| 7:22 pm on Nov 24, 2003 (gmt 0)|
The maximum CPC is determined by your business. You determine what you can afford to pay looking at the expected ROI. The traffic estimator can be used to get an idea about your position with a certain max CPC but your actual position will be determined by your CTR and max CPC.
Shaks advice a few threads down is "start high and lower it over time (when your CTR is strong)" I conccur with that.
My 2cts good luck...
| 7:28 pm on Nov 24, 2003 (gmt 0)|
- Know your website conversion rate
- Know your profits per sale
Find how much it costs per click to break-even. Add 25% to it since avg CPC is never equal to max CPC. That should be your max CPC in the beginning.
If you don't feel like doing all that, go to Overture and find out how much the top advertiser is paying per click. Divide that by two (or three) and make it your max CPC.
Review the results after 24 hours.
| 10:07 pm on Nov 24, 2003 (gmt 0)|
Thanks guys, will get stuck in ;)
| 10:32 pm on Nov 25, 2003 (gmt 0)|
Damn! - This is quite tricky isn't it?
Three main goals that appear to be the keys:
The last I can do, at least it's not my main concern anyway..
- Finding the lowest CPC rate
- Getting super targeted phrases
- Enticing, qualifying copy
The first two are problematic, the more targetted my kw phrases the less traffic they draw BUT I can't compete effectively on 2 word BROAD phrases..
Where's the middle ground here?
| 11:14 pm on Nov 25, 2003 (gmt 0)|
What works best for me might not work best for you, but I am quite happy with the CTR I get from using broad phrases with lots of negative matches. Definately start with a high CPC, get your ad established, then lower it back.
| 2:18 am on Nov 26, 2003 (gmt 0)|
|...the more targetted my kw phrases the less traffic they draw BUT I can't compete effectively on 2 word BROAD phrases. |
Yep, general keywords get a ton of impressions, and very specific keywords get fewer.
However, general keywords usually = low CTR, and very specific keywords usually = higher CTR.
So the trick is to use a appropriate selection of very targeted keywords, which between them will get the number impressions you want and a high CTR as well.
In brief summary:
* one great keyword is better than 100 terrible keywords, but...
* 20 great keywords are better than one
| 6:52 am on Nov 26, 2003 (gmt 0)|
Right, so it's broad terms with negative matches then guys?
Much thanks again ;-)
| 9:17 am on Nov 26, 2003 (gmt 0)|
* Use a combination of broad, phrase, exact matched and negative keywords packed into the adgroup. Example -
pub conference boston
pub conference shak
* If a particular phrase (say "pub conference boston") gets more than 200 impressions/day, you may want to hive it off into a seperate adgroup and add pack more broad/phrase/exact keywords into that adgroup.
* Use the Google keyword suggestion tool extensively. Use the keywords under "expaned broad matching" and "additional keywords to consider" and create new adgroups for them.
I think the new tool has made the highly overrated WordTracker obsolete.
| 3:49 pm on Nov 26, 2003 (gmt 0)|
Once your site is established, and you can track how much a "life time customer" is worth per month/year on your site, you can be a bit more creative with CPC.
Determine how many repeat sales are made on your site by people coming back directly to your site, and not through paid advertising. You can take the money you saved by not paying for them to return, and adding this to your cpc for new consumers.
For some sites, it's worth breaking even, or even losing money on the initial sale because of the repeat sales they generate over the course of a year.
| 3:02 pm on Nov 27, 2003 (gmt 0)|
Vibgyor, you mentioned:
* Use a combination of broad, phrase, exact matched and negative keywords packed into the adgroup.
I saw a post here sometime this month that suggested to someone that a Google bug exists wherein it is not advisable to pack a combo of broad, phrase and exact match all together in only one ad group.
It mentioned the better way was to create a separate ad group for each. Say instead of packing together these keywords for "blue widget" in only one ad group:
Ad Group 1
He said there was a bug, so this would entail three ad groups for the keyword "blue widget".
Ad Group 1
Ad Group 2
Ad Group 3
Vibgyor, pls. confirm if this bug does exist. Do I really have to make three ad groups for one keyword to fight this bug? Help, it's confusing....
Thanks in advance.
[edited by: Woz at 8:08 am (utc) on Nov. 28, 2003]
[edit reason] No Sigs Please, TOS#13 [/edit]
| 5:33 pm on Nov 27, 2003 (gmt 0)|
AWA..? Is there a bug like that? When will it be squashed?
| 5:44 pm on Nov 27, 2003 (gmt 0)|
I would like to know more about this too since that is how I have by Adgoups set up. They seem to be functioning correctly for me.
| 3:40 pm on Nov 28, 2003 (gmt 0)|
I suffered from that bug a few weeks ago, however, in the past 2 weeks, I've not seen it at all, at least in my adgroups. It was a pretty random problem, affected some adgroups, and not others.
| 7:01 pm on Dec 1, 2003 (gmt 0)|
|I saw a post here sometime this month that suggested to someone that a Google bug exists wherein it is not advisable to pack a combo of broad, phrase and exact match all together in only one ad group. |
It mentioned the better way was to create a separate ad group for each. Say instead of packing together these keywords for "blue widget" in only one ad group...
|AWA..? Is there a bug like that? When will it be squashed? |
I must've missed that thread (or I just don't remember it). Anyone have the URL? I'll be happy to look into it.
In the meantime, I am not aware of such a bug. As far as I know it is entirely possible to mix and match keyword types in an Ad Group.
| 11:26 am on Dec 2, 2003 (gmt 0)|
A little more on this if you'd be so kind...
If I want to find out what I'd have to pay to compete effectively what would I do?
Would I just keep increasing the max CPC unittl the estimated traffic maxed out?
| 7:18 pm on Dec 2, 2003 (gmt 0)|
Your market and your product determines the rules you need to set to compete "compete effectively".
If you want to know how much traffic you can get a day regardless of price, which is what it seems you're looking for. First you probably need to increase your spend per day to about 20% above the cap. This will make it on any given day you will show for all the queries. You also need to check the stats each day to see if you're still hitting the cap.
Next, make 3-5 compelling ads for each group.
Every thousand impressions or every two weeks (depends on the impressions - see this thread [webmasterworld.com] for more info tweak your ads for a higher CTR rate.
Raise your max CPC. Look at your ad position in a search. When you get into the premium position, or number one side position, then determine the CPC you are paying throughout that day.
Assume at this point, you're in the position because of your CPC and not CTR as you've been raising your CPC to get here. This is the max CPC you should pay. Now, once your ads start to get a higher CTR rate, this CPC will drop, how much depends on your competitors, but as your CTR goes higher, you will spend more money as more people will click on your ad.
Once your ad stabalizes into the top positions, and your CTR levels off, you'll be able to see the CTR vs CPC required for various positions, and will be able to manipulate which position your ad is in by raising and lowering your CPC.
| 5:44 am on Dec 3, 2003 (gmt 0)|
Here's the logic that I used... I want to spend somewhere around $65 a month (this is a family site and I am "donating" the promotion), but I want a decent amount of clicks. I started out at a limit of 25 cents per click, with a daily limit of 3.50 - I am promoting an online catalog for an offline business, and I just wanted to get started and see where it went. I have about 6 reasonably focussed keyword combinations in my adgroup.
Then I only turn the ads on for Friday - Sunday, trying to target the days most people would be looking for the type of items we sell, and because the store itself is only open on weekends - during the week he delivers and picks up more product.
What I get is more clicks on the weekends because X number of days pass that don't rack up any clicks because the ads are off, so I can spend the extra money on the days they are on.
I am getting about 11000 impressions a month out of this, and my CTR varies between 3.0 and 3.5%. I am usually in the top 3 or 4 slots for my keyword combinations, and I have not yet had the average CPC hit my limit.
The cool thing about all this is that if I was trying to do a random bulk-mail campaign at this level, just figuring postage at 25 cents a pop and no cost of materials I would be spending over $2800 a month, and instead of "clicks" I would be hoping for phone calls and store visits, and I would have to sell a helluva lot of product to pay for that each month.
The way I see it, out of the maybe 300 clicks a month I am getting for my $65, if even 3 or 4 people buy something I paid for the whole month's advertising, maybe even covering a slow month. Of course if I was getting reimbursed for this I might even spend more, but then they wouldn't be able to keep up anyway. I also think by having the ads I am getting more exposure on the other search engines, as I notice that I get spidered more while the ads are running. Lately I have been tweaking my cost per click and dailies up to see what impact it has on my ad position, but I'm really not going to complain about being in the top 3 - why pay more unless I start to slip a bit...
For me this is a small business person's dream... I couldn't even imagine the impact if we were more than a mom and pop store and did this for real!
| 6:31 am on Dec 3, 2003 (gmt 0)|
>> Use a combination of broad, phrase, exact matched and negative keywords packed into the adgroup.
That's really interesting. So what's the logic behind that - what does that do exactly?
| 7:13 am on Dec 3, 2003 (gmt 0)|
|>> Use a combination of broad, phrase, exact matched and negative keywords packed into the adgroup. |
That's really interesting. So what's the logic behind that - what does that do exactly?
It can improve the chance of your ad showing up because even though phrase or broad matching will catch the same terms as and exact match, you can spread out the CTR depending on what triggers the search. It is assumed that CTR for an exact match is higher than broad or phrase matches, so you won't water down the CTR with be broader terms "wasted" immpressions.
Interestingly, some terms I have do much better CTR for phrase matching than exact. This could just be the randomness of the searchers.