantage for SEO, there is an even bigger market in the research area. This is another area where we make money. I like all the secret competitive SEO stuff, but am just as interested in comparing wordtracker results to Google results... given that your data comes from metacrawlers it just is not representative of the true masses searching on the web. Is there a better way to translate, compare, and contrast your data to publically available Google data? If so, this would help me a lot.

  31. Sounds Great

  32. would like to know more

  33. These improvements sound great, but what I'd like to see more than anything is for Wordtracker to pull from a much wider pool of data that's more representative of the searching public. Most of our clients are niche B2B companies, and Wordtracker in the past has told us search volume is 0 when it's not.

  34. I'm new but want to start at the top! Seems to me I need competitor info, who is linking to them, what keywords they are buying. And then how do I keep others from doing the same to my sites?

  35. I think it's a great idea looking forward to it coming out.

  36. Is there anyway you guys could incorporate a geographical search like for different countries. As an example if i wanted to see how many searches are being conducted in say Australia for a specific keyword.

    I have also found knowing the Commercial Intent on a keyword is very useful in making sales it is true when they say keywords are not created equal.

  37. Jaan I couldn't agree more.

    How does one calculate the approximate (real) search volume for a keyword phrase in Wordtracker?

    I love Wordtracker for many reasons but this is the one thing that is difficult to understand why it is not explained better.

    I been a member for two years and still have to use multiple keyword tools to supply clients with potential search volume data for individual keywords.

    I believe it was (leaving name out) who said "you can multiply the 'Count' column by 5 or 10 times to get a closer picture of the monthly search volume"... but when comparing the same keyword with Google that doesn't match up well either.

    best buy coupon 268 best buy coupon 22,200

    Maybe he meant 100 times.

  38. Not sure why the value of the KEI is being questioned. It has been a very big tool for me when using Wordtracker. Is there a reason to suspect that it doesn't "mean" what the material about it suggested it meant?

  39. To answer some of the questions in the OP:

    I definitely need the daily traffic prediction, broken down by SE. I want to know whether it's worth bothering with going after a phrase, or if that big-seeming figure is going to end up being 10 hits/day. And I do NOT want to have to do the math to arrive at the broken-down figure. That's what YOU'RE for!

    The last few times I used WT, I have had misgivings about its accuracy. I think you're underreporting how many searches words get. This didn't used to be a problem, but lately it not only is one, it's glaring.

    As Jaan mentioned, your reports are way lower than Google's. And my own sales stats suggest that the truth is somewhere in the middle--but more toward Google's guess. If your reports were right, I'd be getting all of one sale a month for words which I know can bring in a sale a DAY! So even though I've used WordTracker (paid) for years, lately I've been using Google's tool more often. Since their tool is FREE, you are going to be toast if you can't make your accuracy better than theirs.

    As for the KEI...that's got to be the most overrated thing on WordTracker. Isn't worth a darn! Why? Because it counts all the times a keyphrase shows up in an engine as "competition" - even though 90% of those mentions may be from sites like personal blogs mentioning the phrase in passing. Sites like that are not competition, they are just "there," and any kind of serious SEO will jump right over them.

    I want to know how many REAL competing sites there are. Not how many times "Jane" happened to mention "I got a new Mickey Mouse toothbrush today, even though I'm 50 yrs old, tee hee!" in her LiveJournal update. She's not even trying to rank under "Mickey Mouse toothbrush" and is probably at Listing #990 for it, so her site shouldn't be counted as any kind of competition. But it is. So I always have to go check Google and Yahoo myself, to see if there's ANY real competition there. Sometimes your site will show a miniscule KEI for a phrase, and when I check, I find it'd actually be hard to NOT rank in the top 10 for those words.

    On the opposite end, sometimes there'll be a high KEI for a word. "Only 1000 competitors" I think! Then I check, and those competitors are huge rank-powerhouse sites like Microsoft, Intel, AMD, etc, for the first 100 listings. The KEI for that word should be ZERO but since there aren't many listings, it gets rated highly by you.

    A GOOD THING is the fact that you're finally going to include single searches in the results!!! I am sick of having to just guess whether certain very-long-tail keyphrases are worth going after.

    On the other hand, I have been glad that some keyphrases are not well-known to the SEO community...so maybe it's not so good after all! Someone may actually find those 40 searches/month keyphrases I have been enjoying :-/

    Oh, and don't forget, when it comes to the interface: KEEP IT SIMPLE.

    Overall WordTracker has been a good tool. Of course this post has my complaints, because, what am I going to do, list every other feature and say "don't mess with that part, it's okay?"

    But that fairly recent accuracy issue is something you need to fix, NOW. The rest of it doesn't matter if WordTracker keeps telling me that actually-good keyphrases/words aren't worth bothering with.

  40. Hi there

    I need free keywords,would you help me, please give me some idea about that.

  41. there is still now support for outside usa/uk, so this tool is still limited in its users

  42. Maybe I missed out, but where is the Evaluate-option? This is a GREAT tool to determine the competition for a keyword within the 3 biggest search engines.

  43. I would like to see the relationship between keyword numeric value.

  44. This one's for Pabitra, who is looking for a free keyword research tool. I'd suggest Good Keywords, at http://www.goodkeywords.com/#othertools

  45. Having read most of the comments, I agree with those people questioning the volumes and numbers.

    We have 30 + clients across all fields, sizes and budgets, with the volumes we see, across their websites/adverts being much closer to the Google numbers than the current WT ones.

    We like and use WT to look at cluster words but currently that's about it.

    For all of us it's about competitiveness. We're in this to get our website or our clients to the top of Google (or the search engine of choice) for phrases that will generate income.

    Considering solely the UK, US and European markets for a moment, knowing the competition for a phrase in relation to Title and Page Rank numbers would be an ideal place to start our research from.

    Looking forward to seeing the new version though...

  46. As an ex-client of yours I would like to get access to your new tool.

  47. I bought 25 domain names based on the values I saw on Google Adwords keyword tool. It reported more than 200000 searches for each of those keywords and I was shocked to see those domains available. I purchased them all in a snap! BUT What! I got zero traffic! Then I checked on wordtracker just to see the actual search counts were near to zero.

    This was a big lesson for me: Always double check your keywords with different tools, when possible.

  48. Is it a replacment of wordtraker or it would be a upation of the older one..I am using SERP analytics one -http://www.serpanalytics.com/tools/top_keywords

    and i need one more with this...

    Thanks for the post anyways..

  49. I have to agree with Yuri, especially about the region-specific results filters. I need Canadian results for some of my clients and WT does not seem to even include them, let alone allow me to focus on just Canadian content! WT just doesn't seem to be meeting my needs right now..... hopefully soon?

  50. In order of importance:

    1. keep estimated daily volume
    2. improve count ACCURACY (as many have stated above)
    3. 12 months of data and ability to see seasonality
    4. assistance with geographic search
    5. ability to search for this + that - the other thing

    There is NO relationship between the counts I see in Google Adwords, Wordtracker, Keyword Discovery and Google analytics. None. That makes it near to impossible to set, and then meet, client expectations.

  51. Thanks everyone for your comments - some replies follow, but please do keep them coming.

    @Jaan Kanellis, it's worth remembering that Google doesn't say what its sample size is for Adwords tool, or what their extrapolation algorithm is. They are different datasets, but most SEOs will look at more than one tool to get the information they require, and draw their conclusions from them

    @ John Mills - we're looking at ways to grow and improve the dataset.

    @ Bill Urell - with the new tool, you can generate ideas with the Lateral Search, and then research them without leaving the page. It's good to use other tools, and we wouldn't advocate otherwise, but it's probably worth having a closer look at how Wordtracker can speed up your workflow

    @ Andy Dawson - The new tool will provide an allinanchor metric to start with but our data crunching has given us allintitle and allinurl as well. Depending on demand, we could provide these in the near future.

    The metacrawler data has served us well over the last decade, but additional sources of keyword data do give more information. We're testing new data sources now and when we're certain of their quality, we'll be incorporating them in the new tool.

    @SEO - its' not that the data is necessarily skewed in either set, but the dataset sizes vary. Wordtracker can tell you the size of the dataset, but Google doesn't release this information.

    @ Alex - Wordtracker is going to be introducing new metrics which will supercede KEI - these will be much better indicators of keyword use and commercial value.

    @ Bill Treolar - These are apples and oranges - 'searches' is the number of individual search queries that we analyze to create our database of unique keywords. So from a sample of over 1 billion searches, we create our database of 197 million unique keywords. Our sampling methodology has been mathematically validated.

    @ Alp Bozkurt - for more information about OCI, it's best to refer to this document: Algorithmic Commercial Intent Detection Analysis

    @ More Than Members - nice ideas, thank you

    @ Don - we're looking for data sources from other countries, and when these are available to us, we'll add them. Geographic search is also something we'd like to be able to provide in the future.

    @ Georjina - we aim to present the tools and the dataset in the clearest cleanest way possible. It would be easy to add widgets and gadgets, but you're right, it's helpful to concentrate on the task in hand rather than shiny distractions.

    @ Nancy - KEI when it was introduced 10 years ago provided an excellent way to rank keywords and decide on which to concentrate. Many of our clients still find it very useful. However, we believe that the new metrics we will be introducing will be even more useful. We'll be addressing this in a longer post shortly.

    @ Good Bulbs - you raise some good points here. As far as accuracy is concerned, Wordtracker is very careful to not over-report searches, and filters for spammy behaviour to produce the cleanest dataset we can.

    @ Pabitra - for free keywords, you can use three of our free tools:

    http://freekeywords.wordtracker.com http://freekeywords.wordtracker.com/gtrends http://labs.wordtracker.com/keyword-questions/

    @ Article2Video - an excellent point, thank you.

    @ Nick - the new interface is essentially the same tool, but with better usability, more efficient workflow and the ability to present other metrics.

  52. New SEO Tools...

    ...

  53. Hello Mike, I think one of the metrics you may be missing is relevance. It may have to be a pre-defined set based on single, major keywords or somehow built into the SE alg. I think you would be somewhat remiss by excluding it as a metric, as I use optimization software that does, and it returns color coded KEI which is extremely handy, cuts to the meat of the best keywords/phrases and allows me to optimize pages fast.

  54. Will this tool be free?

  55. I have learned SEO from the great bloggers such like you. I thank you always for such great post and that comes from the sole of my heart. Just as you always help i would alywas help you by telling my secrets. I use a free keyword tool from serp analytics and it can be found out here http://www.serpanalytics.com/tools/top_keywords..it is a good and not such like Google and yahoo which are showing some broad results..I have been using this from past one year and thought to share it through my comments.

    :)

  56. @ Greek - The new tool will be available to Wordtracker subscribers - we do have free tools that you can find at:

    http://freekeywords.wordtracker.com/ http://freekeywords.wordtracker.com/gtrends/ and http://labs.wordtracker.com/keyword-questions/

  57. I am really looking forward to the new features. It's too bad it sounds like it will only be available to paid subscribers. I hope it resolves issues with accuracy. I was always a little disappointed in the accuracy of low volume keywords. Here are some other free keyword tools that may be worth checking out.

  58. My problem with Wordtracker is:

    1. There are tons of other keyword finding tools, many free, some embedded into SEO software, which is handy.
    2. Many have "proprietary algorithms"
    3. Google's Adwords tool is pretty comprehensive, it is from Google, but it is a time sucker.
    4. I have been using the free version of WT and been crashing and burning.
  59. And I have to focus on Local Search. Wordtracker will not return any results for any local terms, because they are not popular enough, or who knows why? Doesn't that leave keyword search for the purposes of local search impact in the dust?

  60. I hope this new keyword tool can address the issue of local search, have some geo-targeting component, track IP's, zips or something.

  61. Hi, www.wordtracker.com - da best. Keep it going!

  62. Thanks for article. Everytime like to read you. Thank you Robor

  63. The tool is great!But its usability is limited mostly to English keywords. I wonder whether Wordtracker in this new extended version will cover also other languages than English.

    Marek

  64. Thanks for article. Everytime like to read you. Thank you Joker

  65. Keyword counts have got to tally with google, otherwise what use is wordtracker now. It's got to be easy for complete for novices to use

  66. Hi!

    Provide a Rich Site Summary

    : by using external links

    If it doesn't, who cares, for you will have Google-Proofed your site against any sudden drop in your site's PR rankings and broadened the reach of your site by using the high PR of other sites on the Web.

    Content will always be king -- make sure it reigns supreme on your site.

    I recommend goowall.com advertising service to everyone who like to have a successful business on the Internet

  67. Hello

    Google PageRank is one of the very main factors which decide the ranking of Google search results

    If you don't have the Google Bar, you can check any site's PageRank here: http://www.GooWall.com.

    There is one more strategy you must explore if you want to lessen the power and importance of your own site's PR.

    You can also try PPC (Pay Per Click) advertising to provide you with targeted traffic and sidestep the search engines. Well sort of!

    goowall.com knowledge and expertise was very helpful in making our dreams happen on the internet

  68. When are the new search metrics being released at last!

    Thanks Jase

  69. Hi Jason,

    we're still working on the new metrics and the best way to organise the information we have - they'll be coming in the next few months. Bear with us, and watch this space!

    Mal

  70. Google PageRank is one of the very main factors which decide the ranking of Google search results? I don't think so...it's been 2 years since TrustRank is born, and now it's time for the semantic web...

    Google is not about PageRank anymore!

  71. Google's Keyword Tool will NEVER give you accurate numbers. Did you ever notice that Google's numbers are always the same? 49,000 or 12,100 or 680, etc? They will never tell you the right numbers. I think WT has always done a great job of giving more realistic numbers without a ulterior motive of selling ad space.

  72. Hi there, This seems like a great useful new tool, I cant wait to try it..!

    I do have a VERY important question about it though - Will it support UTF-8 encoding, and Hebrew? Will it support searches from google.co.il for example as well as google.com?

    So far Ive had some troubles with your tools regarding the UTF-8 part, and it's really important to me since most of my job is based on it.

  73. @BN7Websites - Wordtracker doesn't currently support Hebrew searches, and as we have data from the UK and US, we support Google searches from .com and .co.uk - I'm afraid we can't at the moment show searches from .co.il.

    Sorry we can't be more help with this at the moment, but please do mail us if you have any other questions (support@wordtracker.com)

    Mal

Post a comment