Advertising on Plentyoffish is some of the easiest money you’ll make in affiliate marketing. I’ve used the platform for over 2 years with varying degrees of success. Some days it seems so easy, other days I couldn’t grasp profitability if it slapped me in the face with a briefcase full of Benjamins.

As with any traffic source, or any advertising campaign, good things come to those who show patience – and especially those who aren’t afraid of losing money in the pursuit of much more.

I’ve traded tips and techniques with many Plentyoffish advertisers, and one of the recurring questions that pops up is how to combat unpredictable conversion rates.

Have you been there before? One day your ROI is impressive enough to book a spontaneous vacation, and the next you’re struggling to break even. You spend your morning coffee praying the marketing deities have woken up on your side; afternoon is spent with fingers crossed, legs crossed, and eventually eyes crossed having obtained a miserable headache.

Okay, if you haven’t been there already, you will do someday. It happens to all of us.

So how can we get some consistency in our conversion rates? How can we find the sweet spot where our ROI trajectory doesn’t resemble the Big Dipper?

NOTE: Before continuing, I have to ram this piece of advice down your throat: split test your offer across multiple affiliate networks. And even split test the offer. I couldn’t possibly overstate the importance of doing this – it can reverse a negative ROI overnight. The lack of loyalty will not endear you to your affiliate managers, but you’re in this to make money, right? Don’t become a network martyr.

An easy way to stabilize conversion rates is to tighten up your campaign’s message. Many Plentyoffish advertisers have an obsession with abstract marketing. They try gaming user attributes to attract a high clickthrough rate, often at the expense of a solid advertising message.

For example, let’s say an advertiser is promoting a typical CPA dating offer. In order to create relevance and catch the eye, his ad may read something like “More Single Smokers Needed, Join Site X!“, with the ad being targeted to single smokers.

These ads are rolled out as frequently as rat meat at a McDonalds drive-thru, but they lack a real message. Does Site X really appeal to single smokers? Is there actually a resemblance of a connection between the ad and the service being promoted? When there isn’t, conversion rates are likely to be erratic.

But why is that so? If the ad was competent enough to convert yesterday, it should work today, right?

The reality is you could advertise a dating service using just about any headline under the sun and it WILL score some conversions. Why? Because it’s appearing on a dating website! There’s no science to poor marketing. It is naturally erratic. The handful of conversions are a by-product, not an endorsement of your advertising work.

To achieve stable conversion rates over a sustained period of time, you need a stable message. Something that adds value to the sales funnel, rather than simply monopolizing user attributes. You need to pinpoint what the unique value of your offer is, create a meaningful message, and then rely on the powers of persuasion to convince your audience that the offer is right for them. This is only possible if you engage in real-world marketing, using hooks of genuine value. Attempting to lure smokers to a dating site under the illusion that they are ‘needed to meet the demand for more smokers’ is a shoddy hook – fitting of the erratic conversion rates it will likely produce.

Creating consistency and value in your message is one way to troubleshoot a wobbly conversion rate. What other steps can we take?

How about taking a structured approach to our testing?

A structured approach is definitely NOT this:

Day 1: bidding 0.55 CPM, 300 login count, frequency cap of 4
Day 2: bidding 0.65 CPM, 400 login count, frequency cap of 3
Day 3: bidding 0.85 CPM, 100 login count, frequency cap of 6, oh and a different landing page.

What happens when your conversion rate drops dramatically on Day 3?

Pick your reason from any of the following:

– The high CPM doesn’t convert.
– Low login counts suck.
– Your new landing page is hideous.
– Ads that have been showing for 3 days in a row stop converting.
– If the user doesn’t click in the first 3 views, he won’t convert.
– Day 3 was Christmas.

The actual reason? Who the hell knows?!

It’s impossible to diagnose a faltering conversion rate when you don’t have a structured approach to your campaign. So instead of fiddling with targeting parameters every 5 minutes, duplicate any campaign you wish to modify and then compare the results to your original campaign (which should still be running). Without controlling your variables, any data you collect is meaningless – and thus very expensive.

A final tip for stabilizing conversions is to limit the range of users you’re advertising to.

Ben has previously revealed on the POF company blog that 28% of the site’s inventory has a login count of less than 100. These users are the ‘fresh’ eyeballs. Another 30% of the traffic has a login count of over 550. These are the users who have seen pretty much every trick in the book. It’s going to take an innovative campaign to light a firework up their asses, but fear not, it can be done.

Between 100-550 logins you have the middle ground. It’s difficult to predict the behaviour of these users as the effects of banner blindness could swing either way.

My advice is to pick one or the other. Either target the first 100 login counts, or exclude them. Some advertisers take the middle ground and attempt to advertise to all users with a login count of less than, say, 300.

The problem with doing so is that it muddles your message by targeting two slightly different demographics – those who are new to Plentyoffish, and those who aren’t. Sometimes you can get away with it, other times you can’t. A better tactic, if you’re intent on targeting logins <300 would be to break the users in to two groups. A smart strategy would include two campaigns:

Campaign A: Targets login count of 0-100 (Strictly new users)
Campaign B: Targets login count of 101-300 (Users who have been around a while)

A more unstable approach is to lump those users together, so let’s have a Campaign C.

Campaign C: Targets login count of 0-300 (A melting pot of new and old users)

The difference in performance may not be obvious until a natural variation in the bidding on the platform.

Perhaps a swarm of new advertisers drives up the prices for the best quality traffic, so the volume in Campaign A drops dramatically – but the conversions are still stable. You’re left with an excellent conversion rate but not much traffic. Not a particularly alarming problem on its own, right?

Now consider what happens to Campaign B during this surge. The new advertisers may be rushing to grab the best quality traffic (perhaps yet another blogger recommended low logins FTW), but they don’t seem to care for inventory with a login count over 100. What then? Campaign B remains exactly the same. The conversion rate is stable like it always was. It’s a decent conversion rate, but not as good as Campaign A. The volume hasn’t changed though.

But let’s say you were to advertise using the Campaign C approach. You now find that you’re barely breaking even and the conversion rate has dropped alarmingly in the last 24 hours. How are you going to tell where the discrepancy came from? Why is the conversion rate on a seemingly terminal decline?

Of course you don’t notice that behind the scenes, you’re receiving less of the alleged best quality clicks because of the increase in competition on 0-100 logins. Instead, your high CPM is now paying for the next tier of traffic, which doesn’t convert as well but is costing you the same money. All you see is an unstable conversion rate, keeping you blind to the source of the problem – a problem that is very simple to fix.

This is just one of many problems that are notorious on self-serve advertising platforms. Different advertisers may swarm a demographic at any given time, which raises the prices, meaning that without knowing it, you’re suddenly bidding on completely different traffic. The only way to avoid the problem is to make sure that you’re bidding on just one demographic at a time, and that you’re setting the bid accordingly.

By doing so, your volume may vary dramatically, but the conversion rate should remain stable.

Of course, this is not science. Self-serve platforms by their very nature have hundreds of scenarios that may affect your ROI. But the bottom line remains the same. There’s big money on Plentyoffish, and it’s perfectly within your reach – but you must use your data as an advantage, rather than a deep pool to drown in.

Recommended This Week

• A much more detailed assault on monetizing Plentyoffish is covered in Volumes 1 and 3 of Premium Posts, which have both received widespread praise. Grab your copies now. Also, watch out for Volume 4 which will be landing next month and covering some brand new topics that I think you’re going to enjoy.

• If you’re a new reader, please add me to your RSS. Also follow me on Twitter. Thanks for reading.