The Funnel & Performance Based Advertising (advanced)

Agencies and advertisers like simplicity. They like to pay a fixed fee for a conversion. Either doing performance based advertising on desktop/laptop, tablet or mobile inventory; it’s all about that fixed fee (stimulated by the well know last click attribution model). Isn’t the customer journey more complex than clicking on a banner and buying a product? Of course it is.

Marketers like the word ‘touchpoint’ for a reason; there are many points a consumer touches before committing to an online or offline purchase. A touchpoint can be a TV commercial, a friend with a recommendation, an Out-Of-Home ad, an ad in print or; a display banner on a mobile device or computer. It’s that first touch that implants a desire to purchase/own a product.

This raises the question as to why the very short sales-cycle, applicable in Performance Based campaigns, exists. Why is a sale-cycle of a person seeing an ad for the first time and immediately buying the product - by clicking on the ad - profitable?  There is more to it.

The funnel

When talking about the sales funnel (or customer-journey), generally speaking it can be described like this; - a lot of people out there, some people learn about you, then some less people become interested, and finally some people buy your product -.

The funnel has different stages that require different tools. For example:

Stage 1 (People learning of you/your product) has a generic character that requires a broad tool like a TV commercial, an Out of Home Ad, an ad in print or a (targeted) display banner on a mobile device or computer. Even if your audience is very specific; a broad campaign targeted on your audience is an accepted starting point if you want to scale*.

*Scaling: It’s possible to make sure you can be found if people search for your product (in this scenario solely basing revenue on the fact that customers will search for your product). However; if you want scale; you have more audiences that don’t know they want to buy you (yet).

Stage 2 (People become interested) usually speaking, relies on tool that has a repeating character. And here is where it becomes vague; this tool can also be TV commercial, an Out of Home Ad, an ad in print or a display banner on a mobile device or computer. Furthermore, retargeting comes into play. In this scenario, a consumer becomes interested and might visit the website of your product. By serving ads that visually repeats the offer or product that has been checked out; a consumers’ interest can grow.

Then there is Stage 3 (a consumer buying your product). Luckily stage 3 has some concrete tools that make it happen: when a consumer has learned about your product (Stage 1) and becomes interested (stage 2); usually speaking he or she uses a search engine to either gather more information before committing to a buy or to search the site (as lazy as most of us are by not remembering the exact URL - if visited the site before). This is where the marketing tool Search comes into play; by serving SEA or SEO initiated results (in the search engine) the consumer can commit to the purchase. Secondarily a normal programmatic display advertising campaign (for example on mobile devices) might cause a conversion. It’s worth mentioning that Retargeting is also known to cause conversions. So even though stage 3 is much more linear, there is still an overlap in tools that create the actual conversion.

You could say that, when trying to achieve conversions by using a single tool and focusing on an immediate purchase, you are not at all utilizing the potential the complete funnel has to offer.

So you could say that when trying to achieve conversions by using a single tool and focusing on an immediate purchase, you are not at all utilizing the potential the complete funnel has to offer

Last click attribution vs view attribution

The most simplistic and popular way to measure performance and attribute a merit to the tools (and companies behind them) is ‘last click attribution’. In case of Programmatic Display Advertising; the last click on an ad – that leads to the purchase - gets the merit.

The first problem this evident situation creates, is that the technique is susceptible for fraud. Many has been written about this subject. Simply put fraudulent tools that generate enormous amounts of clicks, will try (and succeed with) getting the merit.

The second problem is the fact that the merit only gets paid to the company serving the last ad. This does not comply with the full sales funnel, the parties involved and the influence their marketing tool had on the decision of the consumer to buy the actual product.

The key to performance Based Advertising is ensure your top of the funnel (stage 1) is effective and measurable. TV-Ads don’t allow you to do this, nor does Out of Home Advertising or ads in print. Programmatic display advertising does. It allows you to measure if your feeding of the funnel is effective throughout stages 1, 2 and 3.

If you use the same KPIs and attribution metrics to measure effectiveness of top of the funnel and bottom of the funnel tools, you will certainly say that the bottom of the funnel (stage 3) is performing well – as it seems to be responsible for the conversions - while the top of your funnel (stage 1) is not. And you will probably stop marketing activities for stage 1. As a result – most likely – you cannot scale.

You want to pay merit to not just vendors that provide clicks, but views as well. If the Device ID (or IP) behind the purchase saw a banner in stage 1, the vendor responsible should get paid. And here is the good thing; with programmatic advertising, this is possible. If you want to maintain a healthy stream of revenue you should give merit to all touchpoints that contributed to the conversion. With view attribution models, this is possible.

Secondary results of your Programmatic Display Campaign IMAGE

It is known that if you initiate a big TV campaign or a similar big programmatic display campaign, the results of your search and retargeting campaign will drastically improve. With Programmatic display campaigns however; you can measure this.

It’s also worth noticing that serving display ads on mobile and computer devices is most scalable; simply because it’s inventory is bigger than reach of TV, OOH, or ads in print.

When deploying a Programmatic Display Advertising campaign (e.g. on mobile phones) there is a second, less familiar, attribution taking place.

Instead of looking for a low CPA to begin with, one should study the secondary effect of the CPA campaign itself. The easiest way to do this is to activate and de-activate the campaign with big time spans in between. For example;

·     Month 1; on

·     Month 2; off

·     Month 3; on

·     Month 4; off

In many cases the conversions via Search, retargeting or even Social are significantly higher when the Programmatic Display campaign is active (on). These secondary results are (still) often missed by modern tracking tools.

It’s worth your while measuring a CPA, but you should know which attribution logic you use. Whether you will pay someone on CPA or not is less important. The important thing is that your overall campaign strategy (including top and bottom of the funnel) delivers a good ROI. A performance campaign should be measured across all layers and players, not separately one by one.

In many cases it’s irrelevant if the last click CPA of the display is not profitable. The important thing is that overall, you got more purchases by feeding the top of the funnel – when running a Programmatic Display campaign.

Using the full sales funnel

As a company we highly recommend analysing and working with the full sales funnel - using view attribution models. The fact that Programmatic Display Advertising is most scalable and is able to use view attribution models makes it a very powerfull marketingtool. With the right knowledge and technology our clients are able to utilize the full sales funnel.

www.targetoo.com