|Would you rely on wordtracker?|
is it accurate enough for the UK
I have a subscription to wordtracker, and by and large I am happy with the results it gives me.
However, I refuse to believe some of the appallingly odd results - which I presume can be manipulated by certain people performing very odd and unique searches in dogpile and the like (what is dogpile - i never use itm why should wordtracker care less what dogpile says).
At the same time, wordtracker predicted an emerging niche in an industry which I am familiar with, and now you can sit back and watch hundreds of media producers fill that niche. So the KEI system might actually work.
I got my advice on wordtracker from an American, can anyone offer a British perspective please?
I don't use Wordtracker but this it's produced by a UK company so I would assume they know the British market?
my name's Mal Darwen, and I work for Wordtracker Customer Service. I hope the following goes some way to answering your question.
We are still building the UK database and the searches figure tells you how many times that has appeared in our own sampling. At the moment we recommend that people use the UK searches comparatively - i.e. to decide the relative importance of each keyword.
The UK database after filtering out questionable or robotic searches currently holds over 111 million searches giving us a total of 34 million UK keywords. We add new searches and keywords to this every day. We collect UK data from a number of different types of sources and one of the challenges we face is first combing these different types of data and second, removing duplicate searches so that we give as accurate account as we can.
We will continue to update the UK database on a daily basis and will announce major improvements as they happen. In the meantime, thank you for taking the time to raise this.
[edited by: IanTurner at 8:33 pm (utc) on April 15, 2009]
[edit reason] Removed email request - WordTracker aren't that hard to find if you look. [/edit]
|total of 34 million UK keywords |
Are there that many words in the English language? ;)
34 million keywords is alot of keywords :), haven't used wordtracker but looks interesting so will give it a try
I strongly recommend trying word tracker, and I will continue with my monthly subscription for the time being.
I do wonder how Wordtracker can reliably estimate the search traffic in Google, Yahoo and MSN, from searches in Dogpile. I would have thought that a completely different type of person uses Dogpile, than uses Google, or Yahoo, or MSN.
I suppose that the only way to get more reliable data, is to secure an executive position with one of the major search engines?
Good point. Mald?
I mainly perform SEO for UK websites, and I haven't found wordtracker to be anywhere near accurate enough for general-purpose usage. I might use it for verification that a keyword has the kind of popularity indicated elsewhere, but (considerably?) scaling up samples means great potential for data that can't be relied upon.
If you want very reliable data, test keywords via PPC campaigns. You can do this on a fairly limited budget, and you can also start to measure the effectiveness of keywords - in addition to just popularity. Otherwise, I would try to combine data from more than one source (e.g. Google/Microsoft) and base decisions on that.
I have heard it mentioned that Ad Words allows one to test the reliability of keywords.
Having never set up an Ad Words programme yet, could someone describe what sort of basic information I would get for setting up an Ad Words campaign? (I dont mean to promote one PPC method over an other, but I would be using Ad Words for PPC)
Also, would it be more worthwhile to spend my Wordtracker budget on a PPC campaign - because not only will I get more reliable data, but I will also get adverts displayed online that might driver business to my websites.
Thank you in advance,
I don't think Adwords is too hard to work with, but I've been using it on and off since it launched, so that's probably easy for me to say ;) There was a good discussion on this subject a little while ago: Is Google AdWords Too Complicated For The Average Advertiser? [webmasterworld.com].
Whether you need to pay for wordtracker (or any other paid keyword research tool) depends on how good you need the data to be, and what exactly you use it for. I find most paid tools to be sold on features, rather than the quality of the database. Examples would be finding synonyms and related words, working out competition, ease of exporting or saving results. If you get hold of "rawer" data, then you may have to do some legwork yourself.
Personally, I prefer using relatively simple programming and a copy of Excel for most keyword tasks. That way, I can control both the data I put in, and the output I get back.
I think it is far too easy to find serious errors within most keyword databases (including Google's tools) - primarily keywords with drastically overstated popularity, or words that it is impossible regular searchers are actually typing with any frequency. So if you want reliability, then you will still need to process the data somehow.
Lastly, IMO it depends on what degree of reliability you need as to what's the right tool for the job. If I was starting a business based on a keyword, I'd be running PPC listings for those keywords beforehand to get (and measure) real traffic. If I was writing a quick article, I might be happy scanning free keyword popularity results. Most tasks are probably somewhere in between ;)
Good points. I've heard it said that you can get access to reliable data for as little as £10 with an Ad Sense account.
Is it simply a matter of opening the account and getting access to data, or do I only get access to the data for the keywords that I am bidding on....until my money runs out?
You have to spend money to get the clicks (which can add up fast in competitive areas), but essentially you can pay for real world data (and you get the visitors to play with). You can restrict to search traffic form Google.com which will give you as good an idea as possible what numbers are out there. There are still associated problems (you need to maintain a highly visible listing, PPC traffic is not necessarily the same as paid traffic, even for the same keyword). But, if you need that degree of reliability and your target is Google, I can't think of any better sources.
I like this discussion, there are some good points coming up. With Wordtracker's data, its' worth remembering that it's a representation of search volume rather than a global result, and what you see coming up in the results are real searches made by real people - we go to great lengths to filter out spam and robotic searches. There is also no extrapolation on our search counts (which you will find with some other tools :-) ).
Adwords can be a good way to test keywords, especially in a geographic context, but it's good to make sure you've entered negative keywords (which you can find with Wordtracker)to make sure you're not spending too much on irrelevant clicks.
[edited by: engine at 7:16 pm (utc) on April 24, 2009]
[edit reason] See WebmasterWorld TOS [/edit]
If I'm testing keyword popularity with Adwords, I would use exact match keywords (within [square brackets]). That way there is no need for negative words. Broad match is less efficient for pure testing, but is useful for finding search patterns and variations (as long as your analytics can separate the keywords you bid on from the words searchers actually typed in).
|There is also no extrapolation on [wordtracker's] search counts |
Interesting. To me that suggests there is a particular audience for Wordtracker's data providers - I just don't find the output to match what actually happens at a search engine like Google, for instance.