Forget Identifiers: The Future Of Ad Tech Is AI

Apple’s new software update, currently in beta and set to launch this fall, will include a major privacy change that may disrupt the current ad tech industry. iOS 14 will ask users if they want to opt-in before sharing their IDFA identifier with app developers who use it to target them with ads. Facebook has attempted to take control of the narrativecriticizing Apple for making it difficult for advertisers to effectively target users through nonconsensual data usage.

“Like all ad networks on iOS 14, advertiser ability to accurately target and measure their campaigns on Audience Network will be impacted, and as a result publishers should expect their ability to effectively monetize on Audience Network to decrease,” Facebook said in a blog post. “Ultimately, despite our best efforts, Apple’s updates may render Audience Network so ineffective on iOS 14 that it may not make sense to offer it on iOS 14.”

The company also said that without targeting and personalization, mobile app install campaigns bring in 50% less revenue for publishers.

The population is steadily becoming more digitally literate in the wake of privacy scandals like Cambridge Analytica and privacy regulations like GDPROver two thirds of US and UK adults don’t trust how their data is being used, and 71% of consumers worry about brands’ handling of personal data. Customers are concerned about their data and privacy, so Apple is giving consumers a clear choice about the information they share and with which developers. Isn’t it time for informed advertising that doesn’t require using identifiers to market to us?

I spoke to Sheri Bachstein, VP and global head of The Weather Company and IBM Watson Advertising, about what potential solutions exist for advertisers going forward.

This interview has been edited for brevity and clarity.

Q: How do you think Apple’s update will affect the mobile ad tech ecosystem?

I think the value of the changes Apple is making is important for the consumer, who will have more control of their data and their privacy. I feel like anything we’re doing for the consumer as a collective industry is a good thing.

What’s challenging in the changes that Apple and Facebook are making is they haven’t been transparent with the publishers on timing, and haven’t really given them the opportunity to appropriately plan for the changes. I think there’s just a big opportunity for tech companies and publishers actually to work together to increase privacy collectively, but they have to be transparent with publishers.

Q: How much will publishers suffer not having the data that Facebook collects through identifiers?

I think it depends on the publisher, because there are solutions in the marketplace that aren’t really dependent on identifiers. If the publisher has strong first party data and they have a relationship with their customers, then the identifiers are not as important. If they have really valuable content that consumers want, those publishers will be fine. We’re an example of that. We have very valuable content that consumers want and frankly need and so they don’t mind underwriting that content with credible advertisers.

It really comes down to building that trust with your user, and it depends on the publisher’s strategy and what they have in place to mitigate against the identifiers.

Q: What about businesses that need Facebook’s resources to reach the right customer because they don’t always have their own big data collection or strong customer engagement?

I do think that some publishers will definitely be impacted, without a doubt. But I feel like we’re at a point where we can’t fight these kinds of changes as it relates to privacy. We have to start adapting and evolving, and IBM has been preparing to solve for the absence of traditional identifiers for a while now. And we are working to build alternative solutions for marketers to help with the problem. We really feel that AI is really the future for advertising.

Q: What kinds of AI solutions do you have?

It’s time for a new category of advertising that’s not dependent on identifiers. AI has been used in some areas of advertising, but it will become more essential with the absence of the cookie and the identifiers. So when you think about programmatic, it’s simply based on automatic transactions. The difference AI brings is it actually augments the human thought process so that it can recognize a pattern, it can make predictions, it can provide insights on the massive amount of data. So it will enable marketers and publishers to really make those connections at scale.

Q: What kind of data will you feed your AI to learn from? 

It can use first party data, so in our case it can use weather data to help make predictions. There’s a lot of data that actually happens on a platform around behaviors, sentiment, tone, that can be used to build these AI products and make predictions.

On the social side, we can provide insights to publishers to make sure that they’re selecting the right influencers based on content that the influencers may have written about. The AI can look at the tone of what’s written, can look at the quality of content of what’s written and make recommendations on that. So it’s a very powerful technology that frankly hasn’t been that well adopted in advertising because we haven’t really needed it. It’s been seen as a nice to have but that is changing. We feel this tech can help keep the advertising ecosystem alive without identifiers and cookies.

Q: How can AI augment creative advertising without disrupting it? 

AI can measure sentiment and use, and then it can dynamically assemble the creative in real time onto a platform. So for marketers that have creatives, the creatives actually have to provide the elements. What the AI does is assembles it in the right places. So for one creative, you might have five to six elements, like a picture, a call to action, a slogan. What the AI does is, based on what it learns — and it learns over time, so the more someone uses a platform the better an AI can learn — it can assemble the right ad that would appeal to that particular user.

AI, as we think about it at IBM, is not meant to replace humans. It’s meant to augment. It’s there to help them. Could you imagine a creative team trying to put together 100 combinations of an ad? That’s just not scalable, it’s not efficient.

Q: In other words, the AI will reassemble creative elements of an ad to find the most successful version based on collecting data on variables like click-through rates? 

That’s right.

Q: And that’s not a cookie thing?

It’s not based on cookies.

Q: When we talk about these kinds of new ads, would they still be as creepy and specific as the targeted ads we’ve all come to know and fear? 

Not exactly because it’s information that users willingly want to give. So again it’s that relationship between you and your customers. We’ve done a lot of user studies and users actually don’t mind relevant advertising. They actually prefer it because I think we’ve all seen the alternatives. So they don’t mind it, it’s just you have to make a decision with your customer — how do they want to underwrite your advertising business? And that’s really what it’s all about. They can underwrite it with credible advertisers delivering messages that they might be interested in, or they can underwrite it by paying to not see messages at all. And I think having that choice is really important.

Q: I’m sure many businesses will be affected at least in the short term by the changes Apple and Facebook are making. What can businesses learn from IBM about pivoting advertising strategies? 

I think what’s critically important is that you build trust with your consumers, and one of the best ways to do that is giving them the opportunity to control their data and being really transparent about it. And what I would really hope is that we start embracing new technologies and leave identifiers behind. Let’s embrace AI in ways that don’t use identifiers, that are safer, and let’s start evolving.


Original post:

Leave a Reply

Your email address will not be published. Required fields are marked *