Some Notes on Affiliates & Referrer Marketing

I like referral programs and I love freebies. They are great ways to gather some customers at low cost and generate buzz. A friend was questioning me late last night on referrer / affiliate programs which got me thinking. The problem with most referral programs, especially with non-digital companies, is that they are not ranked against traditional marketing metrics.

Why don’t people provide give-aways to Tweeters?

I think there are a few reasons. As a marketer, if I know you are willing to pay me money for something then I’m going to take your money. The role of a company, primarily, is to maximize profits.

Secondly, Twitter is still niche; although one can argue it is getting more and more mainstream. It is a dangerous precedent to give something away now, only to have the community expand and expect freebies. If not done well it can destroy the mental value of your product. See my post on decision anchoring. Customers unwilling to pay are not always worth keeping.

Finally, there is an implicit, gross cost to giving something away (gross profit is the money you earn, minus the cost of making and providing the profit). Let’s assume the normal retail price of my widget is $50 and my gross profit is $25, meaning a gross expense of $25. That means each time I give something away you need to refer at least one customer for me to break even and three for a profit. Sounds like a lot more work, doesn’t it? Also keep in mind that gross profit margins in many industries struggle to hit 50%.

What I’m not taking into account here is viral buzz but be realistic. Launching a vanilla ‘take a freebie’ will not get you buzz. If you be creative, that’s a different story.

What about paid referrals?

Referrals I do support, but people rarely put solid thought into how to price or structure their referral scheme. When judging referrals I need to take into account net profits over gross profits.

The first thing to do is figure out our acquisition (marketing) cost per customer; in other words, how much does it cost me to get their money?. Let’s assume that $8 of the above $50 is a marketing expense and that every time I spend $8 I find a new customer (the model is repeatable). In the real world I’d have a specific acquisition costs for each channel.

In my experience, because of propagation, referral customers can also have higher than average attrition rates. So, perhaps we should  be giving $6 per customer, not $8, to be safe. $6 makes sense, I am lowering my cost of acquisition so some cannibalization is okay (though, we don’t want to cannibalize so much that higher volume channels are now ineffective).

We also need to remember that referral sales can cannibalize other channels, if I had 100 referral sales it would be safe to assume at least 30 would have been sent by their friends whether I had a referral system or not.

Referrals are great when you don’t have the money to spend on promoting. But, as budgets increase you need to be careful that referrals don’t become too expensive. You should always be lowering cost of acquisition, not increasing it.

Thoughts on bulk discounts in referral schemes

So why don’t we structure a referral system that gives me 20% off for each customer I refer. Again, in the above example my customer pays $50, so I would give $10 off when they refer a customer.Compare bulk discounts to relative costs of other channels; and then look at volume. The referral scheme again is too expensive here. 10% off works, but what happens when I refer a second customer?

Referrals are easier on the Internet

If I buy and consume a product online, it is cheap and efficient to measure referrals. If I’m buying offline, it is harder and the cost of measurement increases; a cost which needs to be deducted from referral fees. Most web startups also take little capital so referrals are a fast way to grow. Larger, traditional organizations find referrals difficult to measure and have the capital to focus on stronger ROI channels.

Fighting for referrers

Most companies do not yet offer fully evolved referral programs nor are they heavily promoted. As measurement becomes cheaper and existing channels become saturated there will be a stronger fight for referrers (those who recruit their friends or promote products for a fee).