Michael Wu – The Future of Advertising

Michael Wu, PhD, Chief Scientist, Lithium Technologies
The Future of Advertising

The goal of advertising is to persuade consumers to take action, which usually means a purchase in most business settings. Despite the abundance of strategies, techniques, and tools, advertisers are often unaware of the behavioral economics and psychology behind human behavior and consumer action.

My research at Lithium informs us that for consumers to take actions, they must have the following 3 things:

  1. Motivation: the desire to take action
  2. Ability: access to all the necessary resources for taking action
  3. Trigger: a call to action. In advertising, the ad itself often serves as the trigger for action

Most importantly, these 3 things must happen at the same time in order to drive action reliably.

What’s Wrong with Traditional Advertising?

Traditional advertising often fails to influence consumers because there is a big gap between the exposure to an ad (i.e. the trigger) and the consumers’ ability and desire (i.e. the motivation) to act. The main problem is the context under which the ad is served to the consumers (i.e. the ad context) is often not the context where the consumers want and can take the action to purchase (i.e. the purchase context).

For example, you just saw an ad on TV at home, but you need to go to the store in order to buy. You saw an ad on your computer, but you’d like to get your friend’s opinion before you decide. The ad may be served to you on your favorite social media platform where all your friends hang out, but you may not want to enter your payment information on such social platform for security and privacy reasons. Even if the perfect gift idea is served to your smart phone, you may still want to get it later for your friend’s birthday.

Although modern advertising reduces the gap between ad exposure and purchase execution by making purchasing really easy, the gap is still big enough to prevent the consumer from making the final purchase at the bottom of the funnel. This limits the effectiveness of the ads. Consequently, advertisers often ramp up the frequency, intensity, and/or reach, hoping to maximize the probability of a coincidental match between the ad context and the consumers’ purchase context. As a result, advertising becomes intrusive, undesirable, and really not all that effective.

Advertising Could and Should be Context Sensitive in 2020?

A context sensitive ad (or contextual ad) is one that is so relevant that there is virtually no gap between ad exposure and purchase execution. Specifically, it means the following:

  1. Personalized: We are being bombarded with way too many irrelevant ads. These ads not only waste our time, they also waste money from brands and valuable resources from advertisers. Therefore, the content of an ad must be relevant to the consumers receiving it
  2. Social: In this information overloaded society, consumers have become very selective of what and who they want to trust. The Edelman Trust Barometer tells us that consumers trust other consumers and their friends. Thus, effective ads must enable social validation through the consumer’s social network and social proof within his community
  3. Location awareness: To shrink the gap between ad exposure and purchase, ads are best served near locations where the product/service can be purchased. Because traditional advertising technology and media are not able to achieve this, advertisers must increase the frequency and intensity of the ad to maximize the chance of recall when consumers are in the vicinity of where the purchase can be executed. This approach can make an ad repetitive, annoying, exaggerated, or unbelievable
  4. Just in time: Ads that have a high probability of driving the desire purchase action are those that are served to the consumers at the right time. And the right time for a consumer is somewhere within the narrow window along his decision journey when he is still susceptible to influence. This window of susceptibility is defined by when the consumer becomes interested in a product, all the way up until the purchase is executed. However, even within this window, there are still many times that consumers may not want to see an ad. Just-in-time means advertisers will need to infer when a consumer might want to see an ad

What We Need to Do Now to Get to Advertising Utopia?

Because a contextual ad is not only relevant in its content, but also reaches the consumer at the right time and right place with the right people (i.e. people he trusts), it will be much more effective. Hence, advertisers won’t need to be so invasive and ads don’t have to be something repugnant. In fact, if an ad can be context sensitive, then it may even become something desirable.

In terms of technology, we pretty much have all the necessary pieces to bring us to this advertising utopia. Now, we just have to piece it together carefully because dealing with human behavior is a delicate science.

  1. Mobile: In order to deliver ads at the right time and right place to the consumer, the consumption medium must be accessible through  mobile devices since they are with us pretty much all the time
  2. Data for context inference: To achieve contextual relevance, advertisers will need a lot of data about the consumers. For example, to deliver personally relevant content, advertisers will need our recent search and browsing history to figure out our interest. Google has this data. To achieve social relevance, advertisers must have access to our social graph to figure out who we trust and what they think about the advertised product. Facebook has this data. Moreover, advertisers must be able to show independent consumer ratings and reviews about the product through social media. Finally, to achieve location awareness, the advertisers will need to access the consumers’ GPS data, which is available from our mobile devices. With big data technology and sophisticated machine learning and analytics, these data sources can be mashed up to make context inference not only feasible, but practical
  3. Pull first, push later: The most challenging aspect of context inference is to determine precisely when a consumer might want to see an ad. This is challenging because there isn’t any platforms that tracks this data at large scale. However, advertisers can start collecting this data by letting consumers initiate the ad service session. So ads are pulled by the consumers instead of pushed by the advertisers. Furthermore, in order to understand when a consumer wants to see an ad, advertisers will also need ambient sensing technology that analyzes the environment of the consumers. This data will allow advertisers to learn when it is appropriate to push an ad in the future, and when is it not (e.g. when I am at work, watching a movie, or driving). Ambient sensing is what’s going to enable the just-in-time aspect of any context sensitive ads
  4. Opt-in by default, opt-out anytime: Because advertisers don’t have access to the data necessary for context inference, consumers must opt-in and volunteer their data in exchange for a more desirable contextual advertising experience. Therefore, all ads must be opt-in by the consumer by default. Moreover, they must be able to opt-out anytime with a simple toggle. If advertiser can achieve 100% contextual relevance, then this requirement might not be necessary because everything they serve is what the consumers want at the right time and right place. Moreover, there wouldn’t be any privacy issues, because privacy is only an issue when consumers are annoyed with ads that are either irrelevant in content, location, or time. However, we are nowhere near 100% contextual relevance and until we get very close, consumers must retain power to opt-in by default and opt-out easily anytime