My foray into understanding, and more importantly, applying Artificial Intelligence (AI) to current business practices has forced me to throw everything I ever knew out the window. Experience, best practices, preconceived notions, the tried and true – all these rules become obsolete and irrelevant because the world has changed. Media consumption has changed. Business has lost control over the customer and is struggling to gain and maintain their attention.
Display advertising has had it’s day
The display ad model, in its current form, is not sustainable. Current stats on global digital performance across ad formats reveals the following:
Across all ad formats and placements Ad CTR is just 0.05%
Rich media Ad CTR is 0.1%
Facebook and Google continue to dominate… but they only tell part of the story
In comparison, a recent study from Wordstream indicated: Facebook ads seem to be thriving:
…across sectors clickthrough rates (CTRs) vary from around 0.5% through to 1.6%.
…on average, Google AdWords advertisers are seeing conversion rates of 2.70% on the Search network, and 0.89% on the Display network.
It is apparent Facebook and Google have reaped the greatest benefit from digital ad spend. Last year, over $17.5 billion was spent on digital advertising. But FB and Google walked away with over 65% of estimated revenues, with Google garnering $30 billion vs. Facebook’s $8 billion in ad revenue.
Despite this, these platforms are not without their hurdles. In March, brands like Unilever and agencies like Havas chose to freeze Google and Youtube spending because of ad placement beside “undesirable or unsafe content”. This, on top of the questionable reporting on viewability, and the rising incidences of ad fraud are making brands and agencies alike become more cautious about how they spend.
Consumers are become more immune to ads:
The rise of apps like Ghostery, to detect and block tracking technology, has made things more challenging for publishers and advertisers alike. This latest report on adblocking reported the following:
- 615 million devices now use adblock
- 11% of the global population are blocking ads on the web, a rise of 30% in 2016
- Mobile adblocking grew to 380 million devices, a rise of 250%
- According to this report, adblocking is now mainstream across age groups
The impact on the publishing industry is staggering: by 2020, revenue losses at $35 billion are estimated assuming the rate of adoption continues.
How can we effectively reach target customers when ad blockers, changing media consumption and ad performance are impeding this?
However, the holy grail of truly delivering personalization at scale – and doing that consistently – is hard. Unless you have the full journey from the point of acquisition to delivering value to your tenured customers, your efforts are fragmented at best.
Google knows the score: “How Google Plans to Kill Last Click Attribution”
Google announced this recently. What the giant has realized is that knowing what ads works can’t be done by measuring performance in aggregate. The reason they’ve moved to conversion metrics (CV) is that the Click-through rate (CTR) is a misnomer. It’s no longer a measure of true intent. How you measure intent is not an aggregation of behaviors by ad format (yes, I’m simplifying). Rather, it’s by understanding the events in the buying funnel that attribute to the buying behavior. And here’s our introduction to Artificial Intelligence and why it will be the next evolution in the journey for the CMO.
If you want to improve performance, you need to change your mindset to embrace what’s new in this space:
1) Business Intelligence does NOT have the speed and power of Artificial Intelligence:
The time it takes to develop the right algorithm that works for the business can be a grind. It takes multiple testing and continuous iteration to improve accuracy of results. The human effort to do this can take months. What machine learning has done is automate predictive analytics and allow models to go into production much sooner than traditional business intelligence (BI) models. As new data is ingested, the models “learn” and adjust. This continuous feedback loop allows for greater performance in a much shorter time period with much better accuracy.
2) Resist the urge to make assumptions:
What we have known to work can be thrown out the door. Understanding consumer intent means dispelling KPI’s and known variables that were indicators of performance. I’ve been told time and again to let go of what I’ve known to be true. AI has no pre-conceived biases so the models will find patterns among the events that correlate to the intended business outcomes. It will either validate what we’ve already known or surface entirely new results that would not have been found through human analysis.
Traditional business intelligence has its merits. But in an environment where increasing competition, speed and the need for accuracy is required, this practice is also limited in its ability to be nimble and scale as new norms are introduced. More importantly, we may be hindered by our own blindspots by allowing what has worked (in traditional BI) to continue to be the go-to solution. Because of this, we have a tendency to miss critical insights that would have been apparent under an AI framework.
3) A holistic view of the customer means going beyond siloed ad behavior :
Most marketers know this: A click is not what it once was. The vanity metrics of Likes and Followers do not directly denote intent or affinity. We have seen how these metrics can be easily gamed. Unless you can attribute these events to a customer conversion, it’s unlikely that they, in and of themselves, have any real value.
Acquisition and retention initiatives are also colliding. No longer can business control what messages are seen by current customer or potential prospects.
However, by leveraging AI, across millions of customers who follow a brand on social channels, we can determine the value of a Facebook ad or post as it relates to an individual customer and how that differs for a potential prospect. But this alone may not be enough to move a customer to convert. Perhaps, it’s also a recommendation from a friend via email… or a customer service call.
When analyzed in a sequenced journey, the individual contributions of those events in an individual’s journey can be truly valued.
4) Behavior is a derivative of sentiment:
Customers who are unhappy will not necessarily abandon a brand. When it comes to loyalty, price sensitivity may be the defining factor to purchasing an airline ticket because of the benefit of points, irrespective of the recent awful customer service call he/she may have experienced. However, all these events taken collectively, will allow the marketer to have a clearer idea of the risk of customer churn.
What AI allows is the ability to draws patterns from the complexities of human intent and determine the multiple drivers that may (to differing degrees) contribute to their decisions across millions of customers. It may vary with different products or services, across different times of year, different geographies and demographics.
5) Data is king:
In this new world we have the ability to contextualize human propensities and become increasingly adept at predicting the likelihood of behavior. Content has just been unseated! The principles that elevated content in anecdotally understanding what the new consumer values or what he/she needs, is what the methodology in AI seeks to uncover. Content is now a feature or event that may contribute to a business conversion, along with the thousands of other events, in combination, within each individual customer journey. Business can now analyze its marketing efforts alongside the customer journey to personalize by individual, and optimize overall campaign performance continuously.
AI has made the true definition of 1:1 entirely possible
We’ve come full circle. And for those of us who have practised 1:1 marketing, the computational capabilities, along with the wealth of data that has multiplied in the recent years, has afforded us an avenue to make this holy grail possible. For the traditional marketer, the possibilities are now endless.
This post originated from Marketing Insider Group
Founder at ArCompany, and Co-founder of Salsa AI, Hessie is a seasoned digital strategist, and intelligence analyst having held senior positions for top ad agencies including Ogilvy, Rapp Collins, ONE and Isobar Digital. She also has extensive start-up experience in social tech, online publishing and artificial intelligence like Yahoo! Answers, Overlay.TV, Jugnoo and Cerebri AI. Hessie is the co-author of EVOLVE: Marketing (as we know it) is Doomed! She is also an active writer for Cognitive World, Towards Data Science and Marketing Insider Group.