How many keywords?

Really it’s however many are profitable. There’s money in single word short tail keywords, and there’s money in tens of thousands of long tail keywords. I’ll give you a few examples and then you can research and test new keywords for your campaigns.

Short Tail Keyword :

Auto Insurance

Long-Tail Keyword :

Dodge auto insurance

Uber Long-Tail Keyword :

2005 Dodge Stratus auto insurance

 
Short-Tail Keyword :

ringtones

Long-Tail Keyword :

Verizon LG ringtones

Uber Long-Tail Keyword :

Kayne West Stronger Verizon LG ringtone

Each type of keyword can convert for a specific niche. There’s more volume in plain short tails, but the long tails can add up pretty fast and you can get good volume with them.

Keyword Tools

I won’t really go into too much detail here, as I don’t use too many keyword research tools. I have an account with every single one lol, but I rarely use them. It’s fairly easy to generate 100k+ keyword lists without any tools. If I had to pick a paid tool to go with, it’d either be WordTracker or Wordze. Or something like KeywordSpy if you just want sheer volume of keywords. The free Google keyword tool isn’t that bad either, and it’s free.

Match Types

This deals mostly with Google Adwords. Match type selection is VERY important when it comes to your campaign. The three types are :

broad match
“phrase match”
[exact match]

Let’s cut phrase match out of there because I don’t really use it. I start all my keywords with just exact match, which means if I add [dodge auto insurance], my ad is only going to show up when “dodge auto insurance” is searched for exactly. Once I test the how profitable the exact match keywords, I expand it to broad match and test the conversion rates on them. There’s a lot more volume in broad match, but also things to look out for…

Negative Keywords

When broad matching, negative keywords are sooo important. You have to eliminate people that aren’t looking to buy, since those people are much less likely to convert. For example, say you want to sell plasma TVs. So you broad match the keyword “plasma TV”. You ad will show up for :

plasma TV information
plasma TV help
plasma TV maintenance
fix plasma TV
plasma TV support

All of these people aren’t looking to buy a plasma TV, but they may click your ad on pure impulse. This is where a negative keyword list helps you A LOT. I’d add the negative keywords :

-information
-help
-maintenance
-fix
-problems
-support

That way, my ad wouldn’t show up for any of those keywords described above. We’re now reducing our budget while increasing our profits, yayyy!!!

Outside the Box

Many keywords out there convert that you wouldn’t think of. Completely random keywords can convert on an offer – you never know if you don’t try. It’s harder to test weird keywords on Google, so I’d test out on Yahoo and MSN first. I can’t really say too much about this area except that it does exist, because I can’t really think outside of the box about nothing. It all depends on what campaign you’re running, and what you think the lines are within what will convert and what won’t.

Hopefully this article gave a little more insight as to how I look at keyword research, and a couple important factors that come into play.

While Shoemoney saluted uploading 20 million keywords in adCenter goodbye, this concept isn’t totally dead yet. Just go into Yahoo or MSN and start searching the most random and untargetted keywords. You’ll see a couple reoccurring sites promoting ringtones ;). Here’s the simple process I took with ringtones a few months back (and still have a couple of the campaigns running now). It’s nothing complicated, because the point here isn’t to think :

1) Picked an offer, ringtones were still pretty good 8 months back and it was when I first started affiliate marketing so I chose that.

2) Went into Yahoo (Google’s QS rapes these campaigns) and started searching for any random keywords that popped into my head. If I found a keyword with only 1-2 ads, I marked that down. Once I had my list, I picked some of the keywords and started researching keywords for the general niche those keywords were in.

3) Typed those main keywords into Wordze/Wordtracker (that’s what I used at the time) and got ~10k keyword lists for each niche.

4) Some of the niches I promoted ringtones in so you can get an idea (I’m not saying the ones I’m still in and are doing the best, sorry folks) :

-video games (profited with this)
-furniture (wooden chairs, couches, I actually broke even on this campaign)
-anime (can’t remember)

5) Made a base landing page, then swapped a dynamic header in there saying “Get {Niche} Ringtones!” or something like that.

6) Uploaded the keywords into Yahoo and bid $0.10. If the niche was converting, I upped the bids and got more volume at a slightly lower ROI.

My best campaign using this method was making around $1,000/day at it’s peak, which isn’t bad at all for a random niche having NOTHING to do with ringtones. Note that the $1k/day didn’t come from just that initial campaign. I was able to bid $0.50 with a really nice return, and I expanded the keyword list to about 100k.

For a couple weeks I tried figuring out what keywords were converting by looking at all my stats, but it was just huge lists of completely different keywords. I exclusively used YSM for those campaigns too, I didn’t mess with MSN a lot, or any smaller engines like Ask. So not thinking at all can still sometimes come in handy if you want to make a quick easy buck or two.

I’ll probably write future articles about thinking outside the box (I definitely will actually), but sometimes you can still “exploit” things and just completely forget there’s a box even there.