Performance Based Advertising, Part 2

Stephen Yagielowicz

In my first installment, we discussed how there are many different ways to advertise any product or service, and making sure that you get your money's worth is not always an easy (or simple) task. I also gave you a few ideas on ways that you can maximize your advertising budget - and your results. Today, we'll look at a few more considerations, as well as techniques for testing your approach's effectiveness.

A Shot in the Dark
At the other end of the spectrum from Pay Per Click (PPC), is the Pay For Placement (PFP) ad model, which often involves the un-quantified and unqualified selling of 'ad space' - usually for a set period of time. This is what happens in the traditional ad world, where you place an ad in a magazine or on television, etc. without any real way of gauging its effectiveness. Sure, you can usually receive some basic demographic and viewer / circulation figures from the media, and even employ a direct 'call to action' with a basic measurable response mechanism, such as "mention our ad for 10% off" as a means of gauging effectiveness, but at best these are educated guesses, rather than truly measurable results.

This is one of the biggest advantages of online advertising: you can easily count clicks, and then follow the prospect through the sales process, and beyond. Were I to purchase advertising for a pay site, I would not simply 'buy a slot' without some guarantee of performance. This performance could be measured through a reseller code and / or redirect URL that would track click-throughs, or I could simply purchase 'sign ups.'

Anyone selling advertising should at the very least show the advertiser the amount of clicks their ad received (and smart advertisers would track it on their end as well). Sell me a slot based on the number of impressions (good luck!) or run my ad based on the number of click-throughs, or run it based on the number of signups you send me. Give me 'something' that I can measure my performance against, not just 'you get the top slot for one month.'

When many of the big boys trade, buy, or sell traffic, their terms will often involve initial signups. For example, "Sponsor A" might send "Sponsor B" 100 signups, while "Sponsor B" will then send "Sponsor A" 100 back. A million 'hits' could exchange hands this way, but the currency of the realm is signups.

Back to advertising pay sites on TGPs. If I am the owner of a pay site, I had first better know how well 'typical' TGP traffic converts on my site. Let's use Dawn's pay site as an example. If I posted some of her galleries on an amateur TGP, and then received a decent signup rate, I might be interested in getting more traffic from that same TGP without having to submit more galleries. Based on my knowledge of how that particular source of traffic converts on this particular site, I would contact the TGP owner and offer him a deal, paying X$ per signup, and with a minimum commitment for signups, say, I'll buy 100 signups at $30 each (for example). He would then signup for her reseller program, and use the third-party tracked linking code to measure our progress. Once I have the 100 signups I paid for, I would renegotiate the deal based on the conversion ratios and retention rates that I received. The banner that produced the highest CTR (Click Through Rate) would be the one that I'd try using first on the new venue I wanted to advertise in.

Test Your Ads and the Media
Regardless of the actual advertising approach you use, testing should always be done both on the 'creative' (the actual banner, button, or text link), as well as on the media, but not both at the same time (if possible). In other words, if I was currently advertising on a web site using a banner ad, I would not try advertising on a different site, using a different banner until I had tested a number of ads on the first site. Changing only one variable at a time (either creative or the media) allows for a more valid result, since you will know which change produced the desired results.

If the media was 'online' (a web site, rather than a 'print' ad) where I was able to control the creative in 'real time,' I would try to send equal amounts of the same quality traffic to four different banners, each of which sent traffic to the same destination. The banner that produced the highest CTR (Click Through Rate) would be the one that I'd try using first on the new venue I wanted to advertise in. The worst performing ad would be scrapped, and a new design put into testing to replace it, thus ensuring the highest quality results from the best performing ads, and fresh creative to attract surfers who may become desensitized to 'stale' advertising that they've seen elsewhere.

As you can see, getting the most from your advertising budget is a trial and error process that requires testing for validation. The suggestions outlined here will give you a head start, but the results can never be predicted, as they are dependent on many factors. Try a few approaches, measure the results and your Return On Investment (ROI) to see what works best for you!
~ Stephen