Marketers have for years been struggling to determine how online display campaigns contribute to sales, conversions, and branding.
The industry has made great strides with research initiatives and attribution modeling to determine how online display, and also social media and other internet marketing channels, contribute to “lift.” Lift is a term many publishers use as an outcome to their research. What is lift will depend on your goals; lift in site traffic, lift in search impressions, or lift in conversions.
Back in the day, publishers would come to us and tell us that if our banner ad click-through rates were better than .02%, campaigns were successful. What? Rebook? We know better, don’t we?
Publishers and research companies are finally starting to know better as well. Where top marketers at most firms demand accountability and relevant statistics when determining media budget allocation, the communication is now shifting to branding KPIs when it comes to online display strategy. Consider the following:
3 months ago Google introduced Brand Activate, a new online measurement approach for online display advertisers, and a shift away from clicks and impressions. The idea is that with better metrics, marketers will feel more comfortable allocating more of their money to online advertising. “We believe that the industry’s significant investment in brand measurement efforts can substantially grow the online advertising pie, for all,” according to Google’s blog post back then. Google had introduced two new measures of Brand Activate: Active View and Active GRP.
Active View is a bid to become the new standard for online impressions. The proposal, now being submitted to the Media Rating Council, will count a “viewed” impression as one that “is at least 50% viewable on the screen for at least one second.” Active GRP is the online equivalent of the Gross Rating Point, a metric used by the broadcast industry to estimate how many people saw a given advertisement. Active GRP is a digital version that will calculate the reach and frequency of a campaign, but — unlike standard GRPs — lets advertisers react in real time.
Then today (July 2), media research powerhouse Nielsen announced the purchase of Vizu, a digital media research firm enabling advertisers and publishers to assess and optimize online advertising effectiveness. This acquisition will build on Nielsen’s online and cross-platform advertising campaign measurement and advertising solutions by adding real-time online effectiveness measurement. Nielsen will now offer real-time reporting of online advertising performance broken out by media plan channel, frequency of ad exposure, and advertising execution and targeting strategy. More details from their release:
The science (and art) of marketing attribution is another significant element in determining how display and other media channels contribute to the “lift” as mentioned above. Attribution is the attempt to give credit to a medium that contributed to a defined website conversion throughout a buyer’s purchase cycle, but was not the last medium viewed before conversion. So where the top two initiatives by Google and Nielsen focus purely on branding metrics, attribution relates to conversion metrics.
Traditional funnel metrics focuses on two areas:
- Top of funnel (impressions, views, clicks, opens)
- Bottom of funnel (sales, registrations, leads)
But what’s missing is the media activity that occurs in between first click and last-click conversion, where research and message influence occur. This is where the concept of attribution modeling comes in, to give these influential mid-funnel mediums partial credit for the conversion that had been traditionally credited to a last-visit source. By studying how channels work or don’t work together, marketers can see how channels, or groups of channels, assist one another.
An attribution model is a set of rules that determines how credit for sales and conversions is assigned to touch points in defined conversion paths. For example, “Last Interaction” attribution assigns 100% credit to the final touch points that immediately precede sales or conversions. “First Interaction” attribution assigns 100% credit to touch points that initiate conversion paths. Modeling defines what credit to give the steps in between.
Stepping back, it is an age old issue us former Media Director’s have always grappled with- the concept of allocating media funds toward a “sound” media mix supported by statistics and also “gut.” How many TV GRPs per week, and per month, builds sales lift, and what combination of newspaper, radio, and outdoor impressions within a set budget is “optimal.” The same questions come with digital planning, but because of cookie technology, we can better evaluate digital media mix. The caveat to any media strategy, and attribution modeling, is the power of message, seasonality, competition, ad positioning, and other external factors that affect the relationship between seeing a message, and the persuasiveness to purchase or registration.
Because of the increased demand marketers have on accountability, companies that are using attribution modeling are increasing. Assuming you have Google Analytics, attribution modeling is one of the key differences between the free Google Analytics, and the premium version of Google Analytics. Details here on the premium edition. The premium edition is $150,000 annually. Other analytics tools include some attribution modeling functionality as well.
There are limitations to the free edition of Google Analytics when it comes to attribution modeling and the multi-channel funnel reports.
- Assisted conversions and Last-Interaction conversions are all similarly valued
- Data between first-touch and last interaction is only 30 days. That means longer-buying cycles are not tracked. For Google, they are assuming the most important media activity before purchase or conversion occurs within 30 days from initial message exposure.
Google’s premium analytics allows 5 models of attribution. Below is from Google’s help section, and clearly explains each model-type option.
Buying Example: A customer finds your site by clicking one of your paid ads. She returns one week later by clicking over from a social network. That same day, she comes back a third time to buy — this time via one of your email campaigns.
In the “Last Interaction” attribution model, the last touch point — in this case the Email channel — would receive 100% of the credit for the sale.
In the “First Interaction” attribution model, the first touch point — in this case the Paid Advertising channel — would receive 100% of the credit for the sale.
In the “Linear” attribution model, each touch point in the conversion path — in this case the Paid Advertising, Social Network, and Email channels – share equal credit (33.3% each) for the sale.
In the “Time Decay” attribution model, the touch points closest in time to the sale or conversion get most of the credit. In this particular sale, the Email and Social Network channels would receive the most credit because the customer interacted with them on the day of the conversion. Since the Paid Advertising interaction occurred one week earlier, this channel would receive significantly less credit.
In the “Position Based” attribution model, 40% credit is assigned to the first interaction, 20% credit is assigned to the middle interactions, and 40% credit is assigned to the last interaction. The Paid and Email channels would each receive 40% credit, while the Social Network would receive 20% credit.
What does this all mean?
Google and other higher-end analytics tools is attempting to answer the question- how does online display, email, SEO, and social media contribute to conversions, and how should budget and resources be allocated based on the data?
By analyzing the data, some of the outcomes you’ll get:
- Where to reallocate budget towards channels most impactful to the final conversion
- Whether to update cost-per-acquisition metrics to reflect the true value of marketing channels
- Whether to reschedule timing of media campaigns based on time-to-conversion
- Whether to test various media mix models to optimize the best paths to conversion
- How social media marketing contributes to paid media channels that lead to conversions
Bottom line, we are in quite a transition in data analysis as we continue to try to measure the effect of mediums, traditional and digital, in the buying process, and the role online display and social media have in the final conversions process. There’s no perfect model, but at least we’re getting better at rationalizing the decisions we make. Do you measure attribution?