How Stolen Goods End Up On Amazon, eBay And Facebook Marketplace

A Louis Vuitton with every item swiped from the shelves. in San Francisco's Union Square, A nearby Nordstrom robbed by 80 people. the next day, Similar sprees on Chicago's Magnificent Mile and Rodeo Drive in LA. They're actively looking for about 30 people who are trying to break into other stores. Theft is nothing new, but a recent spree of coordinated, big scale robberies has law enforcement and retailers warning this is not simple shoplifting by those in need. The police say this is a professional crew connected to a six state crime spree. It's organized retail crime by professional crews and it's on the rise for one big reason. What fuels this as an enterprise is the ease of reselling stolen merchandise on online marketplaces. While punishments for shoplifters are hotly debated, there's growing consensus around a solution that holds an entirely different group accountable: the online sites where stolen goods are sold, primarily Amazon, eBay and Facebook marketplace, which all say they're already doing a lot to stop this. EBay is not a place to hide yourself and try and offload some of this stuff. It used to be. you have to go to a pawn shop. You have to go find a place to sell it at a flea market. Now you have the ability to ship it from your home: 20 major retailers, including Home Depot, Best Buy, Walgreens and Kroger, sent a letter to Congress in December asking them to crack down on online marketplaces by requiring stricter verification of sellers. My ask for the online community is: just do more right. CNBC went to Home Depot to see the new tik and decades-old solutions it's now using to stop theft. Generally, anything over $100 is typically behind the bars in this case. Here's what major retailers and Amazon, Meta, eBay and legislators say they're doing to fight the big problem of stolen products being sold online. Welcome to the Home Depot. Please enjoy your shopping experience. Surveillance from this tower of cameras watching over a parking lot in Hiram, Georgia, is one way Home Depot is trying to prevent and track down would-be thieves. We're testing a couple of other things like license plate recognition. If we could know in advance that John Doe, who potentially is a repeat serial organized retail crime offender, we'd know when they're in the parking lot before they walk into the store, We can prepare. Scott Glenn has been in loss prevention for 26 years. Since he joined Home Depot four years ago, he says losses from organized retail crime, or ORC, have grown at double digit rates. It's in the billions of dollars a year. So now Glenn says, Home Depot is spending millions to prevent theft, with new ways and old. This area of the store per square foot is the largest loss area of the store in just total dollars. So what we've had to do, unfortunately, is put some barriers in between the customers and the product. Something new is called Point of Sale Activation. A Bluetooth enabled chip embedded in some power tools keeps them from working unless they've been scanned at a register, And if you bought a stolen Milwaukee product that was activated with the POSA tiknology and you brought it home, it wouldn't turn on Correct. Another new example are carts that lock up at the exit If they haven't been discretely scanned by going through a checkout lane. What it's going to do is the cart's going to lock up on me. Home Depot has hundreds of cameras in each of its stores and it's experimenting with tiknology that would allow them to keep track of items as shoppers put them in their baskets. Partikle recognition, computer vision, whatever you want to call it. We actually do have it in a couple of stores in terms of a pilot. In April, a San Jose Home Depot burned to the ground after a man attempting to steal a cart full of tools set a fire as a diversion. But even with extreme cases like this, it can take years for law enforcement to determine whether it was part of a larger crime ring or just an isolated, desperate incident, which makes tracking theft trends especially tricky. It's heavily dependent on self-reported data from retailers, and so this is a very, very squirrelly thing to actually be trying to track. Retailers say a total of $68.9 billion of products were stolen in 2019.. In 2020, three quarters said they saw an increase in organized crime and more than half reported cargo theft. Think those thousands of boxes strewn along the LA rail tracks after they were swiped from slow moving shipping containers. Some big chains say organized theft is the reason they've closed stores or limited hours, but they're also facing steep competition from online shops and a drop in foot traffic because of the pandemic. Retailers have one goal- one fiduciary obligation to their shareholders right, And that's to increase profit and to increase shareholder value. Critiks like Alec Karakatsanis of the Civil Rights Corps say that's motivating retailers to make the problem sound worse than it actually is. By drawing attention to dramatik thefts, they redirect public sentiment toward policies that better protect their bottom line, like increased punishments for petty shoplifters, not just million dollar crime rings, The collateral damage from their campaign to stoke fear about retail theft is going to have profound effects and more severe punishments for the poorest people in our society and for people who are stealing from grocery stores and drug stores for basic necessities of life. I asked the CEO of Walgreens: Why do you have underarm deodorant under lock and key? They said because it's so easy to pull your arm on the shelf and sweep everything into a bag. They put them online. They get certain quantities of them. It's a massive operation. Prior to coming to Home Depot, I wasn't really a believer. I thought maybe it was a little bit overblown. People were making too much of an issue out of it. Coming to Home Depot, I've seen it in real life. I've seen it growing. I've seen the impact of it, And so it is not only growing over the last five years. I would say it's grown incrementally over the last two During the pandemic. While timely numbers are hard to pin down, Homeland Security says its organized retail crime investigations are on the rise. Last year saw far more arrests, more indictments and more value in stolen goods seized than in 2020.. As long as there are resale markets that are not held accountable for the sale of stolen goods, then this kind of conduct is only going to get worse. Here's how it works: A criminal network hires an individual or crew called boosters. They can be professional thieves or even victims, labor trafficked into the US from other countries. After robbing a retailer, they turn over the stolen goods to someone waiting nearby, called a fence. Boosters often hit multiple stores a day and they can even cross state lines and ship goods to a fence. The fence pays the booster in cash or drugs, usually about a quarter of the retail value. The fence then takes the haul to a home or warehouse where a cleaner removes anti-theft devices or markings, then sends it to a larger criminal network which resells it, usually online. The National Coalition of Law Enforcement and Retail, or CLEAR, estimates that more than $500 billion in illicit, stolen and counterfeit goods are sold on third party marketplaces like Amazon each year. Very recently, a CVS manager was assaulted and remains in serious condition. These incidents are not uncommon. Reported violent events at CVS have doubled in the last year. Ben Duggan testified about this at a Senate hearing in November. He's the president of CLEAR and also leads investigations at CVS Health. There are far less obvious dangers to this crime, including infant formula. That's a favorite target of these criminal organizations. They ignore or manipulate the expiration dates and they're not storing it at proper temperatures. It compromises the product integrity and is endangering the health of an inf.

Unpacking an Alleged $75 million Ponzi Scheme

(bright music). - On today's episode of The Unofficial Shopify Podcast, we discuss an alleged $75 million fraud. Okay, so two weeks ago I mentioned on the show as a top of the show house keeping note, I said: look, it's January. a lot of people are jumping into a new business, scaling up a business or are just generally looking to invest in themselves in their own business. this also means there are a lot of con artists that come out around this time willing to sell you the road map in a get rich quick plan, right. And last January there was really a phenomenal deep dive artikle into Guru culture around e-commerce and Amazon and those experiences that we discussed on the show. - It was literally like a year ago. it was an artikle in The Atlantik, like one year ago. today that was the episode that we dropped. - So today, purely coincidentally-- - Little did we know two weeks ago, when we mentioned it again in the morning, someone was being arrested. (laughs) - Oh, really, Were they arrested? - Well, I don't know. - No, it says his assets were frozen. - All right, fine, He's not in jail. - Okay, So there is-- - We almost were involved in a Ponzi scheme, An alleged Ponzi scheme? - Yeah, let's make it clear, no one's been convicted here. - Essentially, every verb we use in this podcast put the word allegedly in front of it - Yes - 'Cause we're gonna get sued. - So last spring, So I believe it was May. I could look it up, but it was in the spring-- - It was like April or May. yeah - Yeah of 2019.. Someone reached out to me- I believe they were a podcast listener- and said: "Hey, listen to your show. "and I've got some friends who are involved in e-commerce. "and do you wanna go get lunch with us? "It'll be fun". I said, "Sure, why not? "I'll go get lunch". And it was at Gibsons - It was at Gibsons downtown, Which is, if you're not in Chicago, that's like very fancy, nice, high end steakhouse downtown - And not just downtown at Gold Coast, Which of the downtown area in Chicago, Gold Coast is like premium. Gold Coast has a Bentley dealership (Paul laughs). And so the folks I met with were Ken and Kerri Courtright, who are from their company is Today's Growth Consultants, and at that time they were known for running The Income Store - Which, I gotta say, calling your business The Income Store is the biggest fucking con job name I've ever heard in my life. - Well, so The Income--. - Get rich quick ink. - Yeah, So The Income Store is an odd- was an investment vehicle. You, as an Income Store client, invested $100,000 minimum, I think, is what they said-- - I'll run down what the business was. - Oh, you got All right. - All right, I got all the stuff. I'll run down what the business was, and then Kurt could give us his man on the ground perspective from his lunch with these folks. - Okay, hit me. - All right, So The Income Store, what it was is years ago. it was a network of blogs that was essentially doing Google AdSense, that was running AdSense on the blogs. They had various affiliate links and that sort of stuff, and that was how these blogs generated money. and what you would do is you would give Ken and Kerri Courtright or The Income Store $100,000 and they would buy a blog for you And then they would manage the blog and do all the work and then they would keep half the revenue as like a management fee and then they would send the other half of the revenue to you and through this you would just generate passive income that would wildly outperform the $100,000 you gave them and as part of this, they had guarantees and they personally guaranteed that you would get 15% return annually on your investment. - This is the first red flag. - This is the first red flag: Guaranteed returns. if anyone promises you guaranteed returns, they are taking your money. They're Bernie Madoff. That is a scheme. do not listen. - Yeah, so number one, Remind yourself: all investments carry inherent risk. So anyone who makes a guarantee on an investment, unless it's like a savings account doing 1% - Unless it's literally T bills, that is like US Treasury bonds, and even those are not guaranteed because the government of the United States could collapse. Those aren't guaranteed. So they promised you 15% a year and they would, literally would send you that's $15,000,. they would send you a check for $1250 every month And obviously, if the site performed higher than 1250 every month, you would get a check for higher than that And, according to their claims, you actually were the owner of that website. so if you wanted to cancel the deal or do what ever at anytime, you could keep your website and walk away from their management. So obviously, these sites- and I looked at some of them- they were really shady. They were like weird review sites, Weird health focused sites, And I'm guessing they used a lot of black hat, grey hat SEO stuff to boost them up in the Google search rankings. - Well, I've not necess, so that's pure speculation. - That's pure speculation. - We do know that they hired a team of copywriters to create content for these blogs - So that's what they're doing. yes, and they hired copywriters, obviously paying bargain-basement-internet-copywriter-needs-a-job prices, And so let's think about this step that has happened. You have just given them $100,000 and all they are required to do is write you a check for 1200 bucks every month. So where does the other $98,000 go? - That money was used ostensibly to purchase the website that you have invested in - Which you would hope so if your purchasing- let's tok about this. if you're purchasing a website for $100,000, you would probably expect that website to generate, say, $30,000 annually. That's like-- - It's conservative. - It's conservative. I mean. the site is producing probably between 30 to $50,000, or. and then they declare that they were such gurus at SEO and monetization that "well, don't worry that the website we bought "doesn't generate that much money now. "We're so good at management? "we're gonna supercharge the sites revenue? "and it's gonna be worth well over $100,000 "when we're done with it". - And they had a history of managing internet or growing internet business prior to this, prior to The Income Store - I don't know what they did before this. I know they declared bankruptcy in 1999.. - I think a lot of people did, though - Not in 1999.. Things were going good then - [Kurt] Oh - So that's what they pretty much were doing, and here's a letter that they sent in 2017 to their investors: " To start, as many of you know, from 2012 to 2016, "we battled Google algorithms "and were in a constant dance "of two steps forward, one step back? "and, in addition, as many have received, "if we failed on a site, "we'd always replace it with another. "In 2016,, we made a hard charge at Facebook driven assets "as they providing us traffic and revenue "the likes of which have never been seen before". So what they were doing is they were counting on gaming Google to make the money, and then, anytime the Google algorithm changed for their kinda trash content blog network, they then switched to they would just crater whenever Google changed anything, and then they moved onto Facebook in 2017.. - Okay, So how long had this been going on? - I mean, this has been going on for at least since 2012.. - Oh really - Yes - Okay - And so the same thing happened. Facebook then changed the. then the reasons for, "Oh, the reason that these sites aren't generating things"? well, was because people aren't getting the returns they expected. "Oh well, it's because Google kept changing the algorithm", So we gotta somehow--. - But it was a guaranteed payout. - But it was only a guaranteed payout of a tiny amount of money, and these people have access to the backend and see the revenue that's being generated, And the revenue that's being generated is not commiserate at all with the $100,000 that they gave these people. - Okay - So, and it's always again, as with th.

More:How To Sell Your Shopify Store In Under 5 Minutes | Dropshipping 2022

Detecting eCommerce Fraud with Linkurious

so let's start. thanks again, everyone for joining today's webinar. I will show you a treat today, or you can detect ecommerce fraud using linkers- and no hood. so a quick introduction to top. so my name is marcin, I'm working as business developer here at linkers and I am joined today by my colleague, Elise from the marketing team. so the agenda of today's webinar is the following: after quick introduction, we will quickly speak about rock tiknology and then focus on e-commerce row. we will wonder, actually, why detecting ecommerce world is tiknically challenging. after that, we will explain the advantages of using linkers and your project. we will then ask a demo focused on e-commerce fraud use case. after this demo, we'll explain you how it works and finish by presenting you a real use case of one of our custom. finally, we will add quick questions and answer session and after I saw you, please note that you can ask all your questions during the webinar using the chat section and we will answer this again. good, let's start by a quick introduction about thinkers. the soul in church is a French startup test in Paris. we started in 2013 and today we are walking with a bit more than 200 customers. actually, most of our customers are working abroad. we do work with organizations like Cisco or NASA and also couple of governmental institutions like the French Ministry of entities or increased enterprise and linked use SDK products as analysts extract Eden insights from graph data. so what do we mean actually by unlocking the value of growth data, our customer, our data, actually in various formats like excel file, SQL database and so on. the first thing is like the model these data as a graph of entities, which is called nodes or relationships. this is called edges. they then move this data inside the graph database, like near project and neo4j s store and query large blocks. the liquor software makes these data and the power of neo4j accessible to non-tiknical users without easy-to-use interface. they can monitor, investigate and detect inside, even inside their grappa. so our customers use linkers to solve various use cases. for example, we have customers working on cybersecurity. there are several Reuters application, this is something you can see on the screen and the youth linker to detect and English to get suspicious activity buttons with similar data. actually, we have some of our customers were working on IT operations or enterprise architecture use case where, for example, differ from root cause, another analyzing or data lineage. we are also working with governmental institutions who have access to emails, transactions, phone call records and try to detect and investigate criminal networks. this is called intelligence. finally, most of our customers actually are banks, insurance and money transfer companies will work on ng money laundering and fraud. they are looking at connections between clients, their transactions, personal data, financial records and etc. the purple for the proposed- actually of today's webinar will be focusing on a specific kind of fraud called ecommerce fraud. I will now let my colleague, Elise tok about this in more details. Thank You, Martin. so, as mentioned, today's topic is about the graph based approach of e-commerce fraud detection and, Before we jump to the demo, we wanted to explore a bit this world and the challenges it rise or prevention and detection solution. so, first of all, when we speak about e-commerce phone here, we refer to any type of fraud scheme that uses Internet related means to present, for doing solicitations to prospective victims or to conduct, probably prevent, transactions, and those proteins unfortunately come in many forms. among the master current schemes we can find a friendly fraud where a merchant receives a charge back because the paddles are denied making the purchase or receiving the order and yet the good where often actually received. there is also promotion abused or phishing, which is a common practike of sending fake official email to steal sensitive our information. we can find immersion probe, when a foster offer goods at very cheap price but never actually ship them. we also have account takeover, when a fraudster is using another person account to obtain product and services. we also have a serious fraud which refers to any sales activity meant to generate commissions from an affiliate program or we shipping fraud, where fraudster recruits a person to package and receive merchandising person previously with stolen credit card. and finally we are identity theft, which is the fraudulent achilles- acquisition of sensitive personal information. and those schemes are not the only ones. dozen exists in over the years, we've seen that froster keep renewing the tactiks and, along with new types of fraud, we have also witnessed an impressive increase of fraud entities. actually, in 2016, large e-commerce mass merchants lose up to 1.4 percent of their revenues to fraud in average, which actually sounds four billion of dollars worldwide. and there are several five factors driving the productivity. the first one is kind of obvious if the he corrects growth. it has become mainstream and it's forecast to expand during the next few years. notably, we are friends rushing into new international markets. so frightened example: last year, ecommerce sales reached one point nine trillion dollar, and it continued to expand rapidly. the second factor that has benefit to froster in the past years is the adoption of new tiknologies and practikes such as mobile payments, bitcoins or cloud services. those actually offer new methods. Apple fraudsters to bypass existing security measure and to Colonel covered their tracks. and finally, we've seen fraudster organize themselves in networks and take advantage of it. they use the same resources. they buy active data from large bridge and organizing to groups across the globe. for instance, this March, the District of Columbia arrested like nineteen people taking part in various international online fraud activities. that has led to more than 13 million dollars of losses, and one of the indictment was an online vehicle fraud where the partikipants falsified advertised car for sale that they did not whole. so, faced with this situation, ecommerce merchant apps turn to the fraud prevention solutions, and those web probe detection software, or sometimes cloud-based services, run background processes that can scan transaction and score them based on the possibility of fraud. today, many solutions follow a layered approach that assesses activity at multiple levels and use multiple analytikal approach. most of them rely on the three layer displays here: inform centric layer, navigation centric layer and channel transit layer. so for the first one part of the software actually strands information. that encompasses clients of applications, device art, ID or geolocation. the second layer controls include real-time dynamic capture of accounts and customer activity and this is actually used to build a customer profile to determine what is normal and what is not abnormal behavior to disperse to our customers. and the third layer: you open channel centric and monitor entity behavior across one specific channel. but for an efficient fraud detection we actually have Gartner Enterprise advising us to add two more layers to the system. so the layer four and five are about connected analysis and the do require tiknology that does not rely on a relational database model, for instance. the fourth layer is cross channel and is dedicated to the analysis of abnormal behavior across multiple channels. the fifth one is entity link analysis and it's all about analyzing relationships to detect organized crimes and fraud rings. so tiknology using only the three first layers can actually face some tiknical limitation. so first, the monitoring and detection system cannot keep packed with new products or news, new schemes. the structured is simply too rigid to allowe the easy adoption of new rules or of new data and suddenly those syst.

More:Sell These 8 Winning Products in October 2020 (Ep. 208)

"Ecommerce Infrastructure: Evolving Performance, Security, and Reliability Expectations" by JetRails

we're going to continue with our next speaker speaking about e-commerce infrastructure 2022: evolving performance, security and reliability expectations. we have Robert Rand from jet rails about to join us and if you don't know who jet rails is, it's one of your hosting Partners providing the fastest fully managed e-commerce deployments for a Magento, open source, Adobe Commerce shop, where in the Amazon web services cloud and on dedicated servers. Robert it became well known in the industry as the CTO of Rand Marketing, where he helped Merchants Thrive and e-commerce. he now heads up alliances at jet rails and e-commerce hosting provider. ladies and Gentlemen, please give a nice warm welcome to Robert Rand from jet rails. hi, everyone. well, I'm really up here because it's not an e-commerce event until there's a nerd up on stage. so glad to have you all with us here today and very glad to be here myself. so you know I'm going to be toking about infrastructure and you know that's an arena that not everyone in this room is probably thinking all that much about on a day-to-day basis, but there's a lot that is happening, whether it's security, whether it's speed, and we've been hearing already today a lot about customer expectations and customer experience. no one wants to hit Arizona site. no one wants down time, no one wants slowness or latency experiences, and so I'm going to be diving in looking at that through a number of different lenses, including a lot of things that don't really relate to websites that, uh, that need web hosting or to the web hosting itself. but just looking at how do we meet customers today, where they they expect to be met, and, um, you know, overall, that the table Stakes have changed, so I thought we could open up, just, you know, thinking a little bit about, um, you know where things started and how much it's evolved. um, because there was a time when you could throw anything up there, uh, and people were were glad to have somewhere to start, um, they were glad to have the, the overall experience. um, that, uh, you know we are in constant flux and the industry is constantly changing under us. um, so, to start off, I'm going to tok a little bit about the security side of things, because that's always, I think you know, that's what makes a lot of the headlines. um, that's what, uh, what can really do a lot of damage also, um, and there's often things happening that we don't know or we don't think about on a day-to-day basis. so let's start off toking a little bit about fraudulent orders. um, show of hands. who thinks that during the pandemic that, uh, you know, the percentages of fraudulent orders went up, that that the numbers got worse? yeah, most of the room. anybody think that they went down? no one. so, interestingly enough, if we look at the data- and now I know that we've got some folks from signified in the room that really have a lot of expertise in this partikular space- but I ask my friends at Clear sale and, based on their data, actually the percentages of fraud have gone down. now, this for me, as a data nerd, is a little bit of a look at correlation versus causation and what's actually happening in the numbers. so, part of what's happened, as you've heard and a lot of you have experienced, the overall volume of e-commerce orders have gone up significantly, and what we're seeing potentially is that the fraud didn't go away. it just didn't grow at the same pace. so the same fraud challenges, by and large, are still out there. they're. they're just not necessarily growing exponentially, which is a good thing and something that I'm certainly very glad to hear. we are seeing some challenges arise that existed before, but in terms of trend lines, um, we've been seeing gift cards being uh part of those attack vectors more often now. so I actually know a merchant that last year lost tens of thousands of dollars because their gift cards were abused and by the time that they caught on to it- because that wasn't going through their normal fraud filters and things that they're just gift card numbers that are accepted through their checkout- a lot of orders had shipped and so a lot of damage was done. but if you think about it, even if you're selling- um, you know more generic uh, you know cards- there are cases where you might uh, otherwise you might get to go through a more traditional checkout process, um, and have some time before you ship the order. even if it's a few hours to catch on to fraud when there are gift cards involved, it gets a little Messier. so there are a lot of different nuances of how these attacks happen and what types of stores and what types of gift cards, but it's an interesting trend line. we are also, of course, seeing a lot more tax involved. Bots- you're going to hear a lot about Bots from me today. um, they're good ones out there. uh, you know that Google's crawling your website to help get you indexed in their search engines. there's lots of good things happening from Bots, but Bots also cause a lot of trouble, and so one typical attack that we see often now it's called a carting attack, where a list of stolen credit cards and debit cards is run against your checkout, trying to find ones that are working and what are goods, and so those sorts of things are happening more commonly and without the right systems, the right payment gateways, the right fraud prevention tools, we're seeing a lot. we're also seeing a lot of interesting conversations around personal identifiable- you know, identifying information. pii is something that you know. it's really started in the conversations around laws like gdpr. by the way, does anybody know which one of these is fake? can anybody guess? anyone have a guess? Virginia, I couldn't believe that someone actually named one of these CPA. they're all real, but I got you to look away from your screens for a minute, so that was great. um, what we're seeing on this map is that different legislatures around the country, they are passing laws, they are working toward catching up, and so States like California were leaders in providing, uh, this kind of protection for consumers to make sure that we, as businesses, are protecting their data, that we're keeping it secure, and so there are components of some of these laws, like right you know in in Europe, gdpr, the right to be forgotten, um you know, the right to to know what data they're storing, um, Etc. but from a security standpoint, what we're seeing is that PCI compliance was the gold standard in e-commerce- not perfect, but at least set that really important uh standard from the payment card industry to say, if you're going to accept credit cards, if you're going to accept debit cards, you have to have a certain basis of of security in place at all times and you're responsible for it. and there are uh, there are you know- issues that you're going to run into if there is fraud, if, if there is theft of the credit card data and you weren't doing your job. so what we're now seeing is that both individuals and states can also go after you. so this is making it more and more important to really think about what data you're collecting, where you're storing it, who else is getting access to it- and in a lot of cases that comes down to a supply chain. it's not just you and your employees. so what we're looking at actively is how many Integrations you have and what data they're getting access to. if I had to guess, most people in in this room that are actively running e-commerce businesses that have been up and running for a few years. you're probably, in some cases, using dozens of apps that get API access that are otherwise exposed through the website through JavaScript and and other coding, and so in a lot of cases, the security threats aren't necessarily about what your developer did on your website or, uh, you know, or something like a more direct threat, but it's what's going on behind the scenes. it's what apps are you using and and what's their security like. and in a lot of cases, you know, there are one-off issues. I've got up here: uh, there were some Shopify employees stealing data. um, I'm actually very glad to.

Detecting eCommerce fraud with Neo4j and Linkurious

so let's start. thanks again, everyone for joining today's webinar. I will show you a cleat today or you can detect ecommerce fraud using linkers- and no hood. so a quick introduction to start. so my name is marcin, I'm working as business developer here at linkers and I am joined today by my colleague Eddie's from the marketing team. so the agenda of today's webinar is the following: after quick introduction, we will quickly speak about rock tiknology and then focus on e-commerce world. we will wonder, actually, why detecting ecommerce fold is tiknically challenging. after that, we will explain the advantages of using linkers and your project. we will then ask a demo focused on e-commerce fraud use case. after this demo will explain you or it works and finish by presenting your real use case of one of our customer. finally, we will add quick questions and answer session and after I told you- please note that you can ask all your questions during the webinar using the chat section and we will answer this at the end. good, let's start by a quick introduction about tinkers. so linkers is a French startup test in Paris. we started in 2013 and today we are walking with a bit more than 200 customers. actually, most of our customers are working abroad. we do work with organizations like Cisco or NASA and also couple of governmental institutions like the French Ministry of Energy. our increased enterprise and link use SDK products as analyst, extract Eden insights from graph data. so what do we mean actually by unlocking the value of growth data, our customer, our data, actually in various formats like excel file, SQL database and so on. the first thing is like the model, his data, as a graph of entities. this is called nodes or relationships, this is called edges. they then move this data inside the graph as events like neo project and neo4j else store and query large works. the liquor software makes this data and the power of neo4j accessible to non-tiknical users without easy-to-use interface. they can monitor, investigate and detect insights hidden inside their graphic. so our customers use linkers to solve various use cases. for example, we have customers working on tribal security. there are several Reuters application- this is something you can see on the screen- and the used linkers to detect and investigate suspicious activity patterns with similar data. actually, we have some of our customers were working on IT operations or enterprise architecture use case where, for example, differ from root cause, another analyzing or data lineage. we also working with governmental institutions who have access to emails, transactions, phone call records and try to detect and investigate criminal networks. this is called intelligence. finally, most of our customers actually are banks, insurance and money transfer companies work on anti money laundering and fraud. they are looking at connections between clients, their transactions, personal data, financial records and etc. the purple for the purpose actually of today's webinar, we'll be focusing on aspected Pacific kind of fraud called ecommerce fraud. I will now let my colleague Elise tok about this in more details. Thank You, Muhsin. so, as mentioned, today's topic is about the graph based approach of e-commerce fraud detection and Before we jump to the demo, we wanted to explore a big this world and the challenges it rise for prevention and detection solutions. so, first of all, when we speak about e-commerce frontier, we refer to any type of fraud scheme that uses Internet related means to present for doing solicitation to prospective victims or to conduct problems for the transactions. and those proteins importunately come in many forms. among the most common schemes we can find friendly fraud where a merchant we seized a charge by because the cattles are denied making the purchase or receiving the order and yet the good where often actually received. there is also promotion abused or fishing, which is a common practike of sending fake official email to steal sensitive information? we can find merchant road when a foster offer goods at very cheap price but never actually ship them. will our accounts take over when a fraudster is using another person account to obtain product and services? we also have affiliate fraud, which refers to any sales activity meant to generate commissions from an affiliate program. or reshipping fraud, where fraudster recruits a person to package and receive merchandise- person previously with stolen credit card. and finally we are identity theft, which is the fraudulent achilles- acquisition of sensitive personal information- and those schemes are not the only ones- doesn't exist. in over the years we've seen the froster keep renewing the tactiks and, along with new types of fraud, we have also witnessed an impressive increase of fraud entities. actually, in 2016, large e-commerce mass merchants lose up to 1.4 percent of their revenues to fraud in average, which actually sounds 4 billion of dollars worldwide. and there are several types factors driving the productivity. the first one is kind of obvious. if the e-commerce world, it has become mainstream and it focused to expand rapidly during the two years notably, we are friends crashing into new international markets. so frightened example: last year ecommerce sales reached one point nine trillion dollar and it continued to expand rapidly. the second factor that identity to fraudster in the past years is the adoption of new tiknologies and practikes such as mobile payments, bitcoins or cloud services. those actually offer new methods for fraudsters to bypass existing security measure and to Colonel covered their tracks. and finally, we've seen fraudster organize themselves in networks and take advantage of it. they use the same resources they buy act data from large bridge Americanizing to groups across the globe. for instance, this March the District of Columbia arrested like nineteen people testing 14 various international online fraud activities that have led to more than 30 million dollar of losses, and one of the indictment was an online vehicle fraud where the partikipants satified advertised car for sale that they did not own. so, faced with this situation, ecommerce merchant apps turn to the fraud prevention solutions and those web probe detection software or sometimes cloud-based services running background processes that can scan transaction and score them based on the possibility of fraud. today, many solutions follow a layered approach that assesses activity at multiple levels and use multiple analytikal approach. most of them rely on the three layer displays here and porn centric layer, navigation Center player and child centered layer. so for the first one part of the software actually scans information that encompasses clients of applications, device art ID or geolocation. the second layer controls include real-time dynamic capture of accounts and customer activity and this is actually used to build a customer profile to determine what is normal and what is not abnormal behavior from dispersed to our customers. and the third layer is often channel centric and monitor entity behavior across one specific channel. but for an efficient fraud detection nationally of Gartner Enterprise, advancing us to add two more layers to the system. so the layer four and five are about connected analysis and they do require a tiknology that does not rely on a relational database model, for instance. the fourth layer is cross channel and is dedicated to the analysis of abnormal behavior across multiple channels. the fifth one is entity link analysis and it's all about analyzing relationships to detect organized crimes and probe rings. so tiknology using only the three first layers can actually say some tiknical limitation. so first, the monitoring and detection system cannot keep packed with new products or new schemes. the structure is simply too rigid to allowe the easy adoption of new rules or of new data and suddenly those system that exists in product and channel.

Troubleshooting Your Ad Results The Customer Experience Facebook Ads Masterclass 2020

hello and welcome to this video in the Facebook Ads masterclass. today we're going to go through and troubleshoot all of your results to see what to fix and improve on your website. I like to visualize this as a funnel. you're gonna hear that word used a lot in a lot of different contexts. for this context I mean the customer sales funnel from recognition to purchase. before we get started, I wanted to ask everyone a favor. I put out one new video every single day on advertising, marketing, drop shipping and Shopify. if these videos help you in any way, please subscribe to the channel and like the video to get these in front of more people. Michaels, to help make the training accessible to anyone who wants to learn, regardless of their current financial situation. I'm a firm believer that education should not cost you a dime and I want to take the information hiding in those high-cost programs and make them available for free. thank you so much for clicking through. I hope you enjoy the video as those thing in the introduction I like to visualize the customer journey as a funnel on the screen. you can see why, at each step of the customer journey, your audience starts to shrink, which is 100 percent normal. at the very top you have your recognition. this is people seeing your ad and deciding whether or not they want to click through and learn more about what you have to offer. under that we have all these who have clicked through to your website or offer. then we have those who said they wanted it and added your product or service to their virtual cart and the next step is the person who started putting in their information in the checkout screen and then, finally, at the other end of your funnel, out comes the customer. this video is going to be breaking down each step of the customer journey in your funnel and explaining how you can improve the flow of your numbers from recognition to purchase. top level recognition at the top of our customer funnel is recognition. this is how many people are seeing your ad. in most cases, the only thing you could look at here is your CPM or cost per mil. this is the cost to you to have your advertisement shown to 1000 people. when you have a CPM that feels except hi. there could be many different factors related to what is driving up the cost. competition is typically the number one driver of cost, but if you're going after a partikularly competitive niche, you can expect to see CPMs, Jeff, as high as 40 bucks, while lower competitions may only cost five to ten dollars. on average, my campaigns for the big 5 countries on facebook- united states, canada, united kingdom, australia and new zealand- see a CPM of about 1750 cents. there is a new factor that might be driving your cost up: your customer feedback score. as you start to sell more products and Facebook pulls your customers, you generate a customer feedback score. if your customer feedback score drops below two, you'd expect to start seeing some warnings and penalties that drive up the cost of CPMs. and then let's tok about your clicks, your cost per click. this is really a subsection of your recognition. if you're experiencing a high cost per click, you have a few things you can look at to see what you can improve. there are three main drivers for a high cost per click. number one is high CPM, number two is poor ad copy and number three is poor targeting. since we just toked about CPM, I won't go into it, but know that if you have a high CPM, you'll likely have a higher than average cost per click. this isn't always the case, but it is likely to happen. poor ad copy is almost always the problem when you have a high cost per click. most people write bad advertising. this isn't an indictment on your ability to write compelling advertising. it's a testament to the modern consumer being able to tune out what used to work so well. writing ad copy is an art that most people just don't care to learn. the more you write and play around, the better you become at writing these ad copies. that's all there is to it. follow the tips that I gave you in a previous video in this series about writing ads, using our four-step template to make the process slightly easier. poor targeting is something that can happen a lot of time. it is just product audience mismatch. you might think that your pair of silver sneakers would be a great fit to target someone who fits in the interest SilverSneakers, but audience is actually completely different than you initially thought. take a moment to analyze your audience and make sure that your product fits the audience well before jumping to the conclusion about your ad copy. traffic clicks. you have plenty of people coming to your website, but it feels like, no matter what you do, they just don't click Add to Cart. this is by far the worst feeling when you are starting off. it could be completely demoralizing to a new store owner, keep your chin up. we've all gone through this at some point. there are a few things that can cause a customer to show enough interest to click through, but not enough interest to actually purchase. they are: number one: misleading advertisements. number two: product value doesn't match the price. number three: poor product descriptions. and number four: poor page designs. now I would like to think, in the perfect world, the first option doesn't exist, and if it does, they know exactly why no one is buying. you can't advertise a chair, and when your customer gets to your landing page they see a waffle iron and when they leave you can't be surprised. make sure what you're advertising and how you're advertising it is representative to the product you're selling. product value doesn't match the price. notike, I didn't say the price is too high here. you can- and in some cases do over sell a product. think about it. we all feel like we have this internal detector that can sense when something just isn't right, when it goes off. we stopped paying attention and just move on. imagine you get this amazing product description that outlines a product and makes it sound like it's worth a thousand dollars, but the sales price is $19.99. our BS detectors are sounding the alarm. they're trying to steal my credit card info and drain my bank account. no way that this deal is real. you're not getting my money. the inverse is also true. you can, and probably do, under sell the product. let's have an honest discussion about this. you aren't going to sell a crappy Bluetooth headset worth seven dollars for $99. you can only polish that turd so much you might be able to squeeze out twenty to thirty dollars, but if all you do is write out headline and features and put some blip points under it, you're getting $10 for that headset. even if you write highly compelling product descriptions, there's only so much that people are willing to pay for an off-brand Bluetooth headset. and let's tok about that for a second. your product descriptions. if they're poor- let's say your grammars poor, your English is poor and you're going after an english-speaking audience, you're going to have issues. so make sure that your product descriptions are as decent as possible. you don't have to have a master's degree in English for this to work. you just need to make sure it flows like you're having a conversation. this isn't an essay that they have to read about the product. it's supposed to take them off the paper as if they're using the product. lastly, poor page designs. if your website is hard to read and there's colors everywhere, its unappealing, and if it's unappealing, people will not buy. make sure your page looks professional, look sharp and it doesn't have a bunch of different colors and things flashing, not a bunch of popups. your customer is there for one thing: to buy that product. they're not there to give you their email address to sign up for a newsletter for a product or a service that they'd never heard about before. keep it simple. page design is one of the biggest flaws in most ecommerce stores, because people just go and do the craziest things, cuz that's what everyone else says is working it d.