How much time is enough to test?

[levalencia1]

I am not an adwords expert but this is is the strategy I am following:

1. I create an adgroup for each set of keywords that will target the same ad.

2, I set the bid high to appear first.

3. After 3 to 5 days I remove the ones with low CTR over the impressions, anything below 1%.

For example: for this url:

http://www.theprinterdepo.com/hp-laser-4250-printer-q5400a-r

I created an AD, and initially I put 10 keywords,  after one week I had wasted a lot of money, after 4 days I deleted 7 keywords, and after 8 days, I left only one keyword with a very good CTR of almost 3% and conversion rate of 4%.

The question is: how much is  enough time to test and remove keywords? is one day of data good enough?  Should I make this test with the highest bid possible?

Should I keep the high bid forever??  11 dollars for one click seems pretty expensive for one printer of 200 dollars when converted.  Does position 1 guarantee more conversions?  Maybe as long as it appears on the first 5 or 6 then its enough so I can decrease bid?

[levalencia1]

I am trying to focus on long tail keywords that are related to specific products, so instead of using hp printers or refurbished printers, I use:

hp 4250 refurbished printer

that one has a CTR of 4% and a conversion rate of 10%.

But I am paying 10  dollars for the highest bid, wih about 200 impressions per day only. But its very targeted traffic of people really looking to buy that specific printer.

I wonder how can I play with the bid until I find something that doesnt spend a lot, and it still has a good CTR and good conversion rate?

The reason I ask about the bid, is that in the bid simulator it says, the higher the bid the more impressions

[JasmineA]

I agree with this.

However, you should not be bidding for 1st position with the highest bids possible. 1st position costs 25% more than 3rd position and doesn't necessarily have an arguable CTR or Conversion rate.

What keyword types are you advertising on? This is going to make a SUBSTANTIAL difference in how much you are spending.

Are your keywords major head terms for "HP Printer" or are they a little more focused like "Refurbished Laserjet Printer" -- as this second one is probably not as broad as the first, the chances are higher that it will cost less.

But the specifics of your account aside, what is it you are ultimately using AdWords for? Are you research or ROI focused? Because if this is going to stick around a little longer than a one time test, we need to discuss your processes to lower costs over the lifetime of the account.

My colleague Kristina Kledzik wrote a really strong post about how to integrate PPC into your online marketing strategy, and one of the major CTAs toward the end was "Don't give up. Try it for at least 6 months" before you can determine that PPC isn't profitable.

The takeaway is that PPC is hard work, and to make the best decisions for your business you're going to need to dedicate a lot of mental space to making sure it runs smoothly. You don't have to spend a lot over your test times, but you do have to be committed to attempting to make keywords work.

[EGOL]

I don't want to give blind advice.  So, the best I can recommend is to use the Google Adwords training... or read Advanced Adwords by Brian Geddes.

You can blow a lot of dough on adwords and save massive amounts of money with just a little knowledge.  Setting good bids depends upon your conversion rate, profit data and opportunities for recurring sales from new customers.

Be careful.

[danrawk]

Andy is correct. It's going to vary by conversion rate and your specific area of business. What kind of metrics do you use for your website in terms of period of time?  The bottom line is you need a little time to "smooth out the averages" of your data to get a realistic sense of how it's going to behave. This is especially true when considering if your business is seasonal or not.

For us, we look at season (quarterly) metrics, and then again on a week to week basis.  If it's early on and you are looking at dailies, it's useful to use that data to filter out keywords you don't want that are costing you money.

It's not going to be a 4-5 day thing. It'll take some time to tune in.

[SEOAndy]

only you can really test the position out, for different ads the conversion rate will change as for keywords i think.

You'd be best to check every 24 hours i would think rather than wait a few days - especially if you are paying more.