AI 101: Deep Learning

Imagine that you are a marketer looking to run a targeted marketing campaign. What if you had a tool that could easily segment your market on the basis of factors like economic status, purchasing preferences, online shopping behavior, etc. so that you could customize your approach and messaging to each segment for maximum impact and conversion?

These are the kind of insights that deep learning (DL)* can offer.   

DL refers to a family of advanced neural networks that mimic the way the brain processes information and extract goal-oriented models from scattered and abstract data. What differentiates it from traditional machine learning is the use of multiple layers of neurons to digest the information.  

A DL program trains a computer to perform human-like tasks, such as speech recognition or predicting consumer behaviors. It is fed large amounts of data and taught what the desired output should be. The more data it’s fed, the better performance.

The program then applies calculations to achieve that output, modifying calculations and repeating the cycle until the desired outcome is achieved. The ‘deep’, hence refers to the number of processing layers that the data must pass through to achieve the outcome, and how the learning algorithms are stacked in a complex, hierarchical manner. The more levels or layers there are, the ‘deeper’ the learning.

DL can analyze huge volumes of data to detect patterns and predict trends and outcomes. This is especially interesting to marketers, finding application in predicting consumer behavior and campaign outcomes, marketing automation, sophisticated buyer segmentation and sales forecasting, to name a few use cases.

*Deep learning is not magic, but it is great at finding patterns.

4 Technologies That Are Transforming Customer Engagement Online

Consumers engaging with brands, seeking customer services and making purchases through social media is fast becoming the norm. Businesses need to constantly adjust their approach to reach out and respond to consumers with relevant and personalized messages.

Take advantage of these four technologies that will help your social media marketing reach a whole new level.

Chatbots: Manage the Volume of Online Messages, and Speed up Customer Service

Chatbots are seeing a surge in uptake and are helping marketers connect with their audience in new ways. According to Gartner, by 2020, 25 percent of customer service operations will use chatbots across engagement channels, and Facebook reports that its Messenger has more than 300,000 active bots.

Brands are increasingly using chatbots to manage the large volume of customer messages generated across social media, and respond to them in a human-like fashion. Powered by artificial intelligence (AI), bots become smarter over multiple interactions, and can be used to personalize marketers’ messages over time. They also help streamline and speed up customer service, allowing businesses to focus on bigger issues likes complaint resolution by taking care of routine customer queries.

AI-powered Prediction Engines: Access Granular Insights to Better Target Consumers and Boost Conversion

The use of AI in online engagement goes beyond just chatbots. Businesses are turning to AI to make sense of the huge volume of online data, access granular insights, predict audience behavior and boost user engagement. LinkedIn’s sophisticated machine learning algorithms, for instance, score candidates on the basis of location, work experience and other information, and use this to improve job-candidate matches.

On the marketing side of things, data-driven insights allow businesses to identify audiences more likely to convert and target them with relevant and customized messaging. For example, marketers can use Appier’s Aixon, a data intelligence platform, to discover new consumers, predict their behavior, and increase conversions by delivering targeted advertising.

Aixon’s recent integration with messaging app LINE extends these capabilities by allowing businesses to access powerful AI-driven insights into consumer behavior on LINE and push out highly targeted messages to them. For instance, a user visited an ecommerce site and added a dress to the shopping cart. A few seconds later, she would receive a push notification for an ad related this dress on her LINE app.

Augmented Reality: Offer Consumers Useful and Relevant Experiences that Increase Sales

In 2017, augmented reality (AR) moved beyond gaming and entertainment into business, with brands increasingly offering their customers useful AR experiences across social media channels. By 2020, 100 million consumers will shop in AR, and Apple’s launch of iPhone X, which offers users new AR capabilities and facial recognition, will only push more social platforms to integrate AR technology.  

The use of AR can be a win-win for both customers and businesses. For example, when Ikea’s Place allows users to preview how an armchair will look in their homes before buying it, not only is the customer making a better-informed decision, but the brand is also demonstrating how it fits into the customer’s home and life, boosting conversion and sales.

L’Oréal, which has bought the ModiFace AR beauty app, believes that this will boost online sales. Using facial recognition and AR technology, users can try out make-up on the app to see how it looks on their faces before actually buying the products. Even simple face filters like those that Snapchat offers are fast becoming a hot new way to advertise.  

Social Listening Tools: Listen to Consumer Conversations and Better Customize your Content

The popularity of social listening tools is growing as more businesses realize the importance of tapping into customer and peer-to-peer conversations online. Social listening tools will allow you to keep an ear to the ground and be involved in consumer conversations across different social channels. The zillions of gigabytes of data generated from these are a goldmine of insights into consumer needs and preferences.

Marketers can then leverage AI’s capabilities to identify keywords and phrases, and generate insights around demographics, trending topics of interest and sentiment, which will help them better understand customer intent throughout the journey. Marketers can use these insights to be relevant in consumers’ lives, by creating content that resonates with their consumers and offering them solutions that they need.

Technologies such as those mentioned above are helping businesses make informed decisions, reach consumers more effectively, and target them better. Adopt and embrace those technologies to stay visible and relevant in the age of customized marketing.

GDPR Frequently Asked Questions

Q: What is GDPR?

The General Data Protection Regulation (“GDPR”) is a new legislation by the EU parliament that lays out requirements for data collection, storage, and usage practice.

This new law is meant to replace the 1995 EU Data Protection Directive (DPD) to significantly enhance the protection of the personal data of EU citizens and increase the obligations on organisations who collect or process personal data.

Q: When is the GDPR coming into effect?

The GDPR will become fully enforceable on May 25, 2018.

Q: Who does the GDPR affect?

Although the GDPR is an EU regulation, the territorial scope of GDPR is potentially far wider as it can also apply to non-EU businesses in certain cases. Businesses that market their products to or monitor the behavior of people in the EU are required to be GDPR compliant.

Q: What constitutes personal data under the GDPR?

Any information related to a natural person, defined as ‘Data Subject’ in GDPR, that can be used to directly or indirectly identify the person. It can be anything from a name, a photo, an email address, bank details, posts on social networking websites, medical information, or a computer IP address.

Q. How is Appier prepared for GDPR enforcement?

We respect the privacy of Data Subject and ensure our compliance with a range of privacy protection regulatory criteria. These include requirements imposed by EU data protection legislation such as General Data Protection Regulations (GDPR) starting in May 2018.

We strictly follow our privacy policy in regard of any data we collect from Data Subject.

Q. Where can I find detailed information about the GDPR?

Please visit official GDPR resources available online for details.

 

Introducing Josh Shozen as SVP of Enterprise Solution Sales for Appier Japan and South Korea

By Junde Yu, Chief Business Officer, Appier

Josh Shozen, Senior Vice President of Enterprise Solution Sales for Japan and South Korea, Appier

I’m excited to welcome Satoshi (Josh) Shozen to our team as Senior Vice President of Enterprise Solution Sales for Japan and South Korea.

Based in Tokyo, Josh brings over 15 years’ experience in enterprise software sales management and digital marketing in Japan and the region where he has worked with global technology companies including Adobe, Proscape Technologies, and Microsoft.

Before joining Appier, Josh was the country manager for SundaySky Japan, the creator of SmartVideo platform, and helped develop the strategic partner ecosystem for selling into major FSI and Automotive clients.

In his new role, Josh will lead our teams in both countries to strengthen the business of Aixon, an AI-powered Data Intelligence Platform that Appier launched last July. More than 10 clients, including Japanese real estate portal LIFULL, have already successfully deployed Aixon and transformed their digital marketing campaigns.

While we continuously upgrade technology for Aixon’s functionality to help our customers’ business success, such as the recent integration with LINE Business Connect, we also see the strong potential of our business in both the marketing and data intelligence platforms.

With Josh on board, we are looking to accelerate our business in Japan and South Korea, especially the latter, a highly mobile-dominated market where 60-70 percent of e-commerce is made through mobile devices.  

Graduated from the University of Washington, Josh started his career at Microsoft in Seattle as a solution sales specialist. He was transferred to Tokyo in 2003 to manage Microsoft’s Global Accounts segment for Japan as a Regional Business Manager. Josh later moved on to lead the Japan and APAC business as VP for Proscape Technologies before joining Adobe Japan as Director of Digital Marketing Solution Div. Enterprise Sales in 2012.

As an avid surfer, Josh enjoys a variety of watersports and outdoor activities with his family and friends in his spare time, either in Japan or at his other home on the Big Island of Hawaii.


About the author:

Junde Yu is the Chief Business Officer of Appier, a leading Artificial Intelligence (AI) company. He leads the company’s Enterprise business, which includes Aixon, an AI-based data intelligence platform. Junde joined Appier from App Annie, where he was Managing Director of Asia Pacific. He started at App Annie as its first sales rep in the region and grew the sales and marketing team in the region to achieve very extensive revenues across the Asia Pacific region. It was also here that he acquired an appreciation for how enterprises could derive tremendous value from data.

Understanding the New Consumer Journey and Why It Matters

Understanding the new consumer journey

By Fabrizio Caruso, Chief Revenue Officer, Appier

Reaching consumers was way easier in the 70s and 80s — the age of mass marketing. Today, the consumer journey is far more fragmented as they divide their time between different screens, which complicates the path to a buying decision.

The savvy consumer now spends a significant amount of time on two or more screens every day. Globally and in Asia, device ownership continues to grow; more than half of Asian multi-device users (51%) now have two devices per person, while more than a quarter (26%) now operate on more than four devices.

This means that marketers have to do things differently, even though the aim is still about reaching consumers at the point where they can influence the buying decision — also known as the consumer touchpoint.

In the past, that touchpoint could have been through a TV ad at home, or in the newspaper. While this was simpler for the marketer, it also offered limited marketing options. The growth in devices adds complexity but it does open up new opportunities for marketers to influence consumer purchase decisions through different touchpoints in the consumer journey.

Why Cross-Screen Tracking is Hard

In order to effectively reach today’s consumers, optimising campaigns across all screens is a must. However, marketers need to overcome two challenges in understanding the consumer journey:

  1. Understand the ownership of multiple devices by the same consumer, in order to switch from a device-centric to a consumer-centric approach.
  2. Know which channel the consumer prefers, and the preferred touchpoint where the buying decision is made. This could be at the buy phase, or earlier when the shopper is wondering whether to buy, or when they are doing research and evaluations.

Marketers also need answers to these questions:

  1. How much overlap is there between devices?
  2. How can the right amount of engagement be achieved at the right frequency to spur consumers to take action on your brand?
  3. How much is harmful – wasting money or consumer goodwill?
  4. Which devices are best for driving upper funnel exposure, and which are best for direct action campaigns?

Artificial Intelligence (AI) technology can provide insights into the number of users, where and on what screens they viewed an organisation’s ads, and cross screen conversion paths. Insights can be gleaned into the role every screen played in the customer’s journey — not just the last screen they used before clicking the ad.

This allows marketers to optimise their brand assets to each screen to receive higher ROIs— whether it is click-through rate, engagement or conversion.

Drive more consumer purchases

When global retailer Carrefour expanded online, it wanted to reach more potential consumers through its online store. The marketing team wanted to boost awareness and drive more online purchases, while optimising the cost per action (CPA).

The use of Appier AI technology helped Carrefour identify all devices owned by the same user from billions of data points. When a potential buyer visits the Carrefour online store on one device (e.g. a laptop), Appier could analyse his or her profile and deliver customised product recommendations on other devices owned by the same person (e.g. a phone), motivating the person to buy from the Carrefour online store.

Appier’s AI analyses user behaviours both within and outside of the website to determine the activities of a single user as well as to segment different users. Cross-comparisons with the Appier CrossX database along multiple dimensions allow Appier to locate potential customers in similar segments to expand the original user base.

Through this partnership with Appier, Carrefour experienced click-through rate (CTR) of cross-screen users that was 87% higher, and a conversion ratio (CVR) that was 40% higher, than for single-screen users.

Drive more subscriptions

Using AI, popular dating app Paktor managed to drive more in-app subscriptions while increasing the in-app tutorial completion rate.

The Appier AI technology quickly identified users’ cross-screen behaviors and their interests in “socialising and dating” from billions of data points.

After the potential buyer downloads the Paktor app, Appier re-markets social/dating-related information to every device owned by the same user through a cross-screen approach. This deepens the Paktor brand impression and helps to trigger the motivation to become a Paktor VIP member.

Brand Safety Woes

Besides reaching and engaging customers, brand safety has also become a key concern with digital marketing, as organisations worry that their online ads may appear in a context that will damage the their brand.

While there are different aspects of brand safety, one key area is ad fraud, where online ad impressions, clicks and conversions are fraudulently represented in order to generate revenue.

AI can fight ad fraud and help protect brand safety. A study by Appier found that an AI-based fraud detection model was able to identify twice as many fraudulent transactions as a traditional rule-based model, as it is able to pick out ad fraud patterns that are difficult for traditional models to detect.

Now what?

With the revolution in the way consumers make buying decisions, using the right approach can help marketers be at the right place at the right time in the consumer journey.

How are you optimising your campaigns across all screens? If you want to know more about driving more click-throughs/conversions or to do more for your brand safety, get in touch today.

About the author:

Fabrizio CarusoFabrizio Caruso is Chief Revenue Officer of Appier. With over 15 years of experience in the digital and mobile industry, Fabrizio has established strong partnerships with top-tier brands and agencies, mobile operators and media companies in Asia. Fabrizio has engaged in management, planning, marketing and operational activities in the mobile payment, mobile internet content and mobile advertising industries. With insider knowledge of wireless and internet technologies, Fabrizio is also a frequent public speaker and author.

Before joining Appier, Fabrizio served as Senior Vice President Asia at Opera Software, a leading mobile internet and advertising company. Prior to that, he served as Managing Director for Asia-Pacific & Vice President for Business Development at Out There Media and held senior management positions at Buongiorno and Amdocs in the United Kingdom, China and the Philippines.

The Role of AI-driven Insights in Delivering Superior Customer Experience

The Role of AI-driven Insights in Delivering Superior Customer Experience

 

By Magic Tu, VP of Product Management, Appier

Customers are inundated with more choices than ever, while the convenience of digital makes it easy for them to switch to competing services through the web or mobile apps. As businesses come under pressure to gain new customers and retain existing ones in this fluid digital economy, many see improving customer experience as the key to coming out ahead. In order to do so, they need to understand their customers a lot better.

Understanding customers better

Defined as the sum of interaction between an organization and a customer, the customer experience isn’t solely determined by capabilities or pricing, but by consistently addressing what matters most to customers. For instance, a frequent guest disgruntled about a slow hotel check-in is likely to leave with a negative experience despite the hotel’s top-notch pool and gym. Yet a fellow traveller with family in tow may retain fond memories after being offered a complimentary bed for his young children to sleep in, even after that same slow check-in.

Over time, the sum of their experiences accrues to a tipping point that can result in one of them leaving for a competing hotel – or staying on as a highly loyal customer. Delivering exceptional customer experience therefore boils down to achieving a deep understanding of what individual customers want and delivering it.

But while the barista at the coffee shop around the corner can be expected to remember a regular’s preference and greet him or her by name, this isn’t scalable for most businesses. The only realistic way for brands to understand their customers better is to analyse customer records and transaction data for insights about individual preferences and general trends. They can leverage these insights to better tune existing services, develop new capabilities, or to create customised promotions that cater to individual customers.

Unfortunately, obtaining a unified view of customer behaviour is a challenge that most organizations struggle to overcome, and the multiple channels through which consumers communicate with brands today further complicate attempts to correlate them.

The power of AI

This is where artificial intelligence (AI) comes in. With the ability to tirelessly and accurately analyze huge volumes of data, AI can help businesses gain a competitive advantage by giving them insights they would not otherwise have. With AI, organizations can move away from a “gut-driven” marketing approach, where a marketer or product manager rely on their instincts, towards a data-driven strategy.

By cross-referencing customer data from a multitude of sources, AI can help build a highly accurate model of your customer base and determine the best ways to reach them.

AI can even take it a step further with predictive insights into the future behavior of your customers. One of the most exciting tools available today to marketers is predictive audience segmentation powered by AI. As part of the larger category of predictive analytics, predictive audience segmentation has the power to help companies identify a target audience with the highest potential for conversion to a sale or click or install, whatever your KPIs (key performance indicators) are.

The most advanced predictive audience segmentation tools look at behavioural patterns, collected from customer databases as well as from the internet, and combine them with demographic data to identify trends to single out the most promising leads. It goes even further in segmenting customers. It analyzes data to make recommendations that help you find and grow your target audience.

Commonwealth Magazine, a leading media group in Taiwan, used the predictive audience segmentation capabilities in Appier’s Aixon platform to discover new customers, increasing subscribers and purchases on their website by over 400 per cent.

A data-driven world

There is no question that AI is a vital technology for our data-driven economy. Many traditional marketing promotions are either conceived from past assumptions that are no longer be valid, or which cannot be scaled due to an over-dependence on experienced employees. On the other hand, the judicious use of AI can reveal pertinent insights that can be leveraged to craft more relevant promotions.

Of course, it is important to note that AI is not magic and cannot be expected to solve every business problem. There is also a structured process that that must be adhered to for successful AI application, starting from the gathering of the requisite data. Yet there is also no question that properly implemented, AI with its ability to deliver tireless analysis is a crucial building block to drive innovation and deliver tangible benefits to businesses.

About the author:

Magic Tu talks about the role of AI in driving superior customer experienceMagic Tu is VP of Product Management at Appier. He is responsible for streamlining the process of product development, from product ideation to product launch, collaborating with various teams including R&D, product management, marketing, and business development. Aixon, the AI audience prediction and analysis platform, is the first product that his team has launched.

Prior to joining Appier, Tu was the director of software product management at HTC, a smartphone and virtual reality device company, leading the overall software experience planning for the HTC Sense smartphone and accessories. He also led the program management team that took care of third party vendors and partners including Google, Yahoo, Microsoft, and Nokia for cross-company projects. Before HTC, Tu was the lead developer in Springsoft (now Synopsys), an electronic design automation company based in Taiwan. Tu has BS/MS degrees in computer science from National Taiwan University.

Technical Insights: Basic C++ 11/14 for Python Programmers

Our new blog post features a short list of some common python programming patterns and their C++ equivalents. This can help programmers learn C++ in a more efficient way if he or she already knows Python.

Leave us a comment below and let us know what you’d like to see covered in our future posts!

█ Read More

Technical Insights: Introduction to GraphQL|goo.gl/d7PyXH

█ Join Us

Our current openings|goo.gl/rx1jce

 

Predictive Audience Segmentation: Take the Shortcut to Identify Your Target Audience

By Junde Yu, Chief Business Officer, Appier

The days of general advertising or marketing campaigns targeting the masses are numbered. Whether you are selling a bank loan or clothing online, you have to know your specific target audience.

The hard part? Identifying the right audience to help you maximize your marketing return on investment (ROI).

Analyzing and segmenting online traffic can be a painfully manual process. Efforts can range from educated guesses to applying simple analysis tools on data. While these will work when you are analyzing a handful of dimensions, the real challenge is when there is complex data or a combination of over 80 dimensions to analyze.

Powerful AI tools to target your audience

Today, one of the most exciting tools available to marketers is predictive audience segmentation powered by artificial intelligence (AI). As part of the larger category of predictive analytics, predictive audience segmentation has the power to help companies identify a target audience with the highest potential for conversion to a sale or click or install, whatever your KPIs (key performance indicators) are.

Predictive Audience Segmentation

Commonwealth Magazine, one of the most influential magazines in Taiwan, experienced dramatic results when it used the powerful predictive audience segmentation capabilities in Appier’s Aixon platform. Not only did the publication discover valuable readers they hadn’t been able to reach in the past, the campaigns exceeded the Return on Ad Spend (ROAS) by 300%, and subscriptions and purchases increased by 404%, compared to KPIs.

The most advanced predictive audience segmentation tools look at behavioural patterns, combining it with demographic data to identify trends to single out the most promising leads. It goes even further in segmenting customers. It analyzes data to make recommendations that help you find and grow your target audience.

An AI-powered platform like Aixon can help you to:

  1. Get an accurate audience picture.

Getting a unified view of customer behaviour is hard, as organizations struggle to consolidate data from disparate sources like the corporate website or mobile apps, as people often consume content on multiple platforms. Also, the data may be piecemeal and disconnected.

Aixon is able to unify an organisation’s data from disparate sources and platform into a single user view. That view is layered with Appier’s own considerable mine of billions of anonymized device profiles in Asia, to provide a better contextual picture of the different customer segments. This precision in data allows for more accurate predictions and segmentations, distinguishing between customers who are likely to churn, refer other customers, or likely to make a purchasing decision. It may even unveil the top revenue generators.

  1. Drive more timely sales or conversions.

You can better reach customers when they are ready to make a purchase as they move through the buying cycle. Predictive audience segmentation can help to highlight customers when they are most receptive to drive more sales or conversions.

  1. Discover new markets.

Predictive audience segmentation technology can help to uncover new markets by analyzing data based on your objectives, and identify the top opportunities within them. The insights from the behavorial data together with the demographic data may identify new target segments. As new prospects are targeted, the algorithms will learn the company profiles and help to further refine your market segments.

An example is Japanese real estate information service provider LIFULL, which had a rich vein of CRM data that was untapped. The company has since worked with Appier to integrate and analyze its vast online and offline real estate databases. LIFULL is using Aixon to cut through the clutter of massive data to drive more effective online marketing programs and to develop new innovative businesses.

  1. Identify users who might leave your service and re-engage them online.

Aixon can use churn forecasting to identify the patterns and trends of customers who had churned or left in the past. Based on that data, it can forecast how likely existing customers may churn or leave your service. This insight can be invaluable to marketing and sales teams who can then follow up and re-engage with this segment of customers.

  1. Boost your revenue by finding the right audience for your advertisers.

By segmenting customers based on their behaviour and demographics, this allows you to find the right audience for your advertisers. This means that marketing, recommendations and promotions can be tailored for different customer segments.

Drop us a line…

Predictive audience segmentation tools can be easily used to boost marketing efforts, complementing existing technologies. Whether you are working on predictive marketing, personalization or just wanting to improve your marketing efforts in general, Aixon’s predictive audience segmentation technology can help you to seamlessly bridge the gap between identifying the target audience and marketing execution.

Aixon’s additional demographic and behavioural data, interests and keyword insights about the audience provides a richer level of detail that can help you to single out the most promising leads, and drive towards a more optimal conversion rate for a higher volume of sales.

Please don’t hesitate to reach out to me or anyone from my team to find out how Aixon can help you take the shortcut to targeting your audience.

About the author:

Junde YuJunde Yu is the Chief Business Officer of Appier, a leading Artificial Intelligence (AI) company. He leads the company’s Enterprise business, which includes Aixon, an AI-based data intelligence platform. Junde joined Appier from App Annie, where he was Managing Director of Asia Pacific. He started at App Annie as its first sales rep in the region and grew the sales and marketing team in the region to achieve very extensive revenues across the Asia Pacific region. It was also here that he acquired an appreciation for how enterprises could derive tremendous value from data.

How Appier Fights Ad Fraud with Artificial Intelligence

Ad fraud is costing the industry billions of dollars. Joe Su, Appier’s Chief Technology Office, described in a blog post how Appier is using artificial intelligence to fight ad fraud. This infographic summarises how AI can help combat one of the top scourges of the advertising industry.

About Appier

Appier is a technology company which aims to provide artificial intelligence platforms to help enterprises solve their most challenging business problems. For more information please visit www.appier.com.

Fighting Ad Fraud with Artificial Intelligence

By Joe Su, Chief Technology Officer and co-founder, Appier

Advertising fraud, ad fraud for short, has become a major threat to the digital advertising industry. According to the Association of National Advertisers in the US, ad fraud will cost companies an estimated US$6.5 billion in 2017. A recent report by Juniper Research paints an even grimmer picture, estimating advertisers will lose US$19 billion to fraudulent activities next year. This figure, representing advertising on online and mobile devices, will continue to rise, reaching US$44 billion by 2022.

The industry has spent considerable resources looking for effective ways to mitigate the effects of ad fraud. I use mitigate deliberately because just as with cyber fraud or financial fraud, there is no way to totally eradicate the problem: you can only hope to stay one step ahead of the bad guys.

Most ad fraud countermeasures have centred on rule-based methods and these are effective ways to combat simple ad fraud activities. However, the ad fraud attempts are becoming more sophisticated and traditional countermeasures are inadequate today.

An AI-based approach

As ad fraud attempts become more sophisticated and difficult to detect, so must our fraud detection mechanisms evolve in tandem and the only way that this can be achieved is using artificial intelligence (AI).

An AI-based ad fraud detection system actually starts with a rule-based approach as the base but through self-learning, builds layers of defence that learn from each suspicious activity that it detects. An AI-based model also has the advantage of being able to view patterns on many more dimensions than a traditional system.Artificial Intelligence approach to fighting ad fraud

Traditional rule-based models typically analyzes activity on between one to three dimensions. An AI-based model analyzes over 80 dimensions at a time, enabling it to detect extremely sophisticated ad fraud patterns. With self-learning, AI-based models can evolve as ad fraud patterns evolve to evade traditional systems.

A Real World Study

To demonstrate the advantages of an AI-based approach, Appier examined data on its own network over four months from May to August this year involving over 4 billion campaign data points including ad clicks and app installs. What we found was that the AI-based fraud detection model was able to identify twice as many fraudulent transactions as the traditional rule-based model. The AI-based model also proved to be more cost-efficient for advertisers, yielding a 3.6 percent higher return on advertising spend (ROAS) than the traditional model.

The greatest advantage of AI though, was its ability to detect sophisticated ad fraud patterns not previously reported. On pattern that our AI system flagged is what we call “the chameleon”.  This is where dishonest publishers disguise themselves as legitimate publishers at first, only to generate fraudulent installs at a later date.

Another suspicious activity detected by our AI is what we have termed “inventory burst”. With this pattern, a fraudulent publisher will generate an abnormally high inventory count in the absence of an appropriate level of in-app registration activity.

Final Word

You can download the full report of Appier’s study here. Ad fraud is costing the industry billions of dollars and has become extremely difficult to detect. Traditional rule-based methods are limited in their ability to detect new and increasingly sophisticated ad fraud patterns. AN AI-based approach with its ability to analyze multidimensional data and with self-learning is a better approach to fighting ad fraud.

About the author:

Joe Su, CTO, AppierJoe Su is CTO and co-founder of Appier.  Su has been hacking and building systems since high school, where he won first prize in the 3rd Annual National Center for High-Powered Computing Programming Contest Taiwan. Since then he’s been involved in system design and development in a variety of areas, including social games, VoIP, distributed computing and online geographical information. Prior to founding Appier, Su co-founded and ran Plaxie, an independent game studio focused on developing intelligent mobile and social games. Previously, he joined Artdio Technology as a programmer and served as a researcher in Computer and Communications Research Laboratories (CCL) of Industrial Technology Research Institute (ITRI), a leading high-tech R&D institution in Taiwan.