rocknbil - 3:30 pm on Jun 4, 2009 (gmt 0)

I spent a few hours a day over a month or so setting it up: finding the right keywords, trading "high cost, low quality" keywords for more focused less expensive keywords, determining peak usage hours to schedule when the ads run (this is really important on a budget,) identifying the ones that performed and which ones didn't, concentrating the budget on the ones that did. One path we chose was instead of letting "keyword envy" get the better of us and throw large chunks to expensive keywords - ones you'd think were important - we'd let those go and spread the dinero among three or four other keyword ads that were less expensive. The overall effect was were were spending say, $1 per click over 3 or 4 ads instead of $1 per click on a single ad, costing less and getting more exposure.

After it got into a "comfortable" position - ads that displayed in the first page of searches, and at least half the clicks let to conversions - we just let it run.

So basically all of the ads we displayed came up on the first page of results, during our peak traffic hours, or they were kicked out the door.

We left the content network alone, too much abuse there. Also a large input factor in determining "what works" is frequent comparisons against Analytics stats.

A second client of mine had an existing campaign that someone had set up, obviously clueless to The Way Things Work. He was spending $100 per day with a very modest return on leads. Some of the clicks were costing him $3-$5 per click in a very specific niche. I reviewed his account, set up a second campaign with focused keywords; his ads now ran from 5 cents a click to 25 cents a click, with the most expensive one at 50 cents. I also scheduled his ads to run during his peak hours only.

His leads increased to %200 over previous and his daily budget was now $15 per day.

So to answer the question, initially, 2-3 hours per day for a month or so, after that, pop in and check in it every other day, 10-15 minutes.