Using A.I. To Measure Marketing. An Interview with Adgorithms’ Or Shani

Or Shani, CEO and Founder of Adgorithms, wanted to change the way marketing programs were executed and how marketing decisions were made. In an attempt to streamline marketing processes (from execution and optimization to analysis and calibration), he created an artificial intelligence-based marketing system he calls Albert (for Albert Einstein). “The complexity in the way we were buying, analyzing and conducting marketing was immense, and over the years it has gotten even worse,” he explained. His company, Adgorithms, purports to break through the complexity of digital marketing with Artificial Intelligence.

Charlene Weisler: How does Albert work?

Or Shani:  We developed artificial intelligence technology, which helps to do many of the time-consuming, manual tasks involved in modern day digital marketing without the complexity. We currently integrate with nearly 30 vendor platforms within the martech and adtech ecosystem. Albert comes in and acts as a single centralized point of contact. Within minutes, we can run Google Search, social, paid and non-paid campaigns on the fly without any manual input (other than the KPIs and target customer information the marketer gives us upfront).

Charlene Weisler: What does your company contribute to TV measurement to gain a greater understanding of how TV is being used/consumed?

Or Shani: When it comes to TV measurement, we integrate with any third-party solution the advertiser is already using. By doing this we’re able to use their viewer data to inform all facets of their digital campaigns, which brings online and offline efforts into greater alignment. Ultimately, this approach gives TV and digital efforts a shared focus, where digital insights inform TV efforts, and TV data informs digital targeting and conversion.

For instance, Albert can use TV data to correlate certain consumer behaviors that he sees online—on websites, social, and search—with specific TV advertising spots. Equipped with information about the relationship between TV and specific user patterns and trends, Albert can now make assumptions about different audience micro-segments and begin acting on them. This introduces a new way for brands to convert customers online and on mobile, who they’ve initially identified through TV. Whereas most second screen solutions are primarily focused on pairing conversion opportunities directly with specific show content, we’re able to use viewer data to “find” specific user-types and lookalikes online, which we can then target and convert using the content that's most likely to appeal to them (rather than TV-specific content).

Charlene Weisler:  What data metric do you use to match digital to TV?

Or Shani: In matching digital advertising to TV, Albert utilizes both deterministic data from the different media providers and different targeting methodology.  For example in a digital campaign coinciding with a TV one, Albert could target users based on a number of different interactions they have with the specific TV program during which the TV Ad will be shown.  This could include targeting users that have the TV show as one of their interests on Facebook, liking its social pages, or even those that are following the show's main stars on Twitter.  When it comes to the Search channel, Albert could target users that are searching for the TV show or related searches before, after, or during the time of the program.

Charlene Weisler: Do you work with segmentations?

Or Shani:  In a way, yes, but segmentation works differently in our system. For example, one of our clients might come to us not knowing exactly who they should be targeting. Maybe their target customer only represents 2% of the market they’re in and they don’t know exactly how to find them. Or maybe their recent campaign didn’t produce, so they are reluctant to keep targeting the same audience over and over. Albert will step in and find the right audience for them.

One way we do this is by integrating with our customers’ CRM and getting further data about their clients or customers. Albert will then start with mini-campaigns, creating micro-segments or audiences of one, as he goes and learns what works and doesn’t. Once he’s determined this, he rapidly expands and continues calibrating along the way until he meet the marketer’s KPI.

Charlene Weisler: What metrics do you use?

Or Shani:  Technically we can track everything. The measurement trend in marketing is to match back to more and more concrete results that are tied to revenue. And that is a good direction for the industry. Some clients aim for clicks and eyeballs but there is more demand for sophisticated optimization metrics that tie back to revenue or sales.

Charlene Weisler: Can you track engagement?

Or Shani:  Everyone has a different definition of engagement. Some define it as one minute spent on the site. Others define it as a visitor reading an article, viewing a video or downloading a whitepaper. It depends on what works for your brand. These things can be easily tracked, but anything based on sentiment or an overall feeling about are hard to assess through analytics. With a system like ours that is committed to performance metrics, we can bypass these more abstract measurements and simply look at the results. Did we do our job? It’s an easy yes or no. The trick is measurement which is very hard. Analytics cannot provide that type of measurement solution.

Charlene Weisler: What about the roll-out of Smart TVs? Will you be able to measure that?

Or Shani:  Theoretically, yes, but it is very complex to do so. Different connected TVs have different operating systems. And just like mobile, it is still evolving. It is not as simple as tracking IDs.

Charlene Weisler: Do you see marketing as a creative job or as more quantitative?

Or Shani:  Marketing has and always will be a mix of art and science. Due to issues I mentioned previously, however, the pendulum has swung more in the direction of “science” in the past decade-plus. As a result, the creative aspect has started to suffer. Marketers own the brand, customer acquisition and retention. The best way to build a strong, sustainable brand, which supports the overall business growth, is through storytelling.  This requires a completely different part of the brain than the part used to analyze things until you’re blue in the face. And marketers as a whole aren’t particularly good at this type of analysis anyhow (mostly because it’s nearly impossible for any human to scour that many channels and produce meaningful insights on each in a short period of time). The need to tell a good story doesn’t go away with data, and marketers are good at telling those stories. Let them focus on tugging at the hearts and minds of customers, while technology focuses on the data aspects of marketing.

Charlene Weisler: Looking ahead the next five years, what do you see happening in the media measurement landscape?

Or Shani:  There will be many more AI-driven technological solutions that will make life easier for us, like self-driving cars or personal assistants. These things will become more common. In media, CMOs will be liberated by technology. Marketers will be free to make marketing fun again and not just focused on execution of a campaign plan.

This article first appeared in

No comments:

Post a Comment