Barely avoided PPC trainwreck. Please help! (long post)

[BobDolemite]

Hi everyone, I'm new here and new to AM. I watched and read several guides, and I decided to get into PPC. The one guide that I've been using the most is the case study video at Cash Tactics. I originally thought this would be pretty straight forward, so I went ahead and began preparing my first ever campaign: an email submit campaign on AdSense using direct linking (testing for now). I set it with a budget of $50/day with a total budget of $350, although I can up these numbers if it's recommended. Fortunately, I paused and deleted the campaign before it started, but this is how the near trainwreck happened:

1. I made a huge mistake of dumping all of my 300 keywords into one single ad group in AdWords. It's my understanding now that I'm supposed to group my similar keywords (ideally under 10) into ad groups. How many ad groups and how many keywords do you run per campaign? The case study video actually said to use thousands of keywords and to use one keyword per ad group. But this has me REALLY confused because he also used the same ad for all of them. From other guides, I'm under the impression that I should be creating at least two ads per ad group so that they're better targeted. So let's say I end up with 30 ad groups, does this mean I need to create at least 60 ads total?

2. If I'm doing split testing, with at least two ads per ad group, how do I set this up in Excel?

3. I'm using Tracking202. Based on the example of 60 ads total, does this mean I have to recreate all of these ads in Tracking202? And then do I have to create 60 different tracking links and then feed them all back into AdWords? This seems really tedious.

4. Any general advice? Should I be reducing my total number of keywords? I thought it was a good idea to spread thousands of seeds all over the place to see what grows and what doesn't, and refine from there. The case study video even mentioned using thousands of keywords. How can I realistically manage that many ads?

5. Should I be advertising on search only or content as well?

6. Any recommendation for using broad, phrase or exact?

7. Should I be using manual or auto bidding at this stage?

Please help a newb out!

[dburk]

Hi Bob,

1. Yes, create 2 ads per ad group. The idea is to split test to see which performs better.

2. If you know excel well, then you can use that. You may want to consider downloading Adwords Editor from Google.

3. You could do that, that's one of the drawbacks of Tracking202, the setup. It's very important to track so take the time to setup a tracking system that tracks earnings down to the individual keyword level.

4. It is generally a good idea to have a lot of keywords, that have little competition, to keep your CPC as low as possible. You should consider a shorter list while you are learning the ropes.

5.Test both, but create separate campaigns for the different networks. stick with just the search network until you have it working satisfactorily, then expand to the content network.

6. Avoid using broad match, concentrate on exact match and use phrase match keywords to pickup the long tails.

7. Manual bidding, unless you have invested in a sophisticated bid management tool.

[nettech]

If you are using targetted keywords then I would disagree with NOT using broad match. I would use broad/exact and phrase match. One tip that always goes un noticed is the power of negative keywords. A lot of people forget to use these and end up with a lot of click wastage.

For example, I used to have a huge cruise client. Becasue they had a huge budget, even though I advised against it they were adament they wanted to target the term 'cruise'. Anyway, cut a long story short, we implemented the negative term 'tom' so all those looking for Tom Cruise would not see the irrelavent ad and costing my client thousands. In general you can use words such as free, info, review and lots more to really laser target your terms.

With reagrds to Ad Groups, yes you need to target relevant keyword types into a single ad group. For example, 'property' and 'properties' need to be in seperate ad groups, some will disagree but this is something I tested over a period of time and saw a much better CTR.

I'm not sure about the courses you are referring to but I was self taught and learnt that to really make campaings work, it all has to be done manually. Forget any bidding software. Focus on improvng CTR all the time and in the long run this will help your RoI. I would go against 1 keyword per ad group, thats just riddiculous, just make sure that they're very very relevant and you'll be ok.

With regards to ad copy, make it highgly relevant too. Use the most relevat keyword in the title as well as Dynamic Keyword Insertion (although this does not imprive your quality score). Also, ensure you use it in the ad copy and in the URL as this will help quality score, but DONT spam.

Sorry, I don't want to sound like Im preaching how to do PPC but I realised a long time ago thats its all in the planning, if you plan your campaigns correctly, you'll generally do a lot better.

Are you using the Google AdWords software? That should help you manage your campaigns better.

Dont expect to have incredible CTR's from the outset as it takes time to build up campaign history, I would give it about a month before you can really start to see any patterns, however, do ensure that you optimise every day if you can.

I hope this helps!
Z

[nettech]

Hi,

I've read about this somewhere, excuse me if I don't sound overly excited. Doesn't free actually mean that you need to spend money with the Google AdWords program? Just to clarify, myunerstanding of this program is that you DO spend money but hopefully you make money back or something. Woudl apprecate some clarification on this thats all.

Originally Posted by ibnu

A new breakthrough secret is all you now need in order to get your Google AdWords pay-per-clicks FREE!

A gentleman from New York discovered what he calls an "oversight" on the part of 99.9% of all marketers that allows him to get otherwise paid-for advertising at Google as well as all other search engines that allow sponsored ads.

And no, nothing about his "secret" is illegal - nor does it require that you know someone on the "inside" at Google, Yahoo, MSN, Overture and others.

Instead, the New Yorker boasts proudly "...this is something that I caught onto just before 2000 when there was so much search engine craze running around, and started doing small just to test things at first ... but which I later expanded on after getting the hang of it."

This same fellow went on to start and operate sixteen separate online companies selling everything from pet food, DVDs, children's toys & games, books, software, and sold not only his own manufactured products but became an affiliate for other web businesses - all the while applying his mastermind secret.

Over the course of nearly eight years the New Englander confesses "I've actually gotten over $87 million in advertising that using my secret I never had to pay for ... and the largest share of which was more recently in Google pay-per-clicks as well as other forms of pad advertising at search engines ... all of which I got for free ..."

[JeffMitchell]

As far as I am concerned with broad, exact and phrase match I usually do a setup like this.


ex. Say I want to spend max $0.40 per click.

Exact = $0.40
Phrase= $0.30
Broad = $0.20

That was I am spending more for precision then I am on flawed algorithyms within the google system.

Hope that helps

Jeff Mitchell

[JeffMitchell]

p.s Here is a neat little free software that will help you work on your adwords campaigns pretty easy.

I have no affiliation with this product (just found it in a search) there are no upsells or anything that I am aware of. I have never even recieved an email from the creator....besides a congrats message and a download link.

Hope it helps you out.

Free PPC Tool | PPC-Maverick.com

Jeff Mitchell

[Lucid]

Bob, do exactly as Don says.

[BobDolemite]

AAAARGH! I just lost my post because the forum logged me out. Let me try again...

Thanks guys! VERY helpful! Here is my new game plan and a few more questions:

1. I'll use about 5 ad groups with about 10 keywords each.

2. A combination of broad, phrase, and exact. When using broad, I will have to be smart about negative keywords.

3. I'll start on the search network until I find some winners, and then I'll move on to the content network.

4. nettech, you mentioned that I should aiming for a month's worth of data. My original plan $50/day for seven days. What is a reasonable daily minimum if I want to do a month of testing? The average CPC for all of my keywords is $0.05, although the first page bid estimate is as high as $2.50. My current max CPC is $0.08. Should I change my bid?

5. I'm trying to get the hang of AdWords Editor, but can someone please explain Placement Max CPC? I read up on it and I understand it as being a means to set a max bid for a specific keyword. However, I'm not certain why I'm seeing it at the ad groups level where I can't specify a specific keyword.

6. Any daily optimization tips?

Thanks again!

[nettech]

Sounds like a good strategy although dont limit yourself to only 10 keywords, I would put as many relevant keywords as you find, even if this means 100.

4. nettech, you mentioned that I should aiming for a month's worth of data. My original plan $50/day for seven days. What is a reasonable daily minimum if I want to do a month of testing? The average CPC for all of my keywords is $0.05, although the first page bid estimate is as high as $2.50. My current max CPC is $0.08. Should I change my bid?

I've tested even with a few $$$s, your average cpc rate is quite low so you could probably spread that 7 day budget out over the month. It totally depends on your niche and how popular it is. If you target keyword phrases that get a LOT of clicks then you may risk your account stopping at a crucial point of the day as its been maxed out....this is just life. What you're ultimately trying to do is pick up patterns and trends, if you know that users convert between a certain time, you can schedule your ads to be displayed at certain times and therefore make use of your budget, this is why I say that its best to have about a months worth of data. I would initially display ads equally and then once I know certain ads convert thats when Id target those more often...this is where you start to split test. As an average, what are the top 5 positions paying, any idea? Again, once your ads have some history then depending on how good your quality score is, you can ourtank ads higher than you for less than what they pay, this is the beauty of PPC when its done properly.

On that max $0.08 bid, are your ads being displayed on the first page and during peak times? IF no then you may need to increase it. i dont know the niche so its hard to say.

6. Any daily optimization tips?

I used to go through all my ad groups and keywords, look at the ads and see how they compared to the day before and the last few days, I was obsessed with CTR as this was an indication of how users received ads, look at conversion rates, average positions of the keywords fom the last few days, just use common sense to be honest. You will start to pick this up when you start to work on your campaigns.

Thanks again!

You're welcome, feel free to click on the thanks button...lol

[nettech]

I'd pretty much agree with all that, only the broad match part, I would use broad match but just make sure they're long tails thats all. the rest is excellent advice!

Originally Posted by dburk

Hi Bob,

1. Yes, create 2 ads per ad group. The idea is to split test to see which performs better.

2. If you know excel well, then you can use that. You may want to consider downloading Adwords Editor from Google.

3. You could do that, that's one of the drawbacks of Tracking202, the setup. It's very important to track so take the time to setup a tracking system that tracks earnings down to the individual keyword level.

4. It is generally a good idea to have a lot of keywords, that have little competition, to keep your CPC as low as possible. You should consider a shorter list while you are learning the ropes.

5.Test both, but create separate campaigns for the different networks. stick with just the search network until you have it working satisfactorily, then expand to the content network.

6. Avoid using broad match, concentrate on exact match and use phrase match keywords to pickup the long tails.

7. Manual bidding, unless you have invested in a sophisticated bid management tool.

[BobDolemite]

I must be blind because I don't see a Thank You button anywhere. Am I completely missing something?

1. I'm thinking about going with $20/day for 30 days, although the amount of days may change depending on what trends I notice.

2. How can I find out what the top 5 positions are paying?

3. I just noticed that the Search Volume Trends for all of my keywords say "No Data." Is this a bad sign?

4. When you were watching your stats daily, what actions did you take in response? Increase/decrease bid? Pause keywords? I was planning on doing split testing on a near daily basis. Starting with completely different ads, finding the right angle, and then honing from there. Do you only start split testing after the first month? I guess I'm having a problem with figuring out what is considered too drastic of a change and the timing for those changes.

I would love to run the best campaign possible for my first go, but I'm thinking I should just take the plunge and learn from my (expensive) mistakes.

[Lucid]

Thanks are bottom right corner of the post.

You cannot find out what anyone is actually paying. Don't go by the estimate in the keyword tool. That's just some calculation based on average QS and historical data. It can be easily beaten.

I don't see "no data" as a bad sign. Simply that they don't have enough data. Of course, it means don't expect too many impressions from that keyword.

You must wait until you have enough data before making any changes. I and others recommend at least 30 clicks for each ad. There are other factors involved too.

Use phrase and exact matches more than broad.

[BobDolemite]

I swear, there is no thanks button in the bottom right corner. I see Quote, Multiquote, and Quick Reply. I even tried checking with another browser. Do I maybe have to hit a certain post count?

Thanks for the 30 clicks recommendation!

Originally Posted by Lucid

Thanks are bottom right corner of the post.

You cannot find out what anyone is actually paying. Don't go by the estimate in the keyword tool. That's just some calculation based on average QS and historical data. It can be easily beaten.

I don't see "no data" as a bad sign. Simply that they don't have enough data. Of course, it means don't expect too many impressions from that keyword.

You must wait until you have enough data before making any changes. I and others recommend at least 30 clicks for each ad. There are other factors involved too.

Use phrase and exact matches more than broad.

[nettech]

I totally agree with this, ermm thanks is on any other posts except your own...lol

What I used to do each day, was go into the campaigns and just have a quick overview and see if there was anythign alarmingly wrong. Its really acombinations of CTR, conversion rates and also things like position, obviously in your ad is averaging position 10 then you may not convert as well. Its quite hard to go throgh every detail but just be reactive to any changes that need to be made, just focus primarily on CTR, thats what I would really suggest. CTR covers a whole array of things so its a good starting point.

Originally Posted by Lucid

You cannot find out what anyone is actually paying. Don't go by the estimate in the keyword tool. That's just some calculation based on average QS and historical data. It can be easily beaten.

I don't see "no data" as a bad sign. Simply that they don't have enough data. Of course, it means don't expect too many impressions from that keyword.

You must wait until you have enough data before making any changes. I and others recommend at least 30 clicks for each ad. There are other factors involved too.

Use phrase and exact matches more than broad.

[BobDolemite]

LOL, okay this thanks button thing is really bothering me. I just tried to post a pic of what I'm seeing, but I'm apparently not allowed to post links or images until I hit 15 posts.

[BobDolemite]

THERE IT IS. Okay, so in order to thank someone you need at least five posts.