Ok, so you set up an Ad Group with 2 ads for split testing and over X amount of weeks you see the clicks come through for the keyword 'Widgets' yet it's simply not converting. - How many clicks would you give it, before turning it off? - Would you try it with other, different ads first, before considering turning it off? Any other advice would be very much appreciated.
i consider low or no conversions a landing page problem, not a creative problem. Low or no clicks is a creative problem
But surely, not all keywords convert for everyone. I run a business and the aim of my website is to recruit people - a conversion is when they request further information on the business. Not all keywords will bring in the right sort of people, which is what I'm trying to find out.
honestly it really depends on what your ROI is. Some campaigns consider a keyword a failure after 100 clicks others consider it after 1000. It depends on how many visitors it takes to make a conversion and what the ROI associated to it is.
Yeah that's true. My ROI isn't concrete as I earn indirectly from people who come into my business - it's not simply a case of each conversion is worth £x.xx however I do pass some leads as part of an Ad Pool so in that scenario there is a ROI in a way, as I'm 'selling' leads on for a set price. So basically, if a lead is worth £2.50 and each click is costing 25p, anymore than 10 clicks without conversion would be classed as a failing keyword?
not really, if its getting conversions then its successfull, you just need to optimize and get the costs lower and the conversion higher
That's my point - some of my keywords aren't converting. Hence why I'm wondering how long to give them before just giving up on them. Some have had over £10s worth and still not yet to get a conversion, but something at the back of my mind thinks, 'what if I suddenly get 5 conversion from that keyword in a short space of time?' - 'what if it's the ad thats too inviting and so bringing in impulse clickers?'
Generally I allow 100-200 clicks just to get data. I usually throttle my CPC way up at first to get first position and thusly a wicked high CTR. Then adwords will naturally start to lower your costs. After about a week of high CTR and me gradually lowering my CPC while position preferencing the #1 spot, I then let lose and position preference 1-3. CTR will stay high, CPC will drop drastically. You can get 10-20 cent clicks where you were previously getting $1+. In this case you can spend 5-10 times more per conversion, so even with a low converting keyword profits are still very possible. Almost every campaign loses money at first so the idea it to get some numbers and stats so you can see what you can afford to pay per keyword per click. Then it's a matter of battling Google to keep the CTR high and the QS high while lowering your bids.
i agree, a minimal of at least 100 - 200 clicks should be considered. Also if its broad match that is getting clicks and no conversions you need to start exact matching whatever broad match is mapping you to. Remember bidding on 1 broad match is like bidding on 100 keywords at once. So conversion is typically less, but can be helpful in keyword discovery
Cheers for the advice. If I had a broad keyphrase such as "blue widgets" and someone typed in "blue widgets for sale" it would probably show, now if I made a new keyphrase, "blue widgets for sale" would this automatically be used instead of "blue widgets" in the above example, as it's an exact match?
Yes it would, and that's a good way to improve your keyword targeting. Use a search query report to determine exactly what people are searching for when they match your broad keyword. If the search was something that could potentially convert, then add it as a phrase and exact match. If its something that probably wouldn't convert like "blue widget instructions", then add "instructions" as a negative keyword. Gradually you should find that your broad keyword gets less searches/clicks and eventually you can safely delete it. You'll end up with higher clickthrough rates.
I've tried running 'Search Query Reports' before but all it ever tells me is something like '1 other unique queries' - nothing about what the actual search term was, that the user typed in. A bit strange really as that's the actual reason why I ran that report! To find out what people were typing in to get my keywords showing in the advanced state - very confusing.
yes, google sums up all the variations, you need 3rd party tracking to expose all the rest. I recommend opentracker.net