Results 1 to 3 of 3
September 10th, 2007, 08:41 AM #1Opinions on Time Frame
Looked for a thread on this but didn't find one that addressed this exactly.
When each of you create a new ppc ad campaign what is your time frame before tweaking the campaign in terms of ad changes and bids? Do you micro manage from day 1, do you watch for a week, a month before making changes?
Obviously once you reach a point where you trust the feedback and data you continue to tweak, but how long do each of you wait to start the tweaking?I would rather live my life as if there is a God and die
to find out there isn't, than live my life as if there
isn't and die to find out there is.
September 10th, 2007, 09:18 AM #2
I would think a minimum of three days, to make sure you get distribution beyond just Google. The first day or two can be deceptive since it'll only be showing on Google most or all of that time. Those results (with just Google) are usually (but not always) better. Once the search network (and/or content network, if you're using it) come into the mix, things change.
Beyond that, I usually look for a number of clicks rather than a length of time. I like to see enough clicks that 10 sales are generated (or should have been generated based on expected conversion ratios) so that I'm working with more statistically reliable data. With less data than that, you're making decisions based on statistics that could be quite different than what you would see in the long term.
Another thing I like to do is to work with individual keywords when I have enough data to be statistically reliable, then take all of the lower volume keywords and consolidate their statistics and work with those as a group. I give those "long tail" keywords a little more leeway and higher bids, since they haven't been individually analyzed and they are lower volume. As the collected data grows, keywords eventually get enough data so that they can be analyzed on their own and I adjust at that point. Obviously this would be very time-consuming to do manually, but you can use the API and automate things.
September 10th, 2007, 12:06 PM #3
Good stuff Coley! I'd add that I always avoid using the content network with anything new I'm trying for the first time, searchers are almost always better buyers than browsers / readers found in the content network (though there are exceptions). I also start with just phrase and exact match as well, to save myself the time involved with developing negative keywords, but just in the beginning. I do fairly severe dayparting as well, I like to set everything towards the conservative ends of things - I'm looking for failure I guess, more than success. I want to save time and money discovering early that something's not working well at all - if something's working okay, great or even at a somewhat narrow loss, I'll go very deep into exploring different areas. But the early tests, are designed to quickly identify when it's time to bail completely, there's too much to choose from and I've got too much work to do than to be chasing poor programs for long periods of time. I want to find winners or near winners and then work on maximizing them. I also don't do time based, volume extremes can make that the wrong way to measure things, so I tend to look for 300 clicks or so, unless there's other reasons or events that make me go in another direction. Like Coley mentioned, 10 sales is enough for me also, so if I've had 10 sales after only 200 clicks, there's no reason to wait for clicks to reach 300 to decide to commit to research and begin experimenting with all sorts of variations. And in some niches / programs, where I know from experience what things should generally look like, I may not wait for 300 clicks or 10 sales. As you do it over and over, you'll codify it in several ways, some very tangible and rigid and others that are less so. Good luck!
By AnonD in forum ShareASale - SASReplies: 5Last Post: June 6th, 2010, 12:47 PM
By TLDnetworks in forum AvantLink -AVReplies: 12Last Post: May 10th, 2010, 04:18 PM
By Rhea in forum Midnight Cafe'Replies: 10Last Post: December 7th, 2005, 03:20 PM
By Haiko de Poel, Jr. in forum Midnight Cafe'Replies: 19Last Post: February 21st, 2002, 11:11 AM