Direct Linking PPC : Let's give a try

Hi

In the general idea of mastering the Top Best Practices of Marketing to make a lot of money, I'm beginning by setting up the more simple channels :
Adword PPC "Direct Linking" to Affiliate Products Sales Page.
I'll post my actions and learnings here, and will avoid as much as I can to talk about paying products and sources I use.

If you know less than me or practice differently, maybe you'll learn things. If you know more than me or practice differently, maybe I'll learn things !)

Day 0 (and the weeks before)
I've set up the PPC on Yahoo, as Google has become out of order for Direct Linking (so as some famous marketer would say, "Screw Google"...).
For the Affiliate Products, I've chosen a niche which is :
1 - getting quite a lot of advertising (which means there is a lot of money in it),
2 - having heavy emotional significance to target people (desperate, personal, painful...) so that they are about to spend money on it.

Then I go for a product (to begin) :
1 - Sufficient Gravity on Clickbank to ensure other marketers bother selling it
2 - Suficient Earning per Sale to ensure that with a 1-5% selling on a .10 to .75 per Click I'll generally pay for ($30-$75)
3 - Sales Copy that seems to match efficiency check-list (emotional, bonding, and heaps of other factors...)
4 - Matching the "Avatar" of my lead when I put myself in its place
5 - Establishing Credibility, Overcoming the reader's objection and Ensuring a Full Guarantee
If I was On Google, I would have a look at how their Site would be Quality Score compliant, because it can have a LOT of impact on your campaign cost.. hence it's profit.
I begin by 1 product, then I'll go to three and tune to the best-earning one.

I organize One Campaign.
- I Get HopLink (the page to which I'll direct the clicks)
- I set for US market (US people have money and pay for value, which is not a general rule for my own country, I'll extend to other english-speaking markets after).

Then I search for keywords :
- Google Keyword Suggestion Tool on the SEO Competition (sounds great, uh ? I typed the main keywords on Google to get 20 domain urls) and the Sales Page analysis ;
- PPC Competition Analysis (who do bid on it for long enough)
My goal is to set a first burst of the main Keywords of the Niche.
** There are strange keywords those guy bid on ?? Not related to the market ?? Let's try too, that may not be a mistake on their part.
That's about 445 Keywords. No way to do a 12 000 keywords list if its most targeted/meaing 445 Keywords don't bring money at first.

Then I set up 30 or so tightly (semantically) focused goups of Keywords, so that the Ad would match one Avatar sub mindset/stride of search.
On a Google Campaign, I would have gone for a 1 or 2 keywords per Group basis, because it has great impact on cost and display rate.

I then set some tracking system.
I did not found a fully free one I could get API or Databse access to, yet it wil repay itself soon by cutting my losses on non profit-bringing Keywords and bidding higher on money-making ones.

Then I go for the Ad writing.
I look at the PPC competition there is.
Ads here for a long time get my attention even more. Those Ads match what I expected to find on a heavy competition niche :
- keyword relevance / repetiton
- "Avatar" matching (how would my client think/feel/react)
- interrupting mental patterns
- low on heavy/dirty/bad overmarketing stuff
- heavy duty on emotional appeal (that cool old Roman Recipe... Blood, Intrigue, Scandal, Sex and Games...)
- invisible influence techniques
I familiarize the best ones, make some model mixes. Then I add some more, some NLP, Influence and Advertising tricks.
I try for my 30 groups some variations on half a dozen models. I intend to soon rotate ad models on groups, so that a "bad performing" group won't be penalized by an underperforming Ad copy.

This is tiring, so I'm happy I programmed for myself an Application to handle the fast copywriting and the url linking to my Tracking web server and the sendings to Yahoo (erh, the most difficult was to wait for 1 month to get an API key, I don't understand how these guys can take so much time for this).

Day 1
Mmmm.... Costed me $5, no sale
40 Clicks. Ok that's not a lot, but it's for a first look to the market and with a small budget (Bids are .10).
- CTR (how many out of 100 people seeing my ad click on it) seems low : .29%. Sure, there is a lot of competition.
- Two Groups make it for the bulk of the Ad Printing : the groups I got from the two main products pages for the niche, myne and the other. Competitor has 3000 prints but CTR .62%, and my product keywords 10 000 prints with .19% CTR.
That's logical these groups bring a lot of views, those guys know their market. I think I'll concentrate my efforts on the main competitors even more next time.
- It took me a lot of time to set up the 28 other groups to get 30 prints, yet I guess on the long run they will prove very well targeted and transforming well even if small in audience. From now I'll keep this targeting for "Phase 3" (Phase 1 is "Try The Market", Phase 2 will be "Tune the Campaign").
- I just guess that Competitor Keywords List is well targeted, or that the Ad was more efficient.
So now I look in depth in the two groups... First my product's site words :
- I see that most prints come from one .19% CTR word, that does not appear (IMO) to me very well targeted.
- Some other words got 2.08% CTR on average 8th position, I'm proud (all the more so : they are in-Avatar).
- I see 6 words that go beyond 140 prints and got no click. They are hardly matching Avatar, and most people seeing them are not avatars. Those are ill-qualified prints. Sometimes those can bring some money, but on Google they would probably kill my campaign quality score. I'll wait for 200-300 prints, won't change the ad cause it seem to perform well on qualified people, and if they don't get clicks I'll pause those.
- I'm eager to see some Avatar-qualified keywords go beyond 4 or 10 prints to see if I'll get Clicks for them.
When I go for my second heavy duty group, the one got from the competitor website :
- With 3 times less impressions, this group brought more clicks. I like this ;D
- I see a 190 prints word that is out of place :"business to business" (on a personal life product) with 0 click. Ok, I put it off : no click, no relevance, no avatar matching, and on an ad that seems to perfom correctly.
- On a non-Avatar (broad situation, and out of mindset) Keyword, I get 1.65% CTR ?? Talking about weird things. It may happen that the clickers end up not being mere tourists. I re-read my ad, it really tells about the product, so it's no mistake if the visitors clicked on it. When I'll have 100 or 200 Clicks, I'll see if they buy and bring me money.
- Other CTR are generally .25% to .55%. Far from the 3% click rates I used to get on well perfected campaigns, but it's a beginning on a new market, and I'll be happy with it if in the end those clicks don't cost me more than they bring (it's about maximizing profit after all...).

Ok. What I have got from this first day :
- I have seen some strange opportunity (1.65% CTR !) that I could have dismissed if I had not tried. "Competition" spying has a lot to teach
- and this competition-study alone gives you what you have to know about Keywords to first try a market rather than hitting blindly some low-traffic words.
- Yet "Competition" is not right on every point. I had to get rid of some non profitable Keywords.
- Got rid of heavy printing words that don't get clicks. Not as important as on Google, but worth taking care of.
- successfully tested my Yahoo API Interface (made to get a lot of work and control in a little time)
- first contact with my ad model seems to confirm that good copywriting on qualified words get more clicks
Ok. 40 clicks, most of wich may not be in-Avatar, it's too early to know.
I must have my ads print more often. I set the bid to .40, this will bring my ad printing to explode quickly. Next step is to see if it transforms when it gets clicks.

We'll see tomorrow how it works on day 2 with more heavy printing.

Any constructive return is welcome !)