The BizProfits Blog

So you’ve got a campaign, now what? (How to cut placements)

Written by: Kate Almog Promotion Tips

How to cut placements

So you picked a CPA offer, have registered at a traffic source, picked an offer, made a few banners, and they started receiving clicks… now what?
Common sense tells you that you should be doing something – optimizing, pausing, adding new campaigns – and you are absolutely right in thinking so. In this blog post, we’d like to talk about testing, choosing and cutting placements.

A quick reminder – what is a placement?

A placement is a website that the traffic source will put your banners on. At most traffic sources, you can choose such placements directly. Another option is running RON (Run of Network) campaigns, where you don’t really know where the traffic is coming from. Placements offer particular spots for banners, e.g. here is a description from TrafficJunky (The two highlighted spots are available for advertising on this website.)

How to cut placements

What should you do with placements?

Testing is an integral part of any affiliate marketing campaign, which means that’s what you should do with placements – test them! Some websites are better than others in terms of traffic and thus will bring you more profit.
It’s all very reasonable, really. What you need to do is identify which placements perform well and focus on them, drop the underperforming ones, and keep testing those that are in the middle.
Easier said than done, however. If you hoped that in AM you’d only have to count money, prepare to be disappointed – we’re getting down to statistics.

Why do you need statistical methods?

It seems that you can identify which placements are worth keeping without any measurements, but in truth – you can’t. There are too many questions – how many clicks do you need to make a decision? How much money should you spend on testing? Are you going to drop placements that will start performing just fine in a couple of days?

With a little statistics, you can predict rather accurately which placements are going to be the winners.

How to cut placements

Let’s start with statistical significance. In terms of campaigns, statistical significance means getting a similar result with multiple repetitions. Here is an example: let’s say you ran a campaign for a day and received zero conversions. You decided to continue and hit five conversions on the next day. To understand which one of these results is more likely to be consistent, you need to run the campaign for a few more days – that’s reaching statistical significance to make a decision.

In other words, you need to be sure that the results were not received by chance.

How to calculate statistical significance?

There are formulas to calculate statistical significance, but most experienced affiliates use offer payout to identify when to cut a placement.

  • If a placement has spent more than 1x offer payout with zero conversions, discontinue it.
  • If a placement has spent more than 2x offer payout and the ROI is negative, discontinue it.
  • If a placement has spent more than 5x offer payout and the ROI is over 30%, you have yourself a winner.
  • If you have an “in-between” situation, keep testing.

How to cut placements

In fact, there are self-made calculators that you can find and download on the STM forum. The best minds have created and shared them generously with everybody who wants to try. The benefit of using such a calculator (which is just an Excel file with appropriate cells where you can input data) is that you have a rather accurate prediction without getting into formulas.

Such tools calculate if the predicted minimum ROI of a certain placement will be higher than your desired ROI (a winner) and if the predicted maximum ROI of a certain placement will be lower than your desired ROI (a loser).

Then what?

Once you have identified winners and losers, there are two roads to take:

  • Blacklisting – blocking ID’s of placements that are unlikely to be profitable
  • Whitelisting – setting a campaign with only the best performing placements

How to cut placements

There isn’t much to explain with whitelisting, and when it comes to blacklisting, the advice is – don’t overdo it. Sometimes placements that performed badly for one banner or offer will do well for another one. Unless you block a placement full of bots (80% of bot traffic and more), there is some use in testing an under-performing placement with a different offer or a lower payout.

Again, sometimes you can stumble upon a list of “never again” placements that are made and shared by affiliates. But even in such cases some of the other affiliates find placements that work for their needs on such lists. It’s all too relative.

Our conclusion is not different from what it always is – keep testing. The methods you use will depend a lot on what offer and in what geo you are running, what creatives you use, etc. And if you don’t make a profit in the first few hours, don’t worry – success comes with time and practice. You only have to be persistent.

Start with selecting an offer from this range: (most of them have creatives provided). Contact your affiliate manager for details and consults.

And don’t forget – we are always here to support your efforts.

Share it:
Kate Almog

Affiliate Manager


Join our Network Today And Be Part Of a Money-Making Online Business!

If you would like to learn more about our CPA network, publisher/advertiser benefits or how to get started promoting with us, feel free to visit our FAQ page or get in touch by contacting us.