|Initial bid price|
Best way to start campaigns
When starting a new campaign, do you find it's better to:
A) Start off initial bid prices higher, then lower them later (an almost guaranteed way to start off losing money).
B) Keep bids at 0.05 and work upwards (potentially lots of impressions with no clicks = low CTR).
C) Something obvious I'm missing?
And whats your reason? How would this apply if you were adding a lot of keywords at once? Surely you couldn't bid high on them all and then lower them slowly, as the initial high price on a lot of words would cost a lot initially?
But then if you do it at 0.05, it'll take forever to get traffic, right?
Here is what I do on a new site. I set a moderate price on my better keywords
Not too high or low. Put a pretty good size daily budget on. Keep it on for
24 hours daily. When the high bidders have spent their budget or vacate their
their normal time slot, you should be able to move up the page and garner some
traffic. Conversion of course always depends on the site, content, the offer, etc.
This may or may not fit your marketing program, but it has always worked well
There's a correction built into the Adwords ranking formulas that makes some allowances for the fact that being higher or lower on the page will affect the ad's CTR regardless of what the ad says.
If your ad is effective enough to get a better-than-average CTR for the rank it's at now, you'll get credit for that in the formulas even if you're starting low.
The effects of a better CTR will take longer to kick in if you're starting really low, especially if you're not on the first page, but it still works.
Like so many things in life, if you want to spend less money you'll need to allow more time.
Could you be a litte bit more clear as to how you decide what a moderate price is (how do you compare it is moderate to the market) and also how you decide what your better keywords are pls (as it will be a new campaign)?
Thanks for your time.
[edited by: Gav101 at 2:01 am (utc) on Oct. 30, 2007]
I've always done it by starting about 50% higher than my max bid (so if my max bid is 30c, i'll start off with a 45c bid), but put a low campaign maximum. This way I can establish a click history so i'm not penalized right off the bat by having bids too low on historically high cpc rated keywords.
How do you estimate the average CTR for a position? For someone who is relatively new to this PPC - any suggestions to on how get an CTR estimate for different positions?
I have too many campaigns and too many keywords to get too worried about it at outset.
If it's a brand new niche that I'm not familiar with, I start everything between 25 cents and 40 cents, and let it run at least three days like that. The control is the daily budget (which is dependent on what the customer feels he can spend) My experience is that the Google estimations aren't even close, and even a single day's data isn't necessarily representative.
After three to five days, I take a look at what I have, and adjust. I may end up going back to the client and adjusting his expectations as well.
What I do next - well, all I can say is "it depends" because there are too many variables to list here - niche, customer budget, my estimation of time spent tweaking, competition, etc etc etc
On any new campaign, I figure I won't get anything CLOSE to a clear picture of the market, a reasonable CPC, and Google's idea of my QS for at least a month. At least. I don't like to make hasty changes without sufficient experience and information to hand.
YMMV, but it works for me, and I believe that most of my campaigns have been pretty successful over time.
I'm with peermedia, start out high.
We usually run/sign customer campaigns for 6 to 12 month, and the higher CTR which results from starting out high will eventually give lower click prices in the long run.
(where S(D) is quality score of a document or website)
|From google patent: Adjusting ad costs using document performance or document collection performance (United States Patent 20060004628) |
 If there is not enough data to set S(D) with a desired degree of confidence, the computed S(D) can be mixed with some default value. For example, the score S(D) may be set to (a) the default value if there is no confidence in the calculated S(D), (b) the calculated value S(D) if there is complete confidence in the calculation, and (c) some combination of the default and calculated values for intermediate levels of confidence. The default value might be set low for brand new documents from new publishers (to have them prove themselves).
What you might find: when you start off with 5c bids that a lot of keywords min bids will jump very quickly to 10c, 15c etc… and you will have to bid that much or little higher to get history going and when your min bids settle down you should adjust your bids accordingly.
The problem here is that when you bid higher you might find yourself in a negative ROI territory, and you have no idea when your min bids will come down and to what level. I had a campaign where all min bids started at .05c then went up to .10c and was going back and forth between .10 -.05 for a month before min bids came down to .04 -.05 range and stopped fluctuating. Then I adjusted the bids to .07-.08 range making ROI where I wanted it to be in a first place.
If I were you, I would start bidding a little higher on your main (most profitable keywords). When few of your secondary keywords go inactive keep in mind those min bids could come down later as your main keywords gain good history.
Thank you all for your answers.
It seems to me it is a better idea to start out with fairly healthy bids for the main keywords, and then maybe add the less targeted keywords at a later stage, after running up a decent history perhaps, and getting an idea of the market reactions to your ads etc. I guess it is better to spend some money fast and find out quickly.
Thank you all for your responses, you're really helping me, as I'm a total novice when it comes to PPC.