The question was of course, what do we care, from an advertiser perspective? How does this affect our bottom line?

I looked through my ads to see if there’s any correlation between Social% and click cost. And on the surface of it looks like the answer is definitely maybe.

The Social score of my ads most mostly under 10%, with a few in the 10-20% range. I checked to see if those higher scoring ones had lower click costs.

Here’s an example ad:

social%: 15.1%
impressions: 1,202,491
clicks: 1,948
CTR: 0.162%
avr. CPC: $0.02

The campaign was running in the US and I would never expect to see 2c clicks in the US with a ctr like that, even for a long running campaign. There does seem to be correlation with the Social metric.

The highest score I’ve seen so far is 54.8%, for an ad that was launched last year and had a total of 3,833 clicks. It also had a surprising low CPC.

This is not conclusive evidence, there’s a chance that those ads had very high CTRs to start with which produced the low CPC and the history carried them through til the CTR dropped but the CPC stayed low. I’ll need to watch some new ads and see how they perform over time. Other people have however reported similar observation.

Why would the Social metric effect CPC?

Because its part of Facebook’s ‘quality score’. One of Facebook’s priorities is a good user experience. Hence them being quick to remove or penalize ads people complain about or flag, for example. Conversely, makes sense that’d reward you for ads that people actually want to see, indicating this by ‘liking’ them or in this case clicking the ad. if the ad is clicked, it is shown to that person’s friends, although its not clear to me if the ad is then shown only to ‘friends’ also in the target demo or to all friends.

Also its very likely they’d be taking this into account when raising your spend limit or having your account white-listed or black-listed in various ways. High Social scores mean a lot of people like yours ads, high quality score, so go ahead and have some more inventory – thus raising the overall user experience on Facebook.

This may mean that soon we may need to divide your campaigns into ‘clean’ and ‘dirty’ accounts. Similar to how it was done with Adwords, for example with accounts for Content Network campaigns, which produced lower account quality scores, and separate accounts for Search campaigns.

It will be interesting to see if having a poor Social score across your campaigns for a few month begins to choke off traffic to the account, raise your bids, or result in having your spend lowered. Many of us got the
warning emails about the ‘quality’ of our ads – now we know more about that means. The emails were accompanied by spend limits being slashed. The limits were raised back up – but some of my accounts rose quickly, some still have not. Ive gone through too many campaigns since then, but again – there’s some kind of Social? Quality Score in effect, and this metric is likely to at least be a part of it.

Looking at old campaigns it looks like the metric has been around since at least the start of the year. My campaigns from last year don’t seem to have it, ie have it at 0.

It will be at the Dolphin Hotel in Surry Hills, which will probably be the regular venue. They
have several areas for hire there, of various sizes, so we can upgrade as the group grows,
and some with screens we can easily plug a laptop into for presentations.
There’s also a restaurant and we have the option of ordering canapes or having a full
sit down dinner.

This time round it will just be another casual event, no presentations planned. I am still
kind of on the fence about presentations – considering the diversity of the people that
came along to the last one, whatever the topic it has the potential of alienating and
boring a large percentage of people. I felt last time for just perfect to be able to chat
to people and hear everyone’s stories about their journeys and battles in online
marketing. Some really interesting backgrounds amongst the group.

As you may have already gathered the last meetup was a lot of fun. We had it in Darling Harbor,
in the afternoon after Adtech. About 15-20 people came along, quite a few affiliates,
some people from networks and various other internet entrepreneurs who had online
businesses selling physical products and other services. The feedback has been great and
a few of us were there for quite a few hours late into the evening discussing the biz over
some beers.

I was stoked to meet some great people from the industry and can honestly say
I am looking forward to seeing everyone again this Thursday.

So if you’re in Sydney and are in any way involved with Affiliate Marketing, drop by
and say hello!

Basically with email submits the experience falls into one of the following 5 categories:

1. 1. No conversions from the start. Complete shave, throwing money to the wind. Run like hell.
2. 2. You’ll get a conversion on the 1st or 2nd clicks, then no conversion at all, while the clicks stack up. They wave a carrot, then cut you down. Particularly amusing when they give you a conversion on a test click, without you even filling out the offer. Chuckle, then run like hell.
3. 3. You start off with a realistic CR then it suddenly drops as they start to shave, after you get a few conversions- sometimes 10-20, sometimes more. This is irrespective of how the offer is backing out for the advertiser, as they’re not checking – ie, they’re simply waving a bigger carrot, before sucking you dry. These can make money initially, but you have to watch them like a hawk. Grab a few similar submits from this advertiser on that network and throw them into Prosper 5 at a time, rotating out the ones that go dead. Careful, the advertiser may drop block you from their offers, and even fresh ones will start off as cat 1. I found a lot of Amped Media’s submits behave like this. Another strategy is to network surf – milk the offer on one network, then find it on other networks and throw them into that same campaign til you run out. One offer that performed really well for me like that is A Maybelline submit that was around a while back. However the last two times I saw it on a network it was already a cat 1! An offer going from 30%+ CR to nothing on another network was obvious bullshit. These first 3 categories are all shaving, rather then scrubbing.
4. 4. From the start they aim for a certain EPC, say 5c. No matter what you do, you get shaved down to that EPC and either you beat it or not. Can make money if you can get an ultra low CPC. Don’t confuse you own crappy, untargeted traffic which is simply not converting with this though. Technically, if the advertiser is evaluating the data before limited creditted leads its scrubbing, if they’re just doing it across the board, its shaving, but who in the end, who gives a shit.
5. 5. Offer converts at a reasonable rate for that offer, which can be anything from 10% to 40%. Unlike category 3 or 4 the advertiser actually checks whether its backing out for them on a case by case basis. If it is they may let you keep going, if not your AM tells you to change traffic source, take a lower payout or stop running the offer.. if you have a good AM and are running with a decent network and the advertiser bothers to inform them. The advertiser may start to shave you down to a lower EPC that matches their own before the AM tells you anything or you ask. This is scrubbing, rather then shaving, btw.

Fact is most advertisers know its just not going to be worth it for them to actually pay out listed by the network. They know most pubs won’t bring a decent enough quality on something as shitty as an email submit, so they skip the quality control and simply rely on noobs running up conversions for them – there’s always more standing in like waiting to be shaved to oblivion, hoping to find that magics submit that actually converts. Why would they bother wading through the flocks of fraudsters, noone actually holds them accountable anyway, one offer runs dry, hey, they’ll throw up 5 more next month.

The better submits have become harder and harder to find over the last few months.

As an example, the other day I ran iPad campaigns from 5 networks – MB, NB, LFM, XY7 and Flux.

Of those 3 fell into category 2 – I got a conversion on 1st or second click, the dozens more clicks with conversions. These were LFM, MB (one of the 1.10 I think, the $3 submits behave differently) and Flux.
The NB was a category 1 from the start.
The XY7 seemed to be a shave down to an EPC, category 4. I had 2 conversions from the first 3 clicks. Then it went down to about 5c EPC and stayed there til I shut it off after a few more leads.

So all these were a waste of time. I cannot 100% guarantee shaving/scrubbing on each of these, you can always argue my traffic just didn’t convert, target being saturated after everyone and their bot army ran the offer, but it sure seems to fit the profile from past experience. Odd also that I was getting conversions on the XY7 (I paused at 6 conversions, as it was going nowhere), but none or just the 1 on the others..

When you get hit with the shave it may transfer to other offers from the same advertiser that are on the same network. They can even drop you down categories, eg all of a sudden all submits from an advertiser give you no conversions at all, even ones you have not run before, as in the Maybelline example above. They used to also blacklist you – your AM sends you an email saying your traffic sucks for the advertiser and they want you off all their offers. Doesn’t seem to happen anymore – they just shave you completely now more often – they’ll still take your crappy traffic for a few conversions if they dont have to pay you anything.

One last tip – if an email submit allows only “Email” as a traffic source it is more likely to be higher category – email traffic converts better for email submits, so the advertiser can afford to shave less. If you have well converting traffic you can ask your AM if you can run the Email only version of the submit, I dont recommend just doing it without asking, before you know it you’ll be getting an email asking your traffic source.

For more info on how email submits work, check out Smaxor’s posts for the perspective from a network owner (he owns Ads4Dough): Understanding How Email Submits Work and My Prespective as an Affiliate, Advertiser and Network Owner on Scrubbing and Shaving.

Brad over at MadPPC.com has a great post on the same subject too, with some further details on the above, if you’re interested in the hows and whys: Submits: How they work and why they are a pain in the ass. He’s got a quote there from Steve from MaxBounty who says about shaving: “no merchant or network would risk their good reputation to do this”. To which I say bullshit, plenty of network and merchants are doing just that on submits.

Anyway, you’ve read all the boring crap, so here’s some candy.

I’m going to share some of the images I used in the above mentioned iPad campaign on Facebook, simply because I got a chuckle out of it and you may too.. and be amazed at what FB let through.

I used about 40 images, mostly the expected photos of the iPad in various titillating positions. The first 20 or so were just regulars. Then in the end I threw a few more interesting images that I thought may or may not get through, for something different and a bit of fun. Worst thing that could happen is they get denied.. best thing an actually decent CTR, considering my images would actually stand out a little. My traffic ran against both types evenly, so no, the below douchebaggary did not compromise the scientific process above.

I won’t share the text, except to say the heading was simply along the lines of “Want a Free Apple iPad?” Well.. along the lines of.

I really have no idea whats going on here:

Hot girl. Apple laptop. What?

OK, this was the absolute winner. CTR above 0.25, clicks were 5c. It lasted less then 48 hours before they actually disaproved it. It is also the only one that got disaproved. On the second day the CTR dropped to about 0.10 when I unpaused it. I suspect people down voted it, so it got a worse position and thus less clicks.

Hot Ass + Apple = WIN

Dont really know whats going on here either:

Random hot girl. I think she’s using her iPad, looks like it to me, anyway.

iLame.

“Well La-De ****ing Da!”

She’s as sick of hearing about the friggin’ iPad as I am.

Feel free to use these in your campaigns! Protip: get these approved for a dating ad.

UPDATE 10/1: WFReview has a new owner, who is currently transferring the affiliate program from C2M to his own. If you get the plugin via this site you’ll get 25% off, but you’ll have to wait until the new affiliate program is live, hopefully in the next couple of days – I’ll post the discount coupon code then. Btw, the new owner is an insane programmer and the author of the Ruby Cookbook out on O’Reilly. His name is Lucas Carlson and the updates he is working on for the next version of WFReview are going to take it to a whole new level and then some. The update will include content spinning, themes, the ability to sort posts by price or overall ratings and thats not even a tenth of it. So get your drool on for this release.

There is a number of review site plugins for Wordpress out there, but none really match the power and flexibility of WFReview. WFReview works with CSV files, allowing you to easily import product info into a Wordpress install, typically for a review site, but there are other possible uses well, such as a Local Directory.

The plugin allows you to easily work with the feeds provided by networks such as Shareasale or LinkShare, which host hundreds of affiliate programs. If you’re a member of one of those networks you can login, grab the product feeds as CSV files and import them into your site, typically simply using the Product Description as the review. You can then use WFReview to generate comments for all the reviews, with the comments containing “customer ratings” for various features of the products you’re “reviewing”. For example the appearance, durability and feature set of a digital camera. You can upload as few or as many products as you like and you can even schedule them to be added to the site over time. Alternatives the plugin can set random dates from a given time period in the past to add a bit of perception shaping.

If you’re a member of the Amazon Associates program you probably know that Amazon doesn’t provide the kind of interface you get with say Shareasale, were you can simply save product lists/feeds as CSV files. Unless you have a scraping script or want to waste some time manually creating the CSV files you’re bit stuck.

Here’s a step by step solution if you want to be able to leverage the power of WFReview with Amazon products but dont have the programming knowledge to write your own scraping tool. Grab these tools and follow this process and you’re set, no technical knowledge required.

The no-brainer Amazon scraper you can use is called Optimal Product Finder (not an aff link, although the WFReview one is). Its not free, unfortunately and at $79 a bit overpriced. Use coupon code WS10 to get $10 off. I’m sure there are other similar tools out there, you may even find a free one. Or you could get someone to code one for you on Elance or something. If you’re planning on doing more scraping or any other tasks that can you can use a bot for I’d recommend you use the awesome Ubot Studio to build your scraper. This tool ain’t cheap, but in this case its easily worth it. It uses a Visual Scripting Language, so you need no programming knowledge whatsover to create your bots. This tool is insane. The price is $249, but you can use the coupon ‘muaffiliate‘ to get $50, so you’ll get it for $199.

For the rest of the post I’m going to assume you’re using Optimal Product Finder. Its the most basic solution requiring the least work.

You should already be a member of the Amazon Associates program and you’ll need to enter your ID into OPF. You’ll also need an AWS API Key and the Secret API Key from Amazon. You can get these by signing up for a Product Advertising account. Its free to join and you can just put Optimal Product Finder as the application you’ll use it for. You need this in order to access Amazon’s product datafeeds using their API. Grab the Keys from the “Manage Your Account” page and enter those Keys into OPF. OPF comes with an instructions PDF so I won’t go into what it does here, but there is one more step you need to do to get things rolling. Amazon change the IDs for the Browse Nodes in their datafeed from time to time. These are described in an XML file that OPF needs to load in to know which categories and subcategories to display in its menus. The XML file you’ll get initially is likely out of date.
I got the correct latest ID’s here. The OPF instructions give a second source, where you’ll have to pay a donation. I tried doing that, but didn’t find what I was looking for. Open up the RootBrowseNodes.xml file that came with OPF – its in the BrowseNodes directory – and copy and paste in any IDs that you notice are different.

You should now be set to grab your Amazon products. You may find your scrape returns thousands of products, which may take a while to scrape, so be patient. If thats not what you want you may want to try just getting the Best Sellers, Most Wished and Latest in the category you’re after. One approach is to go for volume, throwing in thousands of products and getting as many sites up as possible. Another is to get the the 10 best sellers and 10 most wished, write actual reviews of them (I’m a slow writer, so I prefer to save time and outsource) and legerage your now original content and ‘real’ revues to build trust.

Once you’ve got your data and saved it to a CSV (make sure you save it as a comma delimited CSV!) you may want to open up the file and check its all in order. The Description field often comes with a “Amazon Description” heading in it, which I prefer to remove. Add any Custom Fields for your posts here too – just add a column into the CSV named the same as the Custom Field for each one. Its pretty much good to go into your site as is though. So just go to the WFReview screen in your WP install and load in the CSV file. At this stage you’ll need some basic HTML knowledge to lay out the data in your review. Ie lay out the description (which is your ‘review’), the manufacturer, the image, model number etc how you want them to appear in your review post. You’ll probably want to add some buttons here too linking to Amazon with your affiliate link, something like “Read More”, “Buy Now” or “Visit Site Now”. I also add an Amazon logo under “Seller”. If you dont know how to throw down some div tags to format the fields or are feeling particularly lazy, but bought WFReview via this blog, contact me here and I’ll give you my template, as well as some buttons to use with it. Import the data, set up your Categories if you need to (the Manufacturer works fine as the Category, assuming you’re not reviewing a single brand), and you’re ready to generate some comments for your reviews. Set up the review parameters, eg Size, Stylishness and Functions, and press go. WFReview does a good job of guiding you through the process and it has its own support forums, but feel free to contact me if you need help with anything like setting up the cron job or even the whole process.

Thats about it. You should now have your blog populated with products, reviews and even some reader feedback. Promote it via your favorite method, rinse and repeat.

Oh, one more thing. If you also need a theme for your review site, a free one available is aWickedReview. The theme is by ‘radio’ from WickedFire and he is the author of the auction2post plugin (not an aff link, I just like the theme), which you can use it to create eBay stores on a Wordpress site. I have not used it, but have heard good things about. If you want to thank him for the free review theme, buy his plugin. Here’s what the theme looks like. Note that its currently limited to 5 products and if you need to add Pages (like TOS, Privacy, etc) you’ll need to hack the theme. I’ve actually added a second WP install in a sub directory when I used this theme for a review page, so that the main install has all the content and links straight to the review page from there.

So after yet another ‘hey, lets meet up’ discussion on WickedFire that was probably going to go nowhere in the real world I decided to start a Meetup202 group for all the Sydney guys and any other Australian affiliates. About bloody time – even New Zealand has had one for ages.

So if you’re a Sydney affiliate, jump in and join the group here: Meetup202-Sydney

It’ll be added to the Meetup202 page shortly. If you have no idea what Meetup202 is, have a read here.

Just don’t tell your DP friends about it.

I was testing out some campaigns on there, keeping the traffic at a trickle. An hour or so ago I decided to pause the 4 campaigns I had running as the tests were not going very well. I selected them all and clicked Paused. The interface changed to say “You have no Campaign ads currently running or pending review.”, as it should have.

So far so good.

A few minutes later I saw in Prosper202’s Spy view that traffic was still going to those campaigns. Hmm, perhaps it takes a few minutes for it to stop or its just residuals? Reasonable enough. I started to keep an eye on the traffic and what happened next was the exact opposite or what should have happened – the traffic began to increase to at least 2 of the campaigns. Then it increased some more and started flooding in, relative to what i was seeing before.

Not only did the campaigns not pause, which would have been not such a catastrophic fuck up, just a major one – it appears my targeting or my bidding was reset/wiped out/changed. I am not even sure the traffic was coming from the right countries.

I triple checked everything in Adbrite – there is no question all my campaigns were paused. I resumed the campaigns and then paused them again, hoping it would work the second time and tried to find a support email address. The Help page was no use whatsoever. The Contact Us page has only a ‘press inquiries’ address and for a link to a contact form, “in case you didn’t find an answer in the FAQ”. After filling out the contact form it followed up with “You will receive a response within 24 hours.”

I dont know if bigger spenders get an account rep or something, but I’ve never heard from a human being at Adbrite. The support system sucks – had the traffic not stopped was I going to have to wait up to 24 hours to get a response? I dont know yet if I’ll get billed for the traffic, in which I’ll need to chase up a refund or that was a a gift (hey it happens – in fact 3 days ago another platform ran for a few hours without recording impressions, so free traffic for all, thank you very much), but lets see what their support comes up with. I expect them to be as hopeless as their platform.

UPDATE: Well, a couple of days later and no reply from Adbrite. It doesn’t look like I was charged for the extra traffic though, so thanks for the extra 2 leads I got from it, Adbrite! Yeah, if the campaigns were doing better then that I would not have had to pause them in the first place. Wonder if I can replicate triggering Adbrite’s free traffic ‘feature’…

UPDATE 2: Got a reply shortly after the last update, telling me that it takes a while for the traffic to stop after you pause a campaign, after I already explained to them thats not what happened. Anyway, I’m way over bitching about Adbrite and much rather give MediaTraffic an earful right now, but then the blog cant all be about the sand in my vagina can it.

Turns out they are limiting impressions to ANY ads that are similar to other ads being shown, for the same reason that they limit your one same ad being shown to the same user too many times. Here’s a response from Facebook when someone from Affiliit.com asked them whether they ‘throttle’ dating ads. The response was posted on the Affiliit forums:

” In an effort to preserve a good user experience, our system monitors the number of times any given user sees a specific ad. If there is another ad that is the same or similar to that ad, our system takes that similarity into account. This similarity tracking can happen at any level (is independent of campaign or account). If the user things it’s the same when s/he sees it, then the system attempts to limit the ad(s)’ frequency.

What this means in your case is that your ads are probably not different enough from other, similar ads that your audience is seeing, or not different enough from ones that they have seen and found irrelevant or unpleasant. One reason the payout on certain dating offers is so high is that it’s difficult to drive traffic to them.

I recommend differentiating yourself in your current space (different offers, different text, different images, etc.), or changing spaces (e.g. Expanding internationally). If neither of these is possible, you will have to bid more to get the impressions, and therefore CTR that you need.”

On a side note, seems the facebook review doesn’t understand what CTR means, judging but his use of the term in the last sentence.

Affilliit.com, btw, is a new-ish Affiliate Marketing Training Program. I’ve just started having a bit of a look around, I’ll probably post a review at some point. I know the guys that started it back from my PPC Coach days and one of them gave me an account to check out what they’ve built over there and so far I’m impressed. They cover areas including PPC, SEO, Social Media and PPV, but their SEO is their strongest area and they have a huge range of SEO tools, reflecting that. Anyway, I’ll save the rest for a review some time.

I am rarely one for conspiracy theories or donning tinfoil hats but a word of warning, a lot of this post is conjecture. With that disclaimer out of the way, I would like to discuss a trend that I have noticed with the paid side of Google AdWords. I wrote recently about Google flushing out affiliates that use AdWords. A lot of people didn’t agree with me yet Google published an update just days afterward to it’s “website types to avoid” document about the exact topic. A lot of commenters made the very valid point that Google owns and operates it’s very own affiliate network.

The biggest revelation that was drawn from this and the subsequent thread over on WebmasterWorld was that Google was flushing out even the “highest quality” affiliates. They were even flushing out direct linking affiliates that were advertising with the parent merchants consent. I wont go into the rights and wrongs of all this, it’s been done to death on the WMW thread already and I do not want to discuss the whole Google Vs Affiliates topic again either. What I do want to point out is what happened next….

Interesting theory, will keep an eye out on how it pans out.

As for my situation, my older accounts seem to be in order, as are other accounts in my MCC.. Will be keeping an eye on that too. While throwing my campaigns at anything but Google.

So now onto the good news!

Got one of these in my email today:

Dear advertiser,

We are writing to let you know that your Google AdWords account has been disabled due to one or more serious violations of our advertising policies related to Landing Page and Site Quality. As a result, your ads will no longer run through the Google AdWords system and we are unable to accept advertising from you in the future. Please note that future accounts you open will also be disabled.

What ghastly things have I done to deserve this tragic fate? Well, the ONLY campaigns I EVER ran in that account were for 2 Amazon affiliate stores that I had written about earlier. One of those I had bought and know that the previous owner ran Adwords campaigns for (hey, I knew why he was getting rid of it and knew what I was buying, as I explained before) and the other is one that I made myself. I only ran the campaigns for a short time before they got ‘google slapped’, but I was going to turn them off anyway, as they were not profitable and I was doing better with organic traffic.

A google slap for a thin affiliate site I can understand. But a permanent ban from Google? Seems a little excessive. Again, there were never any other campaigns in that account, which I opened around the beginning of this year, but never used until August. I also have two other accounts, which do not appear to be banned and are connected by an MCC account, which includes client accounts (campaigns for local small businesses). Only the client campaigns are currently running in the MCC and I’ll have to keep an eye on whether they keep running. The only other campaign I’ve personally had running in those accounts was an old Clickbank campaign which I turned off about a month ago. Might see if I can reactivate that one. Another funny thing I noticed – a couple of campaigns that had been google slapped a year ago and got zero impressions on for months have been getting a trickle of impressions again after months of inactivity. I have no idea why. I think I’ll delete all those slapped campaigns anyhow, before I get these accounts banned also, if anything they’re the ones I would have expected to be hit for. I’m not sure I’d be able to create new campaigns in them anyway, will have to try. As you can tell, its not like I’ve been doing much with Adwords this whole year anyway, but this still complicates things for the future, if I cant run campaigns for local business clients.

One other reason for the ban hammer may have been that I didn’t have that account added to my MCC, which technically it should have been. Who knows. From the experience of others its useless trying to argue or even ask why the ban occured. Chasing it up will likely simply get my other accounts disabled, if their days aren’t numbered already.

Noone in the contest actually crossed the threshold to qualify for the first or second prize though (1500 leads), but a few of us got over the 3rd place threshold so we’re all getting a Nintendo Wii. So a big thanks to Direct Agents! It makes work a lot more interesting when you get rewarded with prizes for your efforts, so I highly recommend any affiliate marketers out there not doing so to give some contests a run – many networks or even AM blogs run them, for people of all levels. It can do wonders for your motivation.