Life Lessons after 4 years as a CPA advertiser

Great thread so far.

What about chargebacks and Paypal? Do the same CB rules apply to Paypal as well (below 1% and you're fine)? Do they differentiate between legit chargebacks (customer didn't receive the product) and false chargebacks e.g. customer received the product (you provided PP with the tracking numbers) but he is unhappy with the product and decides to open a dispute without you giving any money-back guarantee.

I heard that Paypal likes to freeze your money or forces you to build huge security reserves. Did that happen to you yet?

I've had a single site running off PayPal for about 12 months and can answer that for you - PayPal is much more lenient about letting you keep your accounts, but in return they won't hesitate to slap HUGE reserve holds on them. There was a 3-4 month period where I had 6%+ and all they did was increase the reserve while keeping the account open, and as I was already running at a large enough margin to not need the cash immediately, I didn't really care.

I never ran trials with PayPal though (not even sure it's possible with their system? is it?) but I did have standard recurring (ex. $50 + $50 + $50 etc). I didn't actively try to hide the terms or anything ultra-shady like that, so it's interesting that the rate was so high but hopefully that answers your question at least.
 


When you say typical CPM buy do you mean buying on a network that doesn't operate as an RTB or do you mean buying direct from a site for a set CPM rate?


both actually, since either one would typically be a "flat rate" vs real time bidding.

RTB is more granular, and since you can pay a different rate for every single click, you can more finely tune demographics. So, if you know that early-morning female visitors from a recipe page on Yahoo convert at a 50% higher rate than the "average" clickthru from a gneric Run Of Site banner ad on a Yahoo portal, than you can adjust your bids accordingly and be willing to spend more to more accurately target that specific traffic, rather than a "one size fits all"approach to buying 100k impressions of generic Yahoo traffic all at the same CPM rate.

FB traffic is another area where data can help you out, since you can buy FB traffic on either a CPC or a CPM (on views) basis, so if you know what works best for your offer (could be either, just have to test everything) than you can spend less on your traffic than you would using the "other" method.

NOTE: I should point out that I'm far from an expert on media buys, I dabble in it and know my way around it a bit, but its never been a core strength of mine, there are plenty of peeps who know that game better than I.
 
Interestingly enough, the more customers spend, the less likely they are to return for refund. I've tested this across 15 diff product lines, and people that order the cheapest/smallest are always much more likely to return than the ones that spend more. Every time we have raised prices on any of our product lines, our return rate has gone down

How did the price increase effect your conversion rates? Did the conversion rates drop but the increase in price offset the drop in conversions or did you see conversions increase along with price increases?

Secondly did you change up your landers or USP to cater to a different audience to match the higher prices?
 
I like where this discussion is headed.

To address Amoxicillin's questions, Yes, you can (and probably should) add more transactions to offset/dilute your higher-chargeback traffic, and try to keep that overall rate under the magic 1% rate. It's not quite as simple as suggested though, because the thing you absolutely don't want to end up doing is adding these new, non-profitable transactions to dilute your overall chargeback rate, and end up getting new chargebacks on these transactions, which can happen pretty easily by accident.

The "micro-payment" examples touched on above, where you run a bunch of transactions through for small amounts to get your count up can and does help if done properly. (I should mention here that the "CB %" flag isn't only done on total txn count, but dollar volume can be watched too, so 50 transactions of a dollar each aren't exactly the same as 50 txns of $80 each, but most banks/ISOs don't really notice this, yet, so it works for now.). Running a bunch of these cheapy transactions of a different merchant's/product's can bite you in the ass though, because your credit card descriptor may confuse the customers, so they end up charging back that $1 transaction because they don't recognize it, which really stings. Now, instead of diluting your CB ratio, you are making it worse, on txns that make you no money and cost more in chargeback fees (typically $25 a pop) than you could ever hope to make off them, so its a double-whammy.

Say your product is called "Diet Berriez 360", it does 1000 txns a month,and it has a 2% chargeback rate on its volume. Ideally, you want to keep getting those same 1000 (profitable) txns per month, but you also want to add 1000 new "low-risk" (read: non-profitable) txns to your volume to dilute that overall monthly CB rate down to 1%. You hit up your buddy Joe with a gaming site, SuperMario360,com, and tell him you want to have him push a $1 (loss-leader) gaming tips membership or ebook. He says great, he sells it on the site to 1000 of his customers at a great value(for no profit, probably even a loss) and runs the 1k new transactions of a buck apiece through your merchant account. Great, now your chargeback rate is down to 1%, right? Wrong..

Even if all the customers of the video game site love the $1 ebook/membership, some percentage of them will see that line on their credit card statement that says "Diet Berriez 360" and "WTF, my skinny ass dont need no diet, I'm charging that back" and they do, probably at a rate of greater than 2% even over a buck. It's pretty near impossible for an Advertiser to change their merchant descriptor back and forth (plus you can get slapped with an instant 20k fine from Visa for having a "misleading CC descriptor, which really sucks). Now, instead of diluting your CBs, you have compounded the problem, on sales that have made you no money.

The way to do it best, in my opinion, is through some kind of cross-sell/up-sell of a same or similar product that fits under the same CC descriptor and brand-name, and that is such a clear value that no customer would ever charge it back (some will regardless, but you just have to deal with it). Usually these are add-on products, such as wrinkle remover to add to a skin cream offer, or a jump-rope to a diet pill, anything that the customer perceives as adding plenty of value for just a tiny cost. You will never make a profit off these add-on products, but you could save your merchant account. It's good customer service/karma too, work to add value in their eyes and you are less likely to end up in a bad spot in the rebill game to begin with.

I know Advertisers that charge the product cost and its shipping cost as 2 separate txns, just to up their txn count. That can blow up in your face too though, because a customer angry about the rebill may also chargeback the other shipping charge too, giving you 2 CBs instead of 1. Some merchants "split" larger transactions into multiple smaller ones, like turning a $90 rebill into 1 $50 txn and 2 $20 ones. I dont personally do this, I know peeps who do, and I think it's a dangerous game. Strictly speaking, Visa has new rules that technically prohibit this, but its awful hard to catch. You still have the risk of an angry customer getting multiple CBs against you off the same transaction, plus pissing off Visa, so I recommend against it. If you are going to split larger $$ amount purchases into multi smaller txns, be sure to space them apart as much as possible. First, customer is less likely to notice 5 charges from the same merchant one after the other, Second, banks are less likely to process a later CB if they see the customer has had multiple historical "purchases" over a period of time from that same merchant, its hard for the to say "fraud, never ordered" when they have other transactions from months before that they never complained about.

Just to add a little more food for thought to this discussion:

1) Though CPA offers might not be immediately associated with good customer service, having a phone number with people who are trained to help your customers can drive down your CB rate. Put your phone number on all your customer communication, put it on your product, and remind customers to call if there's a problem. When they call, treat them well. That doesn't mean you can't continue to sell to them, but don't leave them on hold for 20 minutes or make it impossible for them to cancel. You'd be amazed what the power of persuasion and the promise of a couple of gifts can do to change people's mind about your product.

2) Offer a refund policy. If customers see there is a way to get money back from you, they are far less likely to charge back. Sure, it sucks to give up a sale, but you can get extremely creative about ways to recoup the money that you lose to these types of calls and, honestly, it is way better to send a customer their $50, $60, or $70 dollars than to suck up another CB.

3) Do not attempt to deceive your customers, period. Yeah, most people don't read the terms on free offers and that's why they get sucked in, but that doesn't mean you should actively deceive them. Hiding your terms, creating fake testimonials, or doing other shady shit will only get your ass sued and shut down faster than you can imagine. If you run a CPA offer, you need to be ready to face at least a few lawsuits and, in order to win, you need to be 100% compliant. Don't fuck around, it isn't worth it and plenty of smarter people have lost everything because they cut a corner or two.

4) Learn from the big boys. There are a few big CPA companies out there right now. Find out what they're doing. Order their products. Go through their cancellation process. Get a refund. Look at their ads, their landers, and their emails. Pay attention to their messaging. While CPA companies might not be around forever, there are a few (very few) that seem to have positioned themselves for the long-haul. Follow their progress and try to guess where they're going.
 
How did the price increase effect your conversion rates? Did the conversion rates drop but the increase in price offset the drop in conversions or did you see conversions increase along with price increases?

Secondly did you change up your landers or USP to cater to a different audience to match the higher prices?


I never change any other variable besides price at once, it skews the data. As to which price is "better" for total return, it varies. I always have to test, times I've thought a price increase would kill my sales I've been 100% wrong (total sales increased), other times when I thought customers wouldn't notice a price increase, I've seen overall revenue fall, you just have to test. Overall, my additional revenues and decreased hassle from higher prices tend to win out over lower pricing, up to a certain point.

Generally speaking, if there are few/no easy alternatives to your product, you can get away with raising prices, whereas if your product category represents an entire aisle at the local Costco, your customers are much more price-sensitive. For a good read about the psychology of pricing decisions, check out the book "1% Windfall" by Rafi Mohammed, its a good insight into the psychology and market forces behind successful pricing strategies.
 
Got nothing to add, but just wanted to say this thread is the shit. Thanks for the free advice, appreciate it.