Those with lots of gray hair in this business (including the authors) will recall that the eminent dean of measurement, the ironically named David Poltrack, determined that 70% of the effectiveness of a TV ad was related to the creative content —even Google quotes it.
Get the creative narrative right, and you’re more than halfway there, regardless of the media strategy. Of course, in those days the targeting choices were very limited, as broadcast reach was highly concentrated. With limited channels and the Internet in the realm of DARPA researchers, it was easy to reach just about everyone but far more difficult to persuade those to whom the ad was particularly relevant. Targeting in those days was best performed in special interest magazine publishing (we miss Soap Opera Weekly).
Wrong Way
Fast forward to the Internet era. The data hustlers had a new war cry: creative/schmeative!
Targeting is king and creative is barely a second thought. (Enter the ubiquitous “belly fat” banners). If you build the data lake, the right audience will come. Ads will be 100% programmatically served against the data, in tiny standard Internet units. Click away!
Many things happened along the way that derailed the delusion. Cookies were thought to be a great vehicle for targeting audiences. However, with no way to capture the nuances that make humans who we are, it was discovered that 3rd Party cookies were inadequate and were considered to be far too invasive. As a result, they are being jettisoned due to privacy concerns and some of the leading browsers deprecating them shortly. However, people who deeply understand cookie efficacy for anything other than direct response are welcoming new approaches with open arms.
The notion of micro-targeting people through the funnel was an appealing, but oversold claim. Despite the promises, it was never fully possible.
Standards ads — to be kind — are sort of “miss-able”. Spray-and-pray approaches to standard ad units’ low engagement often don’t work. And how many of these ads are objects of fraud? 20% seems to be the recent thinking, which is rather astonishing! There is abundant evidence of bots clicking on standard ads, exaggerating the impact of any measure of consumer engagement.
Even more important, researchers like those at System 1 Group PLC long ago proved that creative is still the king for driving share of voice, which in turn drives share of market. Surprise them, delight them, and the sale will follow. People are a bundle of emotions, and appealing to our natural impulses is the key.
Show Me the Way
Where does this take us? At Undertone by Perion, we work closely with retailers and CPG enterprises who ask us for a simple goal: drive feet into physical stores and incremental sales, most typically through specialized promotional tactics which drive unit volume. We think we’ve cracked the code on balancing the unyielding necessity for creative supremacy, but also driving a better, truer way of personalization that yields powerful incremental results.
Undertone by Perion is widely known for its Commerce Media high returns on advertising spend (ROAS) — as high as 120x. The success comes from essentially a new “formula”, pioneered by Undertone, that fundamentally understands creative is still king, but data is the prime minister and AI is the government. Below is the “formula” for driving True Incremental Impact (“TII)”, and explanations of the variables.
TII=(1PData+1PContent+2Pdata+3Pdata)^(Generative Creative AI+AI Decisioning) + {High Impact Units) * CCD} + M
Where:
1P Data Retailer or CPG data from loyalty, purchases, site actions and the like
1P Content Retailer price, item, art and other promotional content
2P Data Undertone proprietary ad signals
3P Data Contextual data signals and third-party intent data that comes from data conglomerates, such as weather, location, and the like
Generative Creative AI Technology for infinite creative versioning for driving personalized advertising
AI Decisioning Technology for driving real-time offers and optimization
High Impact Units These are ad formats that are high quality, highly designed, structured and engineered to drive attention.
CCS Cross Channel Distribution, i.e., real time selection of the optimal device, format, and media type
M Measurement that is continually fed back into the optimizer algorithm to continuously improve campaign results based on required KPIs, such as foot traffic, awareness, sales lifts, site visits and the like
There is a lot going on in this equation, so let’s break it down.
We begin with data because personalization is key to future success. Without data, the high, true degree of personalization is just not possible. Furthermore, the data is not low-quality cookie data that created fake personalization, but extraordinarily high value 1PD (First Party Data) that retailers discovered is pure gold: loyalty and related purchase behavior, which creates enormous pattern recognition and subsequent marketing experimentation opportunities. Combining customer 1PD data, then adding retailer/CPG price/item/art/stock level variables creates the foundation stone for personalization. Building on the stone is proprietary 2P data, and finally adding well-honed 3P data such as editorial contextual, weather, location, and the like create the entire edifice ready for personalization. To be sure, this is not 1:1, but “1:to few” level, which when using well-defined segments drives extraordinarily high relevance and engagement.
Now let’s look at creative. Even before the use of highly personalized dynamic rendering, creative input was still measured to be 49% of an ad’s effectiveness.
The introduction of Generative AI in dynamic creative, combined with High Impact ad experiences, drives exponential improvements in campaign KPIs. The new ability to use AI tools to hyper leverage human creative talent and bring multiple creative ideas to life, along with using AI to personalize the creative for curated audiences, and keeping a high impact approach is simply a game changer in the industry. This is a new definition of what it means to be creative digitally.
Reader – don’t be fooled by claims of “high impact” that does not employ advanced dynamic tools to render audience-aligned creative in real time — driven by carefully selected signals. True high impact ads employ skilled designers, programmers with years of experience and technology, which then structures, designs and certifies websites to run the unique ad experiences. Using blurred backgrounds and employing primitive creative rendering capabilities not only insults professionals who grind over creating the real thing, it will limit ROI.
We cannot overstate the importance of true high impact ads in this equation. There are DSPs and retailers who attempt on and off site application of this theory, but with notably poor success. One important reason is that standard ads simply do not drive the necessary attention and emotional engagement that leads to activation, as well as the dynamic opportunities. Without advanced real-time creative rendering and alignment, the consumer ad experience will fall short as will campaign results.
The next step is to decide of course where the ad runs. It is well known that the device, the format, and the environment are critical elements to ad receptivity, and using multi-device and multi-format drives improved results. The advent of digital out-of-home (DOOH) complicated the situation, by adding some impressive versioning capabilities outdoor effectively. Not personalized of course, but particularly contextually relevant to retailers. In-store DOOH carries additional power.
Advertisers face critical choices on these allocation decisions. Advanced AI helps to select, monitor, and refine these decisions in real time. The adoption of honest and transparent incremental measurement analysis — not unsubstantiated claims — is critical.
This combination/equation is one of those pivotal/disruptive moments in the industry that will change the advertising frame. The evolution (revolution?) in advertising effectiveness is upon us. Advertising is a war for consumer attention while identifying transparent players in the competitive ecosystem.
Note these tactics are not limited to retail establishments. It is valuable to any entity that collects quality 1P data or sells through those with the described 1P data. That includes CPG companies, DTC providers, Quick Service Restaurants, Airlines, Auto Dealerships and more.
Things I’ve Learned Along the Way
Some things have become very clear as we’ve amassed those gray hairs, and a few are relevant here.
Retain the hero of the past: great, attention-getting creative — and let go of the villains — overstated claims and underwhelming standard ads. Then employ modern, data-based and AI tools to insure true ad effectiveness.
We look forward to helping the future take shape. Won’t you join us?
Daniel Aks is the President of Undertone, A Perion Company (NASDAQ & TASE: PERI), leveraging his broad operating experience in C-level roles for the information, education, & consumer media industries.
Paul Prior is Undertone by Perion’s COO, is a seasoned executive with 25 years experience in technology in general and 20 years of digital ad tech and media experience, focused on building innovative, high growth organizations.
Laura Salant is Undertone by Perion’s Chief Storyteller, providing full service intelligence to clients as head of UT’s Insights Consultancy.
