<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Interceptd - Medium]]></title>
        <description><![CDATA[Thoughts, information, news, reports about mobile ad fraud from Interceptd team - Medium]]></description>
        <link>https://medium.com/interceptd?source=rss----51929d39d414---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Interceptd - Medium</title>
            <link>https://medium.com/interceptd?source=rss----51929d39d414---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 14:53:10 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/interceptd" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[13 Experts Share Their Mobile Success Secrets]]></title>
            <link>https://medium.com/interceptd/13-experts-share-their-mobile-success-secrets-b907e5af173a?source=rss----51929d39d414---4</link>
            <guid isPermaLink="false">https://medium.com/p/b907e5af173a</guid>
            <category><![CDATA[digital-marketing]]></category>
            <category><![CDATA[apps]]></category>
            <category><![CDATA[app-store-optimization]]></category>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[app-development]]></category>
            <dc:creator><![CDATA[Brittney Ihrig]]></dc:creator>
            <pubDate>Wed, 20 Nov 2019 10:30:16 GMT</pubDate>
            <atom:updated>2019-11-20T10:30:16.068Z</atom:updated>
            <content:encoded><![CDATA[<h4>… Just for you!</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ULk7SrisStFNRiel6iEKlA.jpeg" /></figure><p><strong>Exactly how does one make it in today’s competitive and ever-changing business landscape?</strong></p><p>The world is more connected, accessible and idea-saturated than ever before. Possibilities seem infinite, yet there is barely a place untouched by innovation. In 1945,<a href="https://www.theguardian.com/commentisfree/2018/mar/16/university-degree-measure-benefits-tuition-fees"> 2</a> percent of England went to university, today it is a standard addition to any entry-level resume and a ticket to a<a href="https://www.bbc.com/worklife/article/20190401-why-worthless-humanities-degrees-may-set-you-up-for-life"> 50/50 shot at employment</a>. Consumers are yawning at the<a href="https://www.forbes.com/sites/forbesagencycouncil/2017/08/25/finding-brand-success-in-the-digital-world/"> 4000–10,000 advertisements</a> they’re exposed to daily, whilst businesses are competing for a chance to be seen. The world has never been full of so much accessible data and information. Nearly every waking minute can be optimized. You can listen to your favorite book whilst on your morning jog. Do your most analytical work in the morning, whilst you’re at<a href="https://www.businessinsider.com/best-time-day-work-according-to-science-2018-5#your-stress-hormone-levels-also-tend-to-be-higher-right-after-you-wake-up-in-the-morning-putting-the-body-on-high-alert-so-its-a-good-time-to-soak-in-the-advice-of-others-7"> peak mental performance</a>, engage your core whilst you pen some notes and by midday, you can take a<a href="https://www.businessinsider.com/nasa-research-found-the-perfect-length-for-a-power-nap-2019-3"> N</a>ASA recommended 10–20 minute power nap during your lunch break.</p><p>Like a ray of light, peeking through a blanketed sky, we offer you real advice from professionals, who have ‘been there’ and probably ‘done that’. They have hustled and bustled through the minefield of increasing competition, data, and endless possibilities and emerged victorious in the mobile and app industry. Enjoy their success secrets.</p><h3>Know How to Ask for Reviews</h3><p>It is no secret that great reviews are essential to rank in, both, Google and Apple app stores. However, many apps often struggle to get just a few. The most effective method I found to get real quality reviews is to strategically place the “ask for reviews” notification in the user experience of the app. For example: if your app is a game, ask for a review after a user unlocks an achievement. Understanding where your app delivers the maximum value and emotional satisfaction is how I have seen apps go from less than 100 to a thousand reviews.</p><p><a href="https://www.linkedin.com/in/daniel-elias-09a93550/"><strong>Daniel Elias</strong></a><strong>, Marketing @ Jibrel</strong></p><h3>Be Open, You Never Know</h3><p>Mobile Marketing Tip: Over my time in digital marketing, it never ceases to amaze me on the success of a test when it is something I am normally hesitant to try. An example of this is when I didn’t want to run Incent traffic the CEO of my last company was pushing me to work on a ranking strategy to get organics up. After doing this I ended up finding a source ID that was giving us positive ROAS results. After hunting it down, it turned into one of our largest and most profitable sources. So from then on, I try almost anything as long as I can de-risk it through a deal structure.</p><p><a href="https://www.linkedin.com/in/trevor-laehy-bb70678/"><strong>Trevor Laehy</strong></a><strong>, Founder @ Upshot Labs</strong></p><p>Read the full article <a href="https://interceptd.com/13-experts-share-their-mobile-success-secrets">here</a><br>Or here: <a href="https://interceptd.com/13-experts-share-their-mobile-success-secrets">https://interceptd.com/13-experts-share-their-mobile-success-secrets</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b907e5af173a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/interceptd/13-experts-share-their-mobile-success-secrets-b907e5af173a">13 Experts Share Their Mobile Success Secrets</a> was originally published in <a href="https://medium.com/interceptd">Interceptd</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Solutions to Mobile Ad Fraud]]></title>
            <link>https://medium.com/interceptd/solutions-to-mobile-ad-fraud-460c6eb3ac84?source=rss----51929d39d414---4</link>
            <guid isPermaLink="false">https://medium.com/p/460c6eb3ac84</guid>
            <category><![CDATA[fraud]]></category>
            <dc:creator><![CDATA[Devrim Çavuşoğlu]]></dc:creator>
            <pubDate>Fri, 08 Feb 2019 13:57:37 GMT</pubDate>
            <atom:updated>2019-02-08T14:00:15.204Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*yryDGA5dsE4cxo3ZHnyOAQ.png" /></figure><p>Fraudsters have been trying to occupy the fields where they are able to join the game. Not only in mobile advertising technology field, they have been laying siege to every breaching field they find. Their siege weapons are various, but most of the ad tech companies are aware and they have been fortifying their fortress. The strength of the fortress depends on the strength of the whole chain, and the strength of the rings.</p><p>To understand the chain better refer to <a href="https://interceptd.com/effect-of-ip-while-dealing-with-mobile-ad-fraud/">this</a> post. The <a href="https://interceptd.com/">fraud prevention tool</a> is a ring of the chain which deals with the fraudulent traffic. Since the chain mostly does not contain fully clean traffic, it is expected to be broken by fraudsters easily.</p><p>To prevent this, the rings that are part of the chain connect to an anti-fraud tool to strengthen the chain and make it rigid enough that fraudsters cannot break, or let’s say the chain probably have deficits, but it will never be broken since it is continuously improving and reinforcing the chain. Trying to catch fraudsters is like endless hunt.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pcEv-9BKJMxxJfrTBsfHPw.png" /></figure><p>If you are not using an anti-fraud tool, or not connected with one, you are probably in a danger zone.</p><p>The rings in the ad tech chain have different payment models among the other rings and the payment model may also differ within that specific ring for different customers. The payments model usually based on impressions, clicks, installs or events, or mix of these. Some use different payment models like cost per block or cost per rejection. The ones who are curios about the payment models can search on the web for the payment models used in ad tech, I will stick with the topic of this post. I will try to explain what are the solutions for the fraudulent activity hidden in the traffic of the chain.</p><p>As you may have presumed the traffic load on such a network is enormous, it brings us (and possibly most of the rings) loads of data every second. This high volume of traffic/data forces the solutions to be AI systems based on the data. EDA plays a key role for retrieving the information what is going on the chain. It is inevitable to not use ML/DL and other proper statistical models at this point. Analysts can run an exhaustive EDA on the data and try to make points, running anomaly detections and detect possible frauds.</p><p>Let me tell you that in this post I will focus on frauds in the scope of <strong>CPI (Cost Per Install)</strong>, the solutions to the frauds can be visualized with the chart below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*c40_oR1Ve6ZHpQgeNy32JA.png" /></figure><p>Nonetheless, we are not revealing all features used, that is not appropriate for two reasons. First, the fraudsters would probably improve theirselves to penetrate those, or would try to work around the specific cases. The latter will probably not be sufficient enough against strong AI based detections. The second reason is that they vary a lot. We grouped the ones not mentioned in the figure as <em>‘Other .. Features’</em>. These features can be extended and they vary among different ad tech companies.</p><p>‘Other Features’ is a part of feature extraction case in the scope of the data the people have, obviously the data shared through the chain is mostly common for most of the rings, but different rings possible have more data which they gather by theirselves. I will not go too deep in the detail for the sake of the topic of this post.</p><p>By the way, note that the install column does not necessarily belongs to post-install. Most of the time the blocks at that level are mix of both pre and post install cases.<br><strong>How can install level be counted in pre-install ?<br></strong>Pre-install does not strictly covers only the part before install occurs. The thing is that the click and the install flows are in reverse order (refer to the post I mentioned earlier for an illustration and for better understanding the flow.). You can imagine the flow like an electronic circuit. As in its simple form, fraud tools are like switches in the circuit between the bulbs, if fraud happens open the switch, and the flow stops, so the bulb after the switch cannot know if the install occurred or not. This particular case is needed to be explained under another topic and post though. Just think like that, in the flow, if a fraud tool detects a fraud and blocks it even if the information collected by that fraud tool after install, it can stop sending that install information to the next ring of the chain, so the flow stopped, no payments have transacted for fraudulent traffic. Note that, the flow can be stopped by a fraud tool after an install actually occurs.</p><p>There are pre-install solutions as they are illustrated in the chart above, and it is one side of the fight the fraudsters where it takes places before an install occurs. Those solutions prevents fraudster to reach the install phase and steal money from advertisers.</p><p>Large amount of the fraudulent activity is caught before the install occurs, they are stopped at impression or click level. This helps install level shield being attacked by less amount of frauds.</p><p>Some of the fraudulent activity, however, can pass through the shield of pre-install phase although the volume is much more smaller relatively.</p><p>The rest of the fraudulent activity is dealt at the install level. The fortress of Interceptd is almost invulnerable to the fraudulent activity at this level. The defensive strategy for an anti-fraud tool is to detect and block the fraudulent install without losing the clean activity during the block phase. It is important to have fraudulent activity blocked, and having not blocked non-fraudulent installs is important as well as the former one. Most of the time, advertisers miss that point and only consider the performance of fraudulent blocks.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*DDD7Cc9ZaqMpSfGC7fy3cQ.png" /></figure><p>All in all, an anti-fraud tool must handle the traffic by allowing the <strong>highest possible non-fraudulent</strong> traffic and <strong>lowest possible fraudulent</strong> traffic. The former and the latter can be %100 and %0 at the maximum efficiency, any of anti-fraud tools have yet to attain this, unfortunately. This is almost impossible due to the complexity of the field, you have to consider many things while fighting against fraud. I say impossible because even if you provide %100 non-fraudulent traffic, it is highly likely that you blocked some clean traffic as well, vice-versa holds as well. The efficiency depends on the strength of the anti-fraud model (any data based model), and the model strength depends on the features you use and powerful detection of the past fraudulent activity.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=460c6eb3ac84" width="1" height="1" alt=""><hr><p><a href="https://medium.com/interceptd/solutions-to-mobile-ad-fraud-460c6eb3ac84">Solutions to Mobile Ad Fraud</a> was originally published in <a href="https://medium.com/interceptd">Interceptd</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Handle high load traffic with AWS and Go]]></title>
            <link>https://medium.com/interceptd/handle-high-load-traffic-with-aws-and-go-d89431965399?source=rss----51929d39d414---4</link>
            <guid isPermaLink="false">https://medium.com/p/d89431965399</guid>
            <category><![CDATA[scalability]]></category>
            <category><![CDATA[aws-lambda]]></category>
            <category><![CDATA[go]]></category>
            <category><![CDATA[aws]]></category>
            <dc:creator><![CDATA[Arda Güçlü]]></dc:creator>
            <pubDate>Wed, 23 Jan 2019 12:03:24 GMT</pubDate>
            <atom:updated>2019-01-23T12:03:23.948Z</atom:updated>
            <content:encoded><![CDATA[<p>Scalability is the one of the major and prevalent concern of modern software industry. There are huge amount of data and traffic and if you want to survive in this continuously growing/changing/evolving industry, you should be equipped with the right tools in your tool bag. Moreover, if your business runs in the Ad Tech industry, huge amount of traffic and data come into tangible existence. That’s why, in this blog post, I would like to give some details about how we made architectural decisions to handle 10Ks of requests per second.</p><p><a href="https://interceptd.com/">Interceptd</a> is a mobile fraud detection tool in Ad Tech industry. Customers send their clicks/installs/events, etc. via Rest APIs and they are warned via Rest APIs when the Interceptd filters, detects and prevents any type of fraud. Thus, integration between two products should perform well even in the high load traffic.</p><h3><strong>First Phase(Performance Issues)</strong></h3><p>At Interceptd, our Rest API returns responses within ~20.0 milliseconds as shown below;</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pyOxqK6dQ2pCWDojaWpOHw.png" /><figcaption>Interceptd’s Rest API Response Time in milliseconds</figcaption></figure><p>And our customer has lots of tasks implemented in PHP and running on AWS EC2s and whole traffic is drained into that tasks. In our initial step of integration, they had sent their whole traffic to Interceptd’s Rest APIs synchronously in a single thread. After a while, we realized that their task count increased 4 times, due to enabled auto scaling feature. This was not a cost effective and perpetuated in the long term and we did not tend to change tasks from single thread to multi-threaded application, due to the already running tasks in production environment should have worked without being changed its architecture drastically . Nevertheless, we tried the performance of multi-threaded version of the integration and anyway, it did not perform well (~10K/minute).</p><h3><strong>Second Phase(Loose Coupling with S3/SNS)</strong></h3><p>The first attempt and decision we were made was separating integration from the core parts of the tasks to assure the loose coupling. Thanks to AWS S3, whose response time is very low and resilient service, we drained whole traffic to seperate AWS S3 buckets and SNS topics have been subscribed to S3 buckets’ creation events (Current architectural diagram shown below). Furthermore, due to the tentativeness of the traffic, we decided to upload traffic 6 times per minute in bulk files. Hence, we could have a chance to manage the flow uniformly.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/742/1*QtFLGLSK9SkwOxl6ZmUxhw.png" /><figcaption>Customer-Interceptd Integration Architecture</figcaption></figure><h3>Final Phase(<strong>AWS Lambda/Go)</strong></h3><p>After decoupling tasks and integration feature, we had a chance to choose appropriate programming language and stack to overcome the performance and cost effective scalability problems. 10Ks of requests per second need concurrency and Go’s builtin concurrency seemed a good choice to tackle such high volumes of data and traffic. Moreover, AWS Lambda can be triggered from SNS events (because of its uncountably large scalability and cost effectiveness, not because of the serverless hype in the software industry). It performs well as shown below diagrams;</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*eImjKlmzQxlmEiIK_jc-ZQ.png" /><figcaption>Traffic load per minute</figcaption></figure><p>And it can be clearly seen in the below diagram, AWS Lambda with Go completes its process (~1200/sec) within ~2 seconds.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7AvceheoFsg26SBT1r_lnA.png" /><figcaption>AWS Lambda with Go</figcaption></figure><blockquote>One point should be beared in mind that if you invoke all traffic in different goroutines, AWS Lambda throws “<strong>socket: too many files</strong>” error, because you can not increase ulimit in AWS Lambda (limit of maximum connection). Therefore, you have a limit of goroutines you can create. For example, If you try to send 1500 requests in different goroutines, you will get an error. We send these 1500 requests with dividing it 10 within 150 goroutines.</blockquote><h3><strong>Conclusion</strong></h3><p>Go is a powerful programming language with its concurrency support(readability, fast execution time, etc.) and when it is combined with the AWS Lambda, it performs well for tasks have low tolerant to latencies with fine-grained goroutine counts.</p><p><em>You can reach out template Go code which consumes SNS events triggered from S3 via Github link below.</em></p><p><a href="https://github.com/Netvent/interceptd-api-integrator/"><em>https://github.com/Netvent/interceptd-api-integrator/</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d89431965399" width="1" height="1" alt=""><hr><p><a href="https://medium.com/interceptd/handle-high-load-traffic-with-aws-and-go-d89431965399">Handle high load traffic with AWS and Go</a> was originally published in <a href="https://medium.com/interceptd">Interceptd</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What is Mobile Ad Fraud?]]></title>
            <link>https://medium.com/interceptd/what-is-mobile-ad-fraud-7fc8524a6ef0?source=rss----51929d39d414---4</link>
            <guid isPermaLink="false">https://medium.com/p/7fc8524a6ef0</guid>
            <category><![CDATA[fraud-detection]]></category>
            <category><![CDATA[advertising]]></category>
            <category><![CDATA[mobile-advertising]]></category>
            <category><![CDATA[mobile-marketing]]></category>
            <category><![CDATA[ad-fraud]]></category>
            <dc:creator><![CDATA[Ecem Öz]]></dc:creator>
            <pubDate>Tue, 18 Dec 2018 14:58:12 GMT</pubDate>
            <atom:updated>2018-12-18T14:58:12.024Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*sc0RXS6XAxZHCAuf35ly8w.jpeg" /></figure><p>Mobile ad fraud has risen dramatically and, unfortunately, according to estimates regarding 2019, it will continue to rise and be a challenge for mobile marketing industry. The mobile growth offers rich opportunities for mobile advertisers along with fraudsters.</p><p>There are different types of mobile ad fraud and fraudulent activities are evolving day by day as. To prevent fraud, marketing people need data, people and more sophisticated technological methods. In order not to get affected from fraudulent activities, it is a necessity for mobile marketers to clearly understand and be able to analyze the funnel of mobile advertising from first impression to post-install.</p><p><strong><em>Mobile Ad Fraud…</em></strong></p><p>It is basically an activity of stealing from ad budgets of advertisers carried out by fraudsters and it can take various forms like click spamming, click injection, hijacking, device farm and so on.</p><p>The amount of money lost to fraudsters is at a fearful rate: IAB estimated that for every $3 spent on digital ads, $1 is lost due to fraud <em>(adage.com, 2015)</em>. In order not to lose a penny from the marketing / advertisement budget, there are things that can be done and tools that can be used to avoid fraud but before anything else, detection should be the priority and then necessary actions to prevent fraud should be taken rapidly.</p><p><strong><em>Mobile Ad Fraud Detection &amp; Prevention</em></strong></p><p>For detecting, <a href="https://interceptd.com/fraud-will-cause-your-cpi-costs-to-increase-soon/">identifying and analyzing fraudulent activities</a> on mobile marketing campaigns, reports can be great resources. Reports can be counted as the creator of the “data” part which is needed to prevent fraud. In these reports, there should be data (metrics) regarding time-to-install, click to install time, IP address, user agent and so on. The faults in these metrics should be monitored periodically. Monitoring gives the opportunity to immediately recognize fraudulent activities and take the most effective action for keeping hold of the budget.</p><p>With the help of a tool which is a defender against fraud, you can customize rulesets and create alarms based on metrics such as click to install time, install-to-event rates, identify, and blacklist and block sources automatically. If you are the advertiser, it will let you detect fraudulent traffic in near real-time, before click reaches the store page. If you are the publisher, it will show why or when a sub-publisher is blocked.</p><p>It is essential to bear in mind that follow-ups regarding the metrics is needed as much as preventing fraudulent traffic because once fraud is prevented for mobile advertising activities, it does not mean that it will be prevented forever. However, if you are working with or decide to work with a mobile advertisement fraud prevention tool, you should be able to demand transparency regarding reporting and efficiency. Transparency gives you the opportunity to clearly see whether your mobile advertising budget is spent properly.</p><p><strong><em>So, …</em></strong></p><p>In order not to let fraudulent traffic get in the way of your mobile advertising campaigns and to have successful results for your marketing efforts, you should not be contented with the solutions your MMP provides. A strong and efficient fraud prevention tool, which lets you identify the best channels to spend marketing budget, make sure if the MMP used is enough to combat mobile ad fraud and increase the profitability of the app with better ROAS is a necessity in the mobile marketing environment of today where fraudulent activities are increasing in very critical rates.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7fc8524a6ef0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/interceptd/what-is-mobile-ad-fraud-7fc8524a6ef0">What is Mobile Ad Fraud?</a> was originally published in <a href="https://medium.com/interceptd">Interceptd</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[App Install Fraud — The Gangsters, Their Methods & Victims]]></title>
            <link>https://medium.com/interceptd/app-install-fraud-the-gangsters-their-methods-victims-1a3246a41925?source=rss----51929d39d414---4</link>
            <guid isPermaLink="false">https://medium.com/p/1a3246a41925</guid>
            <category><![CDATA[fraud]]></category>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[advertising]]></category>
            <category><![CDATA[mobile-advertising]]></category>
            <category><![CDATA[fraud-detection]]></category>
            <dc:creator><![CDATA[Elif Çetin Ölker]]></dc:creator>
            <pubDate>Fri, 07 Dec 2018 13:37:36 GMT</pubDate>
            <atom:updated>2018-12-07T13:37:36.790Z</atom:updated>
            <content:encoded><![CDATA[<h3>App Install Fraud — The Gangsters, Their Methods &amp; Victims</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F_4AVSKU6RZk%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D_4AVSKU6RZk&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F_4AVSKU6RZk%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/4552aa1209776fed86de431065ae8aab/href">https://medium.com/media/4552aa1209776fed86de431065ae8aab/href</a></iframe><p>Methbot is a huge bot farm that generated millions of dollars by targeting the premium video advertising ecosystem. They were long lasting and very difficult to get rid of. And here is the worst news: there are many groups like Methbot.</p><p>As a result of the attacks of the groups like Methbot, companies regardless of their sectors started to spend some time understanding why and how they are/can be victims of these bad actors. So they wanted to understand the true definitions of some of the terms about the fraud. There are many words to know about ad fraud from click spamming to click hijacking and click flooding.</p><h3>Who are the victims of app install fraud?</h3><p>It is quite easy for us all to think app install fraud doesn’t affect our job or us. But here is the fact: it does affect. Everybody is a victim of app install fraud. The statistics show that app install fraud in Q1 this year generated 800 million dollars, which means that 11.5% of all paid installs are fake installs. These numbers have increased by 30% comparing the last year. And it is actually higher than these numbers up to 20–30%.</p><h3>Who are the fraudsters?</h3><p>They are real people. One of them is a publisher serving real ads within real games; the other is running global ad networks. These two people who are fraudsters have got install bases of their own apps that allow them to implement and run malware on people’s smartphones.</p><p>But, we need to notice that there are hundreds and thousands of people just like these two guys. There are also different types of frauds that we can divide into two, to begin with: attribution fraud and download fraud.</p><h3>Main Methods Of Attribution Fraud To Gain Credit For Installs</h3><p>Attribution fraud is where the bad actor is stealing a real download. The fraudster is stealing the credit or the attribution either from organic or another paid source.</p><p>Below you can see the main methods that the bad actors are using to gain credit for installs that they should not be credited for. These people steal millions of dollars from the companies and this is why it is important to bring this subject to all of our attention.</p><h3>What is click hijacking?</h3><p>It aims to steal another publisher’s legitimate publisher/network paid install. It requires access MMP, the Mobile Measurement Partner, to real, live attribution click URLs from their legitimate side of their business. They install malware on a consumer’s smartphone and what the fraudsters are doing is to listen to when the legitimate publisher is serving an ad through this malware. As soon as they hear that legitimate click, they quickly send their own click URL being the last click and win the attribution and the CPI payout.</p><h3>What is click spamming?</h3><p>It aims to steal an organic install for big apps. The best targets of this type of hack or attack are the developers that have millions of organic installs. What the fraudster does is to send clicks from their malware without the knowledge of the user. They do this with millions of devices so it is a statistical model.</p><h3>What is click injection?</h3><p>It also aims to steal organic download, but for any sized app as long as they are running paid media. It only works on Android and what the bad actor is doing through their malware is listening. Google makes a broadcast when the installation is initiated. Because the malware is configured to listen to Google Play to hear when an app install has initiated, this gives the fraudster the time and opportunity to send a click to the attribution company. When it is installed they get credit.</p><h3>What About Download Fraud?</h3><p>On the other hand, download fraud is when the bad actor is actually faking a download. The download may never have even happened or it wasn’t a human.</p><h3>SDK Spoofing</h3><p>To understand download fraud, we need to understand the dark arts of SDK spoofing. It requires a malware install base to fire the events from real devices or a device farm along with a very detailed knowledge of the attribution company’s SDKs to fake it according to David Philippson. What bad actor does is to send a fake click to the attribution company from a real device. They then know the content of the API call to the attribution company. So the company believes that an install has occurred, which means they succeed in fooling the attribution companies and the advertisers into paying for downloads and user that don’t even exist.</p><h3>Device ID Reset &amp; Device Farms</h3><p>What is interesting about download fraud is that the fraudsters collect real phones like our old smartphone that we don’t know where they are now. They collect these phones and update the operating systems, and finally implement VPNs to fake IP addresses and locations.</p><p>Then, when they see a new install on their legitimate business on their network or in their publisher app, they just deploy the click URL in their device farm and get the credit from the attribution company. Later on, they delete the app and reset the device ID and do this again and again.</p><h3>Who Is Being Fooled?</h3><p>The advertiser pays and trusts the attribution provider while the legitimate publisher and networks are forced to bill from the attribution partners’ numbers. All of the fraudsters’ efforts are going into fooling the attribution companies. To be clear, they are succeeding.</p><p>They are actually fooling the accountant. What does this mean?</p><h3>Accountant and Auditor</h3><p>If the attribution is the accountant, then <a href="https://interceptd.com/">fraud prevention, detection, and reduction is the auditor</a>. With all industries, it is very important that the accountant doesn’t audit their own books. So it is not right for attribution companies to license separately fraud prevention modules. It is absolutely essential that they are built in to their core offering to insure they haven’t got a conflict of interest if they have the capability to reduce fraud and that should be a part of their core offering.</p><h3>Final words and recommendations</h3><p>So, if the attribution companies have the capability to prevent fraud, they can maintain their commitment to advertisers and to publishers. Choose a reliable accountant and an independent third party for fraud solutions to detect, report and reduce fraud.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1a3246a41925" width="1" height="1" alt=""><hr><p><a href="https://medium.com/interceptd/app-install-fraud-the-gangsters-their-methods-victims-1a3246a41925">App Install Fraud — The Gangsters, Their Methods &amp; Victims</a> was originally published in <a href="https://medium.com/interceptd">Interceptd</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>