Algorithm & Blues: Bad Data Equals Bad AI

AI_Data_Feed_3
- Oct 11 2018
Algorithm & Blues: Bad Data Equals Bad AI

Making use of artificial intelligence is an essential part of informing and activating an effective mobile strategy. 84% of marketing organizations have implemented AI and machine learning this year. But today, it seems every marketing solution claims to have best-in-class AI, or the most advanced machine learning available. So how do organizations know who and what to believe? What factors make for effective AI that will generate results?

Beneath the complexity, AI is made up of just two key elements; algorithms and data. Algorithms enable machine learning, but data is the fuel that powers them, responsible for the majority of AI effectiveness. As such, validating the credentials of the data used by a vendor is the only way to check if its AI is ‘best-in-class’, and will ultimately generate the business results you require. Here are the top three things to keep front of mind when it comes to AI.

#1 Fresh, Real-Time Data

From the moment data is collected, it starts to go stale. It’s only fresh and relevant in real time. For example, 25-33% of all email addresses change each year, along with 18% of phone numbers, and 20% of postal addresses. Human beings are unpredictable, and similarly, static data on their mobile behavior, habits, and preferences does not stay accurate for long. This has implications for your AI solution. An algorithm will learn based on whatever it is fueled by. Therefore, if it is not fed real-time data, then its intelligence is deeply flawed. AI needs to take into account changes in behavior or shifts in trends as they happen in order to be effective. So, if you are selecting a provider, it’s important to ensure that their AI algorithms are fueled exclusively with fresh, real-time data.

#2 Granularity at Scale

Algorithms need scale to become intelligent. The good news is there’s no shortage of data available. But quantity doesn't always mean quality. For algorithms to scale intelligently, they require depth and granularity. Historically, being able to identify that a consumer was, for example, in the market for a new car would be considered granular. But today, this is far from sufficient. Your AI solutions need to determine what kind of car this consumer is specifically in the market for. An SUV? Sedan? Electric? Similarly, are they looking to buy, lease, or short-term rent? Without this level of insight, how can you determine if this consumer is worth engaging with? By working with granular, complete, accurate, and always timely data, this level of precision can be achieved.

#3 A Complete First-party View

Aside from quality, knowing where your data comes from should be the primary concern of any mobile marketer using AI. 74% of marketers say that first-party data provides the greatest insight into their customers. It’s no surprise that 81% of marketers showed the strongest ROI when using first-party data, compared to just 61% who relied on third-party sources.

The problem is, organizations can only draw first-party data from user activities within their own apps, sites, and ecosystem. This represents just a tiny fraction of the time users spend with their devices overall. Therefore, they lack scale. In order to fill the data gap, the option most organizations take is to turn to third-party vendors. However, many of these solutions rely on data segments taken from questionably qualified sources. In the era of GDPR, to risk not knowing that your data is consented legally is to risk considerable negative consequences for your business.

Activate Data That Matters With Purpose-Built AI

Ogury’s AI technologies are purpose-built on our own first-party data, drawn from the entire mobile user journeys of over 400m mobile user profiles. In full compliance with GDPR.

This provides our customers with real-time accuracy, data transparency, completeness, and scale. For brands, this means the most precise targeting in the industry, serving their ideal consumers relevant content, and driving maximum engagement at scale. For publishers, increased revenue generation from their apps and mobile websites.

If you are interested to learn about how Ogury’s purpose-built AI, fueled by first-party mobile user journey data can impact your business results, please contact me directly: christophe@ogury.co

 

Christophe Thibault, Chief Algorithm Officer