TechBytes with Jason Atlas, CTO, Adbrain

0 89
Jason Atlas

Jason Atlas
CTO, Adbrain

Adbrain’s Identity AI™ approach utilizes machine learning technology to create cohorts of identities across three dimensions of data: People, Places and Devices. The multi-dimensional identity methodology affords a comprehensive view of identity beyond a single-device context. We spoke to Jason Atlas, Chief Technology Officer, Adbrain,  to understand how Customer Identification Mapping technology helps resolve the challenges in marketing attribution.

MTS: How have the definitions for “unified, cross-platform customer experience” evolved with time and technology?
Jason Atlas: 
The terms have evolved with the creation of new delivery experiences, such as smart phones, tablets, smart TVs, the connected home, the IoT including wearables such as smart watches, and the extraordinary amount of mobility with laptops taking over from desktops as the norm. Given these new technologies, the need for new means of gathering data along the customer journey based off different types of data has quickly emerged. This new period marks an evolution from the early 2000’s, when we were just experiencing the emergence of search and adoption of content platforms like Yahoo and Google. Those platforms collected customer intelligence information, like user agent information from browsers and applications, which allowed those companies to serve their users smarter, more targeted ads.

Technology innovation in the concept of identity is really at the fulcrum for these changes. Single Sign On, Federation of accounts, Oauth tokenization, and other emerging identity technologies have allowed the data science to gain understandings, and make the work we do possible.

MTS: How does Customer ID mapping technology solve the pain points in marketing attribution?
Jason: 
For a marketer, it is about not only knowing, but finding your customer first and foremost. By identifying a user as they move across devices, by creating a singular view of that customer, the marketer can extract a tremendous amount of information. What types of devices are used? How many people are in a household? What are their purchasing patterns (i.e. do they start browsing on mobile or a laptop and on which device are purchases being made?) The marketer also knows a significant number of other key attributes, such as demographics, device used, time on site, and others. The combination of these attributes builds a complete customer profile allowing marketers to master attribution.

MTS: What does Adbrain’s roadmap to leverage AI/ML look like?
Jason: 
Adbrain has been on an aggressive trajectory using more advanced data science methods of machine learning and artificial intelligence. As we have evolved from a statistical and heuristic model, into the use of true supervised learning over the past few years, and expanded our supervised learning into a multitude of different methods, from identifying, to neighbouring, to clustering. As we look ahead, we are going into deep learning, with areas such as neural nets, tensor fields, and others, as yet to be explored unsupervised learning systems to continue to extend and expand the deductions we make from our observations.

MTS: What are the key metrics in Customer ID Map that a marketer and technologist would look at differently? How does Adbrain correlate both these requirements to create a unified personalized experience?
Jason: 
Marketers tend to be focused on getting a deeper understanding of their customers. Who are they? What do they buy? What are their habits? How do demographics and time impact decisions? Technologists tend to be more concerned about what the data means, and how it can be used to derive a better algorithm, delve into a new way of understanding trends, or whether different “features” when combined give different results. It is Adbrain’s job to provide views for the marketing team that are consuming the content of the data we provide, and also to provide data and information to technologists or data scientists, so they can validate our data. Transparency in our approach, and our output is paramount.

MTS: In a hugely disruptive space like data science where privacy and transparency laws are getting tighter, how does Adbrain retain the trust of customers?
Jason: 
We must not only be compliant with all privacy laws and regulations, but we must detail how we are compliant as well. As new regulations are rolled out, we will be telling our customers exactly what we are doing to ensure consent and opt out. We have been thinking about the use of personal data and our systems for many years, and have worked on means and methods of still producing best in class results, without ever violating privacy. As we advance our technology and methods, we continue to have our data science work in lockstep with our business to keep privacy and regulatory obeisance a top priority.

MTS:  Moving slightly away from the clichéd narrative around the use of AI/ML, what are the other technologies that Customer ID Mapping platforms would benefit from?
Jason: 
AI/ML get a lot of the press, and have an almost metaphysical aura around them. One thing all of us who are exploring cutting edge technologies needs to understand is, you cannot throw away things that worked or can still be improved just because they are not cutting edge. Advanced statistics and heuristics are probably more common than true ML, but many folks still call these ML. The more data we have, the more intelligent we can be but to have more data, we need to be able to absorb, process, normalize, compute and output massive quantities of data. The use of Kafka to ingest streams of data.  Graphing databases for producing the outputs. Data visualization frameworks, to allow people to gain insights into their data that were previously unavailable. The fact that we can now ingest petabytes of information, producing identity graphs of billions of individuals is something that would not have been computationally possible even 10 years ago, except in the most advanced labs in the world. This is now a commodity. The tools to do the work are as important as the ML/AI itself. To be able to do math at scale and speed is a huge change and equals the importance, and not coincidentally, corresponds to the rise of ML/AI.

MTS: Thanks for chatting with us, Jason.
Stay tuned for more insights on marketing technologies. To participate in our Tech Bytes program, email us at news@martechseries-67ee47.ingress-bonde.easywp.com

Leave A Reply

Your email address will not be published.