my situation:
i've been trying to spend $100 a day on about 20 keyword phrase variations of "widget" with maximum bid of $1.25. i ran my campaign from Friday 9 a.m. to Monday 9 a.m. to test the waters. i've only spent about $18 a day!
my average CTR was 0.7%, with keywords ranging from 0% to 10%. at this point i'm eliminating keywords with 0.5% CTR and below to get a more focused grouping.
my questions:
1. how many impressions should a keyword have before it is evaluated for effectiveness? are there any impression benchmarks? is three days enough? am i running too soon?
2. should i keep keywords with lower than 0.5% CTR? that would leave me with only five or so words ... and many of those words under 0.5% had less than 100 impressions! but one CT on those words would put them above 1.0%.
3. if i need a longer initial campaign to determine my keywords' effectiveness, how often should i tweak copy or delete poor performers?
i don't want to spend six months trying to make money, so consider me a sponge. thanks!
stephanie
2 Google disable keywords with less than 0.5% automatically (after 1000 impressions of the whole campaign group)
3 Dont tweak copy much. Tweak keywords and use exact/negative matching facilities to get a good clickthrough rate. You can create more than one advert copy per campaign group, Google rotates them and gives you a clickthrough percentage. You can simply delete those that dont perform.
[webmasterworld.com...]
[webmasterworld.com...]
[webmasterworld.com...]
The great thing about adwords (and the internet in general) is you can usually get a read in as little as a day. Try seperating your keywords out into different adgroups. Each adgroup should have a main keyword and a few variations (i.e. mispellings, extended phrases, etc. Use the overture suggestion tool or wordtracker for help.) You should average out around 6 - 10 keyword phrases per adgroup based on this. Focus the copy of each ad on that one keyword. Try a few different copies based on different approaches to copy (hard sell, soft sell, informational or exciting, just to give you an idea) Give it a day or two to see which copy performs best (as explained above) and delete the underperformer.