by Kyle Grant
The use and validity of the post view conversion metric is a hotly debated issue in the online advertising space and, as with search, brings about the issue of attribution. Display is very different to search and therefore requires its own set of metrics. The major differences between search and display advertising can be analyzed by looking at the variation in the user intent across the different channels.
Question of intent
With a search ad, the user has directly entered into a search engine a query, a direct intention to find information. The direct expression of the intent to find information translates into a direct response-style of advertising and therefore can be measured using direct response metrics such as a click. Since a search ad is at the intersection of a user’s intention and the content they seek, a click to get to that content is a valid metric, as it is a response to a question or implicit intention to find information.
On the other hand, display ads typically appear on webpages where the user is engaged in activities that may or may not be related to the content in the ad. This does not imply that the ad is a wasted impression and that the message was not seen; but the expectation that the ad will result in a click is significantly lower. In order to garner a click, the display ad must be compelling enough to attract the users’ intent and interrupt their current train of thought, which is an extremely difficult task. Taking this one step further, should an ad be effective enough to divert the users’ intention and garner a click, the landing page must be equally as engaging in order to push the user through to conversion. If not, it will likely result in the user abandoning the conversion until a later time and resorting back to their previous intent.
Well documented studies state that clicks are inconsequential, random events in the display world with little-to-no bearing on conversions. A study by comScore found that only 16% of consumers ever click on ads. Even more striking in that study was that 8% of users account for 85% of all clicks. Upon further examination those clicking on ads were found to be the youngest and oldest users, and those with lower income(less than $40K annually – typically digitally less experienced). These users are not usually the target market for many advertising programs.
A study by Nielsen found that clicks and conversions have virtually no relationship, and that direct response metrics, such as CTR, fail to indicate whether an ad will impact consumers’ attitude or impact sales. Advertisers who embrace newer, lower funnel metrics and collect trustworthy data on effectiveness across multiple media platforms will be far better positioned and likely to separate themselves from their competitors.
Click fraud (“when a person, automated script or computer program imitates a legitimate user of a web browser clicking on an ad, for the purpose of generating a charge per click without having actual interest in the target of the ad’s link”) is rampant in the display advertising space. Although a great many steps forward have been made with regards to safe guarding advertisers with technologies (such as brand safety filters, viewability metrics, ad blocking, and third party ad verification), the fact remains that the majority of clicks are fraudulent. Knowing that the majority of clicks come from those outside of the target audience or are simply fraudulent, optimizing for post click conversions is actually counterproductive to a display ad campaign’s performance.
To optimize a display advertising campaign to align with the intended target market and drive maximum incremental ROI, advertisers are required to look deeper at the metrics and focus on driving conversions. By focusing on onsite conversion activities that are difficult for bots to complete, advertisers can start to hone in on the target audience and tactics that are driving valuable interactions with the site. However, given the fact that the majority of conversions are not going to happen post click, advertisers must optimize based on the post view conversion activity. Optimizing the campaign for sites, networks, and tactics that are driving real on-site interactions will subsequently remove fraudulent activities.
Attribution Models Required
While web analytics platforms are great at looking at post click engagement, the majority of analytics platforms are not adequately integrated with display advertising platforms and can therefore only track a user from the referring source directly prior to hitting the website. The impact for display advertising is that, unless the ad was clicked on, the majority of the conversions from display would appear to be from another channel; even though the interaction was initiated or influenced by a display ad impression, dramatically undervaluing the effectiveness of the display advertising programs.
Cross-channel attribution modelling is becoming a pre-requisite for being able to measure the true impact of online advertising programs. The last click model must give way to more advanced statistical models based off multiple interactions prior to purchase, including impressions. The weighting, attribution window, and time decay algorithms will need to be customized for each advertiser band based on their specific business model. To be effective, all display creative must be tagged with that information passed along to the attribution program. To further increase the attribution accuracy, viewability can be factored into the model, thereby ensuring that only those ads actually seen are given credit for impacting conversions.
Proven impacts of cross-channel display:
- Display advertising directly impacts SEO, PPC and direct load traffic. Combining display with SEO and PPC can amplify the cumulative effect of both programs. According to a Harvard Business School analysis, display advertising will impact search after a period of about 2 weeks.
- An Atlas study indicated a 4X lift in overall program when adding display to search.
- A comScore study indicated a 50% increase in branded search when using smart display buys.
While post clicks metrics are effective for search, the differing intention of a user interacting with a display ad requires a different set of metrics to evaluate success. The conclusions from copious amounts of research released on the subject are consistent: post view conversions must be embraced to recognize the full potential of display advertising programs. Advanced attribution models, capable of tracking both clicks and impressions, are required. Until these complex attribution models become ubiquitous and fully transparent, the combination of both post view and post click conversions is the only effective method of measuring the impact of a display advertising program.
About the Author
Kyle joined Mediative about 5 ½ years ago and has been working as a Digital Marketing Strategist since that time; although his official title is Performance Media Manager. Kyle’s passion is Digital Marketing Strategies and Integrated Marketing Communications. Although an expert in online Paid Advertising, Kyle seeks to look at the whole picture and welcomes the opportunity to talk holistic Digital Marketing Strategies. Kyle has worked with some of Mediative’s largest Paid Search accounts; managing accounts in excess of 1 million keywords and a monthly spend of over $1 million. He is able to combine his skills in data analysis and marketing ability to develop highly targeted ad messaging and campaigns, aiming to not only access the correct target market, but to speak to them with the right language. Although, his largest accounts have been in the B2B e-commerce space, Kyle has also successfully managed lead generation and branding campaigns in both the B2C and B2B markets. During his tenure at Mediative, Kyle has gained in-depth experience on a multitude of Analyt