I discovered a really cool site today that all you pay-per-clickers are sure to appreciate. It’s a free tool which analyzes your ad data and tells you which one makes you more money!
Sound straightforward, kinda like a no-brainer which you think you already know how to do? Might want to think again. The premise behind this is that most marketers create ad variations, split test, and then the one with the strongest click through rate (CTR) carries on to round 2. That’s a good guesstimate of performance, but at the end of the day, we don’t really care how many people clicked our ad, do we? We care about how many people performed the required action! So we really need to take conversions into consideration too.
So how do you do the math to get a useful result out of all the numbers? Glenn Livingston from Super Split Tester has created a cool tool to do it all for us. Simply enter in your numbers and it will tell you, along with a confidence interval, which ad actually adds more to your bottom line. You might be surprised, the ad with the lower CTR might actually produce more sales per thousand impressions than your previous star performer!
The other day when I logged into my Adwords account I noticed one of those fun alert boxes at the top, advising me that Adwords had a new feature.
Upon further inspection, it appears that Google has finally decided to let people see what sort of search volumes each keyword is getting. This is a great new source of data for internet marketers, who have previously been relying on all sorts of other sources which are then extrapolated to guesstimate total search volume.
For instance, to carry on from my last keyword research post, I’ll use “affiliate marketing” as the keyword again. Previously, if I wanted to research this term I would have gone to the free tool at WordTracker, and punched it in. They give a result of 546 people per day, which is something like 16,000 a month. Now this is only a small percentage of the real searches, so you always need to do some sort of rough extrapolation to figure out where the market was at. I read it was 1% somewhere, so if you multiply by 100 you get 1.6 million searches.
Well, what does Google data say? Google says that approximately 1 million people search for that term every month. So if you throw in Yahoo and Live/MSN the 1.6 million figure is probably in the ball park. Still, isn’t it nice of the big G to give us a sort of semi-firm approximately averaged number for our research?
Do you use pay-per-click? Are you into search engine optimization? Either of these marketing disciplines hinge on proper keyword selection. You want to be in front of your customer when they’re in a buying mood, not a browsing mood, right?
I’ve always found that keyword research is one of the hardest things to when creating an ad campaign. How do you find out which keywords make money, and which ones burn money? Anyone can add all the suggestions from a keyword tool into their campaign - but I guarantee they will lose money unless they take that a step further. So how do we find profitable keywords? Here are three steps to help this process.
Step 1
The first thing to do is have a little brainstorm session. For whatever product you’re promoting, try to get inside your customer’s minds:
What is causing them to be looking for your product?
What pain (problem) are they feeling or dealing with?
What questions do they have?
What solution are they trying to find?
Once you’ve answered these things jot down the different phrases you think they would use to find you.
Step 2
The second step is to use a keyword tool such as the free keyword suggestion tool at wordtracker.com. See if the terms you thought of are showing up on their lists, and find out how many searches are done daily for each keyword. The higher the better, obviously, though the highest keywords might not be the most profitable.
Step 3
Up until now, the information here is probably old hat for most of you. Here’s the golden third step. Head over to MSN’s AdLabs Online Commericial Intention Detector. Click on query, and punch in one of the keywords you think has potential. The web form will then spew out a number between zero and one. Here’s the readout for “affiliate marketing”
Result: NonCommercial (Query)
Probability for Commercial Query:
0.33748
As you can see, MSN thinks this phrase is by and large, non-commercial in intent. Having less than a 0.50 indicates non-commercial; anything over .5 (50%) is commercial in intent. The stronger or higher the co-efficient, the more profitable this keyword is likely to be!
Go to the tool, and compare the output for “personal loan” and “personal loans” - you’d perhaps be surprised to see that the singular is scored at 0.67969, while the plural is scored at 0.89554. So although both look to be quite profitable, “personal loans” is the keyword with the highest commercial intent, of those two.
So now you run your list of keywords from steps one and two through this tool, to make sure that they are all above 50%. If you’ve got keywords that you think would be good, but they’re under 50%, ask yourself why, and see if there isn’t a small variation that can be made to boost the score. If you can improve on it, use the new keyword instead.
Now, I should clarify that although I’ve talked about this as being a way to find profitable keywords, I guess really it is just a method to find the keywords that people are using to buy stuff. This is great, if you are doing SEO. If however you are paying per click, profit takes on a new meaning, as you need to consider what your costs are for each keyword. Chances are strong that the highest rated keywords are also those with strong competition.
However, in all markets, the presence of sharks indicates there is prey in the waters; competition isn’t a bad thing, you just need to find a way to best them, and you’ll be laughing!
I’ve been so busy lately trying to work out some kinks in getting my first webstore online (selling digital picture frames - and yes - I’m trying to get some link love here =) that I’ve somewhat neglected my affiliate marketing endeavors. However, as the site is slowly coming together I find myself starting to think about how I’m going to market it. As I previously posted, I have had no end of SEO related sales calls, some of them totally discounting the value of PPC advertising. I think that is rather misguided. SEO alone is like having one leg on a stool. You need more legs if you want to stand securely!
Pay Per Click advertising gets you results fast, and it also helps you really drill down to find the money words, if you’re doing proper analysis. This can really help inform you when it comes to SEO, and choosing what keywords to optimize around. So I set out on a quest to find some tools to really help me do that proper analysis, and along the way I’ve discovered a couple great ones. If you’re into affilate or pay per click marketing, I highly recommend you seriously consider these.
Disclaimer: This is not a sales pitch, and even though my affiliate links are in here (you can only be an affiliate once you’ve purchased) I honestly do recommend these products. Evaluate them on their own merit.
Affiliate Radar
I’ve been looking for a way to quickly build campaigns that are trackable at the keyword level for a long time. I’m fairly handy with Excel so I’ve limped along with some homebrew solutions for quite a while, but when I found Affiliate Radar I knew I needed it. Their slogan is “Don’t Fly Blind” and it’s true. If you’re an affiliate marketer and you’re not tracking the performance of every keyword you’re advertising on, I can almost guarantee you’re both losing money (even if you’re making money) and at the same time leaving money on the table.
The great thing about AR is that I’m now able to easily implement keyword-level and ad-level tracking on pretty much any search engine I choose. I can send all the traffic to one page if I wish, and the software will handle it all for me. Later on, I simply go into my affiliate network, grab the relevant report, import it into AR along with my cost data and boom - it spits out a gigantic report that gives me cost / revenue data per keyword / ad across all my campaigns all at once! It takes seconds! From here I can quickly identify which keywords are winners and which are losers. This allows me to save a ton of time testing new affiliate offers, and means I’m much more likely to hit upon a profitable combination (even if I’m only running two keywords!).
Another interesting thing about the method of tracking AR uses is that it assigns a unique ID to every keyword - such as 7A000123 for example. The only place this has relevance is in their database, so the manager of your affiliate program will have no idea which keywords or tactics you’re using to drive sales. This effectively keeps your secrets safe.
Affiliate Radar is setup for simple report importing for most of the affiliate networks you can think of (and they’ll add more on request). CJ, Clickbank, Linkshare, CPA Networks, Azoogle, the list goes on and on. One of the coolest features is the ability to quickly export any given campaign to Google, Yahoo, or MSN. What I do is build a campaign in Adwords Editor (another FREE must have for anyone using Adwords), export the whole campaign into Affiliate Radar, add all the special keyword links, then export back to Adwords, adding Yahoo and MSN at the same time. This can really triple your efficiency when building new ad campaigns.
Speed PPC
I just got Speed PPC the other day, and I’ve yet to really take full advantage of its power. However, using these two systems in conjunction can provide some stunning results. Speed PPC allows you to manipulate keyword lists into massively long lists, broken down by category in order to provide maximum relevancy. For example you might have two lists, one of type of shoe (running shoe, tennis shoe, hiking shoe etc) and another of brand names (Nike, Reebok, Adidas, etc). Speed PPC can merge these together into every combination you can think of (Nike running shoe, Nike tennis shoe, Nike hiking shoe) and group them together. You can then do similar things building the ads themselves. The final screen lets you export all that data into Adwords Editor where you can have a campaign with hundreds of keywords broken down into dozens of adgroups live in about a minute.
Now take this data from Adwords Editor, send it over to Affiliate Radar, add keyword tracking, then export it back to the three main search engines and you’re well on your way. A competent user of both programs could have a fully fledged, optimized keyword and ad tracked campaign (with hundreds of keywords) setup from start to finish in just over 10 minutes. Where I come from that is pretty impressive.
Adwords Editor
I thought I should mention Adwords Editor in here as well. I know a while ago I was conned into buying a similar product that allowed me to edit my Google Adwords campaigns from a Windows client; however the software didn’t work that well and I ended up discarding it. Little did I know, probably at the same time, Google was releasing their free tool called Adwords Editor. By far the most impressive feature of this software is the ability to easily import / export bulk data. I can add hundreds of keywords to multiple campaigns from one screen, or change the max bid on all my keywords simultaneously. Did I mention it is free?
The other day I decided to try promoting a new affiliate product, as I hadn’t done anything new that way in a while. So I picked one from CPA Empire, which had a good network EPC (Earnings per click), and paid per lead. I’ve had good success with lead based affiliate programs in the past. I like the fact that there’s a very low level of commitment required of the customer - they don’t have to make a purchase decision right then and there, and I still get paid!
Initially, I went to Google, because Google has traditionally been a great place to get a campaign up and running in a jiffy. Unfortunately, with affiliate marketing you don’t always have control over the landing page. I went to the landing page to check it out and it was one of these one page wonders. It is a very simple page, nothing wrong with it from a user’s perspective; however Google’s crawler clearly thought it was lacking. So only 3 of the 40 keywords I wanted went live, and only after I put my minimum bid up to $1.00 / click. When I looked into it further with Google’s tools they told me the landing page sucked. Uh-huh. Already knew that.
So anyways, I let the thing run, and managed to get one lead that day. Unfortunately I spent nearly twice as much on Adwords as the value of the lead. The extremely interesting thing was that I actually got clicks on keywords that were classified by Google as Inactive. Has anyone else seen this? Is it some sort of fraud, or is Google actually letting me get those clicks for some reason? I thought that was extremely interesting.
So I recognized this wasn’t going anywhere, so I thought for a second, then realized if the landing page was the problem, from Google’s perspective, then maybe I should try it out on a different engine that didn’t care nearly as much as Google does about these things. So I fired up my trusty 7Search account and plugged in basically all the same information as went into my Google campaign, and turned it loose.
The immediate difference: I was paying approximately 14 cents per click for second and third place positions on 7Search, whereas I was paying $1.00 per click on Google just to get active. I knew from the start that at least I wouldn’t lose my shirt if nothing much happened, 14 cents a click is a lot more forgiving than $1.00.
So that has now run for about 3 days on 7Search, nothing spectacular, but I have got 97 clicks so far, at a total cost of $14. Out of that traffic, I’ve managed to generate two leads, at $10 a piece. So I’m actually showing a very modest profit. That’s encouraging. Better than a loss, but still nothing to throw a party for. Now I’ve got to figure out how to grow that so it does that regularly and consistently.
Anyways, I guess the lesson of the day is that you need to look at the landing page you’ve got and then go from there; if it doesn’t have much text on it, chances are strong Google isn’t going to like it, and you might want to try it on a different engine first.
I got an email today extolling the virtues of a new service from my web host, Powweb, called TargetClicks. (I’d like to preface this post by saying that I love Powweb as a web host - I’ve been with them for years and they’re really top notch). I’ve copied in the main bullets of the email here:
WHAT THE TARGETCLICKS TEAM CAN DO FOR YOU
TargetClicks ensures that when people perform Google, Yahoo! or MSN searches on keywords related to your site, an entry for your Web site will be listed next to their search results! All you have to do is …
1. Fill out your site information. TargetClicks will determine your competition, target region and optimal keywords.
2. Pay just $1.75 per click. TargetClicks takes care of ad placement for you and eats the cost, even if the lowest bid price is more expensive!
3. Review your performance. TargetClicks provides easy-to-read e-mail and online reports that track your progress.
Is anyone else chomping at the bit to get signed up? I saw $1.75 per click and nearly laughed out loud. Then my brain kicked in, and I started thinking about it. I realize they’re really targeting the people that have no clue about pay-per-click or internet marketing, and would just rather pay someone to know for them. That is fine, there are things that I don’t really care to know much about and would rather pay someone to do, such as fix my car for instance. Are you an ad mechanic? If you don’t like getting under the hood, perhaps a program like this fits your ticket.
Thing is, I do happen to know a thing or two about pay-per-click ads and the $1.75 seems pretty high. Apparently though, one gets coverage on up to 30 PPC networks, which is definitely a big time saver. My guess is their incentive to create a good campaign is the more they get costs down, the more they get to pocket. The other thing to keep in mind though is what kind of quality are they delivering? I could setup a campaign for $1.75 a click and send thousands of visitors through all day long; but how targeted are they? Seems to me the disadvantage of this setup is that there is no relation to profitability.
At the end of the day, this is what makes a successful affiliate marketer; being able to advertise and get a targeted customer ready to purchase for less than the value per action so he can pocket a profit. If you’re just sending traffic down the pipe for a set fee, there will never be any accountability on quality.
I was over at AskHowie.com and found a clip where he was talking about finding the correct triggers for your product. He gave an example of selling a weight loss product for kids. What are the triggers that would prompt a concerned parent to search for such a product? It could have been a doctor’s visit, a comment from another parent, a comment from the kid themself, or any number of other things.
By understanding the different events that would trigger a motivated buyer to start searching for your product, you can market to them much better. A “triggered” searcher is one who emotionally is ready to buy, right now. All you need to do is get your product in front of them and clinch the deal with some compelling copy that plays on that trigger.
Take the weight loss example again; if you identified a common trigger being a doctor’s visit, and you created an ad that was titled or included “Fed Up With Doctor’s Visits?” or something of the like; your searcher is more likely to identify with that ad because of their recent experience. Immediately they feel a connection with your ad because you’ve identified with their experience. They get the impression you understand what they’re going through.
Taking this beyond the ad, if you can exploit these triggers in the sales letter itself you can take this to a much higher level. Suppose a concerned parent got to your site by searching for “weight loss for kids,” and when they get there, they immediately see a leading statement related to doctor’s visits. Do you think they’re more likely to at least read the whole sales letter?
If you can get inside your target market’s minds, you will have a far higher success rate than simply taking the “if we market it they will come” approach.
PS: today I randomly discovered that 138 people per day search for the keyword phrase “excuses for being absent from work” and there are no Adwords campaigns running! I wonder if there’s potential here for a work from home product… then again perhaps its a bomb? Let me know if you try it out…
The other day I logged into my Adwords account and I had one of those nifty alerts waiting for me:
Looking at my current campaigns, I don’t think this will affect me much; however I’m interested if others out there think this is a good or a bad thing. I frequently run afoul of Google’s URL policy with one of my ads - it is for an affiliate program where I link directly to the merchant’s site. Google won’t let more than one ad run with the same base URL. For example, 123.com and 123.ca can run side by side; however 123.com and canada.123.com are not okay.
The problem with this as an affiliate marketer is that if someone beats you to the punch by staking out an ad to the main domain, you’re basically out! At that point, because they’ve got a pre-existing spot in there, and are deemed “more relevant” than my hasn’t-run-yet ad, they get to keep their spot and I don’t get to run my ad unless I can find an alternate domain to send traffic to, and then convince them to carry on to my main offer. Where is the survival of the fittest in this scenario? It’s more of a survival of the first.
IMHO, Google should at least let my ad run for a few days, to see if it outperforms the incumbent ad. Some of the ads you see out there are so terrible you just know you could make a better ad that would be deemed more relevant, but we’re not even given that opportunity because of Google’s repressive regime.
In Vancouver BC where I live, there is this one intersection on Robson street (one of the busiest areas downtown) where there are no less than three Starbucks coffee shops! Out of four corners available - three are occupied by the same store! These three stores have been there for several years now, which tells me that each is profitable in its own right. If city council legislated that this wasn’t relevant enough for the average visitor to Robson street, then these stores would be denied a business license. How well do you think that would fly? Would Starbucks complain? Would the clientèle complain? Hmm.
I understand that Google doesn’t want an entire page of ads for 123.com, however if all those ads are outperforming ads for 456.com, why not? Markets will organize themselves; however Google likes to play divine architect and dictate what they should look like. Any Economics 101 professor will tell you this kind of intervention leads to inefficiencies in the marketplace and should be avoided.
I read this e-book recently called Project Quick Cash. The basic premise was why bother with all this “build your marketing empire” nonsense when you can reliably make a few hundred bucks from 10 minutes of work? Well, he’s got a decent point I guess; it is at least worth considering.
So one of the methods talked about is a way of capitalizing on a tool called Google Hot Trends. The purpose of this tool is to flag any terms that are skyrocketing above their average search volume. For instance, if national TV suddenly runs another expose on Britney Spears, even though her name is normally searched quite a lot on Google, the sudden increase in traffic due to greater interest at that moment causes Google to register a spike in traffic and call it a hot trend. This list contains the top 100 and is updated hourly.
So as marketers how can we capitalize on this laser targeted interest? Well in theory it should be easy: find a product that people consider relevant to the hot trend and put it in front of them! This can be accomplished in a couple of ways. The most obvious is to find an affiliate program selling exactly the thing people are searching for. For example, often times the hot trends include specific models of a product. I’ll open Hot Trends and see what is up right now to illustrate (click on the graphic to see it fullsize):
The above pic is a screen capture from today, and several items are immediately obvious as specific products. Checkout #19 and #27. These are two keyword versions of the same product, a Kodak v1003 digital camera. So now that you know there are lots of people looking for this particular item right now. The trick is, you’re not sure why people are looking for this item - perhaps it was just featured a “Worst 10 Products of All Time list!” Then again, more likely it was featured as a great product on some review program. If possible, it helps to figure out why it is hot before you promote it, but use your judgement.
So now you have a product idea; now you need to find a site selling it that has an affiliate program. Amazon and Ebay are two obvious choices, though there are many others. Try searching for your keyword and “affiliate program.” Another good idea is to search for exactly the terms shown, find the top site or two and see if they have an affiliate program. Sign up, and setup a Google Adwords campaign using the keywords provided by Google and turn it loose!
A word to the wise: pay as little as possible for your keywords! Using this strategy you WILL get hundreds of clicks in pretty short order; you need to make sure you aren’t paying through the nose for a lot of lookey-loo traffic.
My Experience
So I’ve tried this tactic myself a couple times so far, because the principle of it really clicked with me. It’s like selling umbrellas when it’s raining, and sun lotion when it’s hot, which just makes good business sense.
I noticed that a weight loss book was featured on Hot Trends one day, so I got an affiliate link setup through Amazon and setup a Google campaign. My results:
Clicks: 675 Cost: $80.29 CPC: $0.12
Total Amazon Sales: $351.56 Amazon Commission: $20.67
Net Profit/Loss: ($20.67-$80.29-) = -$59.62
On a different occasion I found a Panasonic digital camera trend, and again used Amazon:
Clicks: 196 Cost: $23.69 CPC: $0.12
Total Amazon Sales: $32.14 Amazon Commission: $1.29
Net Profit/Loss: ($1.29-$23.69) = -$22.40
As you can see, I kind of missed the profit boat on these experiments, losing nearly $80 altogether.
“An expert is a person who has made all the mistakes \ that can be made in a very narrow field.”
-Neils Bohr
Experience rarely comes free: I’ll chalk this up as part of my tuition. Another quote:
“Insanity: doing the same thing over and over
and expecting different results”
- Albert Einstein
Burning $80 just to pass time does no good unless you can learn something from the experience. Here are a few lessons I learned:
Lesson 1: Don’t Pay Whatever The Heck Google Wants You To Pay
Google is now infamous for the “Google Slap” part of which means they can just arbitrarily make your minimum bid whatever they want to. Some keywords are okay at 5 cents, and others Google mandates you need to spend 25 cents, or even several dollars. Presumably, this all relates to their epic quest for relevancy; however often it is hard in practice to see how. Anyways, on the weight loss experiment, I was a cocky little newbie and bid way higher than I should have. I actually accepted some of Google’s minimums (20-35 cents) because I wanted active keywords (the alternative is your keywords don’t run). This led to a much higher campaign cost than I should have had. One thing I noticed though: I had incredibly high click through rates (in excess of 20-30%) and as I checked back in every few hours, Google would let me slide my bids downwards a bit because they recognized the high relevancy. Over time I cut out the high-cost keywords, and my overall CPC came down a bit.
On the camera experiment, I didn’t cave to Google unless their minimum was only a few cents higher than I was willing to pay (I was aiming for 10 cents or less; they offered 12-15 cents for some keywords). However, I still ended up with a 12 cent CPC overall.
I suggest aiming for 5 cent keywords and little else, unless you know you have a real winner of a sales page/product combo.
Lesson 2: Amazon’s Commission Structure Sucks for PPC
When you are making in the 4% range, it is difficult to make pay per click work with Amazon. The interesting thing was that I only sold a half dozen weight loss books and all the rest of the revenue came from items like Brita water filters, camera tripods and DVDs. So I was able to take advantage of additional revenue, because my cookie was on record as the referral. This is a benefit of using Amazon versus a more targeted site; however I suspect a more targeted site would perform FAR better on the promoted item.
I suggest finding a more direct affiliate program where you’re actually able to make a good return on each sale.
Lesson 3: Pick a Good Sales / Landing Page
On the weight loss example, I was too eager to get going because the trend was volcanic hot and it was my first one, and I just KNEW there was money to be made ;). So I looked on Amazon, and they were sold out of the exact book that was listed, except for a couple of used copies, and they didn’t have a real good sales page for it. (Many products they have a good product page, but this one didn’t for some reason). So I linked to the page displaying the Amazon search results for that term. My book was top of the list, but there were a lot of other related items on there too. Mistake. I suspect this caused a number of dropped clicks right there, as people really wanted THAT product, and probably went back to Google to find it.
I suggest picking a good landing page that clearly displays exactly what people are searching for. Make it easy for them to buy! This is PPC 101.
Conclusion: Does it Work?
Although I haven’t made any money on this strategy so far, it makes so much logical sense that I find myself wanting to try it again. I think I’ve learned a few lessons, and I think with tweaking (and the right products) there is good money to be made. One serious advantage of this strategy is the sheer lack of time required to get setup. If you’re familiar with PPC, and have the necessary accounts setup already (Google and a relevant affiliate program) you can have one of these setup in less than 10 minutes.
Project Quick Cash has additional tips to making this strategy successful, but I don’t really have time to get into them here. There are also about five other low-cost strategies to turning a quick buck in the book; perhaps I will experiment with them as well and then share my results here later.
