Any good marketer knows one thing for certain: Good data is gold.
It helps you make informed decisions, pivot campaigns, predict audience behavior, and compound marketing success.
But there's one huge, often insurmountable problem with marketing data.
It's nearly impossible to assess treasure troves of data at scale. This is a problem that artificial intelligence is uniquely qualified to solve, and why Near exists today. It's uses the world's largest behavior-related data set in the real world, and couples powerful artificial intelligence and machine learning models to power self-serve SaaS products. Of course, it does all this with consumer privacy at the forefront.
We talked with Madhusudan Therani, CTO & Head of AI at Near, to learn more.
In a single sentence or statement, describe your company.
Near is a leader in providing intelligence on people and places.
How does your company use artificial intelligence in its products?
We take raw, unfiltered, structured and unstructured data signals from multiple sources such as data aggregators, mapping companies, WiFi partners, census data, public hotspots etc. and build context about people and places using advanced data science models.
This people and place intelligence is productized in our SaaS products, making it very easy for all our consumers to access this intelligence—and act on it.
The data harnessed into the Near platform has characteristics that has motivated the use of AI/machine learning technologies at scale across various stages of data transformations and insights generation. At Near, the AI models are done in-house and are part of the Near toolkit to solve the below problems.
AI investments at Near:
- Data cleansing, clustering, noise removal, and spatial property estimation
- Data fusion
- Consumer profile property inferences
- Place profile property inferences
- Identity unification/entity resolution
- Audience and place cohort properties
- Powering search in Allspark
- Audience segmentation
- Audience assistant that automates audience curation by "understanding" the query.
What are the primary marketing use cases for your AI-powered solutions?
Two major areas of AI applications in marketing include data intelligence and operational intelligence.
Data intelligence encompasses:
- Gaining insights about campaigns through data.
- Understanding audiences and profiling them based on their behavior patterns.
- A/B testing not only creatives, but all aspects of the marketing process.
- Integrating data from multiple perspectives to guide better choices throughout the marketing lifecycle.
Operational intelligence involves answering questions like how to add intelligence to your workflows during design and execution in media planning, ad operations, creatives management, and integrated channel management.The Near Platform and products address both of the above scenarios.
What makes your AI-powered solution smarter than traditional approaches and products?
Our AI-powered solution creates better business results and allows for faster and more efficient media planning.
Clients have achieved superior business results with our AI-powered solution. Near’s solution leverages online as well as offline data, and proprietary data science models to get a holistic view of the consumer. The diverse data sources help Near get better audience context, thus enabling better marketing. For example, Decathlon was able to draw over 150,000 shoppers within the first week of its new store launch with our marketing solution.
Near’s marketing product, Allspark, enables clients to curate audiences on the fly and get immediate brief estimates. This leads to a reduction in the brief response time to a few minutes from the traditional 24 to 48 hours. The product shields its users from the underlying structural complexity with an intuitive interface.
Addressing both of these factors helps our customers address the top-line growth as we provide more options to reach prospects in various nuanced ways. This is all while reducing the total cost of ownership of the martech/adtech operational toolkit.
Are there any minimum requirements for marketers to get value out of your AI-powered technology? (e.g. data, list size, etc.)
There are no minimum requirements to use our current SaaS products.
Who are your ideal customers in terms of company size and industries?
Ideal customers for our products and data include:
- Publishers and media houses with large-scale media operations that need to enhance both from a data and workflow perspective.
- Media buyers, including big and small agencies, aiming to improve their operational and data intelligence.
- Enterprises that are adopting CDPs to enrich and activate their data sets.
- Enterprises that have spatial assets which need to be monetized by adding people and place intelligence.
What do you see as the limitations of AI as it exists today?
More than limitations of AI/ML per se in terms of its algorithms and such, current adoption and implementation of AI tools (in marketing and other domains) is severely hampered.
First, it's hampered by the existing legacy/archaic data systems adopted over the past couple of decades and the limited data they capture. Only by rearchitecting current data stores and overall marketing workflows and capturing the right data sets can we utilize data-centric AI/ML systems properly.
Second, given the fragmentation of control of different aspects of the martech/adtech ecosystem, AI systems utility is limited by the lack of coordinated and common data collection, sharing agreements, and process baselines. Deployment of AI tools by every player hits some blind spots where one has to cross boundaries. Also, the adversarial nature of players in the ecosystem trying to control different resources in the ecosystem limits businesses to realize the full benefits of AI as it exists today.
If these two issues are addressed, current AI techniques can address nearly 60-70% of issues in the coming decade.
What do you see as the future potential of AI in marketing?
First, future AI applications include superior data intelligence. This means collating and deriving intelligence of all data sets—external and internal—so businesses get a true, 360-degree picture of customers in a privacy-safe manner.
Then, marketers can enact superior execution based on this intelligence. This involves deploying resources and devising workflows to act on intelligence available to businesses. AI can help define the best moment for action, and automate it.
AI-based tools will pervade both buckets in the near future. Specific to marketing, as the search for niche audiences increases, more AI tools will be required to support the overall prospecting, customer acquisition, and customer retention processes across all enterprises. Though privacy is an overarching concern, if the end consumer gives consent to disclose their preferences, businesses may be able to deliver personalized consumer experiences.
Any other thoughts on AI in marketing, or advice for marketers who are just starting with AI?
AI for marketing is still in the early days. For marketers who are just starting out with AI, you need to sift through the hype.
Do the work manually and build your intuition before jumping on any specific tool or approach.
Paul Roetzer is founder and CEO of Marketing AI Institute. He is the author of Marketing Artificial Intelligence (Matt Holt Books, 2022) The Marketing Performance Blueprint (Wiley, 2014) and The Marketing Agency Blueprint (Wiley, 2012); and creator of the Marketing AI Conference (MAICON).