SuperSifter Spotlight: TCGplayer and the power of Payment Protection + Account Defense

0:03
Welcome to the first of our new Super Sifter Spotlight series. Today we’re
0:07
talking with TCG Player about the power of payment protection and account
0:10
defense. My name is Eli Mker. I run customer marketing at SIFT. And I just
0:14
want to start with some quick housekeeping items before I introduce
0:17
our host and our super sifter guest. The this webinar is being recorded. It will
0:21
be made available to all registrants and attendees later this week. We know
0:25
you’ll probably have questions for our speakers, so we’ve reserved some time at
0:29
the end for Q&A. And you can submit questions throughout the webinar by
0:32
using the Q&A button. We’ll try to answer as many of questions as we can
0:36
live, but if we don’t get to your question here, we’ll be sure to follow
0:39
up with you directly after the webinar. And with that, I will introduce your
0:44
host, Jesse Danziger from SIFT.
0:48
>> Thanks. Uh, thanks, Eli. Yeah, it’s great to be able to uh be uh on this
0:53
webinar with my good friend Joe. And um just a little bit about me. Um I’m Jesse
0:58
Danziger. I’m a customer success manager at CIFT. I’ve been here for about three
1:02
years and I work alongside fraud leaders like Joe uh to help get SIFT set up um
1:07
or get optimized. Um and uh before I was in this role, I was actually a SIF
1:12
customer myself for three years at a real estate listing platform uh site.
1:17
So, uh, that site actually functioned quite like a marketplace. In fact, uh,
1:21
we had renters and property managers, um, and we were the site that brought
1:24
them together. And so, we face similar challenges with bad actors scamming our
1:28
renters. Um, and so I do have some experience personally, uh, with the
1:31
account defense side of, uh, the marketplace protection environment. But
1:36
Joe actually manages risk at both the account level and the payment
1:39
transaction side as well. Um, and so really excited to be talking with him
1:44
today. I’ve worked with Joe for about about that three years now that I’ve
1:47
been at SIFT and um he’s just a wealth of knowledge and a really fun guy to get
1:51
to uh to talk to and hang out with and so I’m excited to introduce him and get
1:55
into this conversation about um risk risk management in the marketplace
2:00
environment. Um so hi Joe want to give yourself a quick introduction and then
2:05
I’ll we’ll we’ll jump in.
2:07
>> Yeah. Um my name is Joe Gentilly. Um born and raised in Syracuse, New York.
2:12
Uh, and uh, I’ve been fighting the good fight for nearly 12 years now. Um, but
2:21
with CIF since 2018. So, um, I’m it’s it’s just a, you know, just honored to
2:28
be even asked to be to talk and uh, and once you will definitely once you know
2:33
me, I I can’t shut up sometimes. So, um, so I’ll try to keep it brief for
2:39
everyone here. Uh, no, no worries. So, um, so Joe, uh,
2:44
why don’t we start with TCG, uh, TCG player? Um, first of all, what does TCG
2:49
player stand for and what is the brand known for? So, kind of give us the
2:53
background on what you’re known for because really
2:55
>> Definitely. Uh TCG is trading card trading card game player, but we’re an
3:00
online marketplace for collectible gaming uh like Magic the Gathering,
3:05
Pokemon, Yu-Gi-Oh, you know, we service hobby shops and backpackers alike. Um
3:11
anyone who wants to sell, buy, um we’re and even manage the how much their
3:18
collection’s worth. Um we’re that go-to uh site for it. Um, we we pride oursel
3:27
on the market price and and using that as a a great way to, you know,
3:31
accurately, you know, price your in your your cards um that you have on stock. I
3:36
mean, Magic’s been around for we were just saying it about 30 years. Magic’s
3:39
been around.
3:41
>> Um,
3:42
>> it was the hot game when I was in when I was in middle school, I think, is I
3:46
remember when it came out.
3:47
>> Yeah. Uh, we’ve been uh it’s been around the the site itself, the marketplace has
3:51
been around for about 13 years. Prior to that, it was mostly content um articles
3:56
and and such like that, but um we went we went live with the marketplace around
4:01
13 years ago. Um so,
4:04
>> and then of course the the big news is recently um end of 2022 you were
4:09
acquired by eBay. Can you talk to us a little bit about that and what that
4:12
process has been like?
4:13
>> Yeah, we we’ve we were I mean I was hire number 12 with the company. home and I
4:20
was in, you know, customer service, um, helping people, just normal
4:26
servicing issues. Um, and I saw I’ve seen us grow and grow and grow and, you
4:32
know, double our employee size and all that stuff. And we just got to a point
4:36
where, you know, there was kind of two paths in front of us. And and this was
4:41
probably the smartest path was to be acquired by eBay um to get those
4:46
resources and that and that name power behind it to kind of get more clientele
4:52
with with us.
4:55
>> Has that changed your dayto-day at all? I know
4:57
>> it has not changed at all. Um I always like to compare it to the Marvel
5:01
acquisition or Disney acquisition of Marvel. Uh basically they have the
5:05
Disney name but Marvel is their own entity.
5:10
Um yeah, really exciting stuff there. And um you mentioned that you were hired
5:15
number 12. Um and you know, I’ve known you for three years, but I think
5:17
prepping for this uh webinar, I I got to learn a lot more about your background,
5:22
you know, before TCG. I thought that was really fascinating um just what you were
5:26
doing before that, how you got to TCG, and I thought you have a really
5:29
fascinating career path um at TCG getting to trust and safety
5:33
specifically. Um, and I think it I’d love it if you’d be be willing to share
5:37
kind of some of the details around that um, with the group. I thought it was
5:40
really fascinating um, and interesting.
5:42
>> Yeah, sure. Um, my whole life, you know, I’ve worked retail for many years. Um,
5:49
mostly in sales and and and I I was sold shoes for six years. So, I was Al Bundy
5:55
for those who know. Um, then I worked at a bank. I worked in I was customer
6:00
service there. I worked in loan servicing. I worked closely with the
6:03
collections department. Um, and then I, you know, I I got a job
6:07
at TC Player in CX. And what’s great about a small company is you have to
6:13
kind of find your niche or you’ll kind of get left behind. And I’m all one of
6:18
those people who like to kind of like grow as a person obviously and and and
6:22
knowledge base. And I I I have knowledge what what helped me was my background
6:26
knowledge with banking. um the flow of like of funds through a bank statement
6:32
or a debit card, credit card, how does it appear, how does it work? Um opening
6:36
up chargebacks for users who come in, seeing it from that side. Um so I had to
6:41
find a niche. I had to find I had all my buddies who were working at Tsquare,
6:45
they were finding their marketing, you know, they’re really great with words
6:49
and and all that fun stuff. So I was like, I have to find mine, you know, to
6:52
kind of get my place in the business. So what happened was I started just doing
6:58
little things and just looking at loss you know where where we have loss you
7:02
know you know missing packages is a huge thing for us you know we always have pro
7:08
people claiming that they haven’t got especially with tracked packages so what
7:11
I used to do is I used to get on a walking treadmill and call the USPS
7:15
national number and finally get to talk to a person say hey I’m looking for this
7:19
package and they would send me to the right post office and I would get there
7:22
then I started working with chargebacks and people with
7:25
users writing in with unauthorized transactions, a lot of lot of friendly
7:29
fraud, too. You know, a child in a house using their mom’s card, stuff like that.
7:34
And I kind of grew and it kept going and I I started researching more and
7:37
researching on like international, you know, credit cards and just general, you
7:42
know, chargeback processes and and understanding, you know, the flow of how
7:47
things kind of work at our our our on our landscape of TCG Player as a
7:53
marketplace. And then I kind of reached the cap of what I was doing in in
7:59
customer service. I I was I was definitely a fraud
8:03
uh specialist there and I was helping a lot of people, a lot of the team members
8:07
as we grew, you know, we kept doubling in employees every year. Um and as we
8:12
grew and and and I kind of reached a point where I was like, I can’t grow
8:16
anymore. So what I did was I I reached out to our finance department which is
8:21
the CFO and the and the uh VP of finance to uh see if they needed help plus you
8:28
know follow the money you know basically is the the the name of the game here and
8:34
I essentially created a position and then applied for it and and to my
8:39
surprise I got it. So, um, what was great about that at the time was I was
8:44
really focused on fraud and chargebacks. And I would bring that up to the finance
8:48
department, which had no clue. They all just saw a number. They had no clue what
8:52
was behind it, but our chargebacks was really good. You know, they were really,
8:56
really good. We’re, you know, the two grand a month type of situation. Good.
9:01
So, they didn’t say it as a priority. So, what happened was is I became an
9:05
accountant. I was closing books. I was paying bills. I was doing all that stuff
9:11
and then fraud happened and fraud happened in a big way and uh they asking
9:18
me why and I was like well I was focusing on being an accountant.
9:22
So we um we needed to find a solution. Our our
9:27
internal solution was not great. It was a cumulative score based on rules that
9:32
we designed that never moved or adjusted with the with the with the user, the end
9:38
user. So, it just kept growing and growing and we would just be looking at
9:41
the one person, not any connection points. So, we needed to find an answer.
9:46
So, we we actually went through a a great service called Paladin. Um they
9:52
are they’re basically a consulting group that helps companies find different
9:58
solutions for their issues. And we actually talked to a bunch of different
10:02
um vendors uh email edge riskified for all those people. Sift and we found that
10:08
sift was the best answer for us. Um it worked well with our our current systems
10:14
in place and it did what we needed it to do um with the AI help. you know, and
10:19
this is before AI was really in the news. It was machine learning was just
10:24
crazy for everybody, but it’s still like it makes sense. Let’s look at a bunch of
10:28
parameters and then and then put it and spit out an answer.
10:33
So, we’ve been with them since 2018. And my my fun fact with 20 with with getting
10:38
with Sift is um during sandbox mode when we were testing the system, I was able
10:44
to identify multiple rings of fraud that were already on our platform through the
10:50
association and through, you know, all the in the integration that we put into
10:55
place. We were just in sandbox mode. I’m like, well, there’s some fraud. I stop
10:57
that. Stop that. Stop that. I’m like I have to everything had to be done
11:00
manually. There was no like web hooks or anything connected. Um, but it was it
11:05
was pretty interesting to see coming from what we had to what we were going
11:09
to be.
11:10
>> Yeah. Yeah. That that’s a great anecdote and it reminds you know you went through
11:14
Paladin um to kind of find to evaluate different platforms and it it reminds me
11:19
of when I was um uh when when I was a customer of Sift and how we got to Sift
11:24
very different path but uh like you said it was not a problem. It was a customer
11:29
service issue when we had scams and and and abuse on the platform. It was a
11:33
customer service issue. It was like help help the uh the customer just like not
11:36
be too upset about it and move on. And like you said it really well like like
11:40
you said until fraud happens. And it’s like suddenly the dam breaks and you
11:43
have this huge existential crisis that’s affecting the business. And the way we
11:47
found Sift was we actually hired someone for that customer service team who was
11:51
formerly a fraud analyst at a big SIFT customer at the time. Um, and so he he
11:57
was like, “Hey, you you should try SIFT.” And I think that we we thought
12:00
about building out rules internally. And we were all just like, “No, no, no. We
12:03
need something dynamic like you said.” And um, to your point, you know, right
12:06
when we turned it on, like even uh, just turning on initially without even
12:10
training the model yet, just the global model kind of in the background with
12:13
background signals, we were starting to catch like all the the really lowhanging
12:16
fruit that um, is really kind of obvious in a model, but not so obvious in rules.
12:20
And so, yeah, really love that story and thanks for sharing that. Um, and I think
12:25
that’s a good segue into um, you know, you talked about, you know, applying
12:30
SIFT into you have this more like dynamic marketplace now um, with that’s
12:35
a little bit more complex than, you know, a few just sort of yes, no rules.
12:39
Um, so can you walk us through what’s going on there uh, with both on the
12:44
buyer side and the seller side? So, um, what are the points of risk on the buyer
12:47
side and the seller side? How are they different? Where where’s their overlap?
12:50
and then let’s kind of like like unpack that and go from there.
12:53
>> Yeah. Yeah, no problem. Well, with with the buyer side, um that’s been running
12:57
smoothly since 2018 since we kind of launched it and we integrated all that
13:01
stuff. What I love about um SIFT mostly is the fact that it learns based on your
13:07
decisions. So since 2018 and I always like to call it’s a welloiled machine.
13:12
So buyers the buyer side I feel like it’s very good lockdown. We have very
13:15
low chargeback rates. we have very low fall uh false false positive uh rates um
13:21
from agents and if we do we just kind of pivot and or change the workflow a
13:25
little bit all that stuff. So when it comes to buyer fraud it’s it’s standard
13:30
across anything. I always like to say you know it doesn’t matter what you’re
13:32
selling the fraud is going to be a fraud across the the board when it comes to a
13:38
marketplace. You got stolen credit cards, creating m creating multiple
13:42
accounts, account takeover, um friendly fraud, which is very hard to
13:47
kind of identify, refund abuse, policy abuse. Um
13:53
so, so for buyer side, it a lot of times we used to the big thing for us was like
13:57
triangle scams. Um, triangle scams is is some a user a fraudster using using
14:04
utilizing another marketplace to list and sell items that they then use stolen
14:08
credit cards to purchase on our platform to sell to that particular person uh to
14:13
send to. So the the the transaction looks whole except we’re receiving
14:18
chargebacks on these. Um also you have inadvertent because of the the
14:22
connection points. Um you’ll get people who get blocked on our platform due to
14:28
that chargeback because of those associations and they’re like why is
14:31
this? And so we’d have to ask them like well did you purchase something at
14:34
another marketplace and like yeah I bought this and well they are not legit.
14:38
Um so you’d have a lot of those kind of those type of things for seller fraud.
14:44
Um it’s we relatively started looking at them 2020. Seller fraud’s always been
14:51
there. The biggest thing with seller fraud is uh validation at registration
14:58
um as well as manipulating market price. Uh basically selling
15:05
product for less than what they’re actually valued. And since our market
15:09
price is kind of the the big, you know, thing that we kind of pride oursel on
15:15
and trust have people trust in it, it’s something that we really focus on a lot
15:19
is try to catch them before they start listing and selling completing uh sales.
15:25
>> Yeah. And I think that’s actually um a really interesting point and to kind of
15:29
double down on that just uh the market price uh you know to unpack that a
15:33
little bit for you guys is you’re you’re basically like the New York Stock
15:37
Exchange for for trading cards, right? Like you kind of set the market price um
15:41
and it’s based on what’s what things are being sold on your site. Um, I think I
15:46
mentioned like a year or so back a cousin of mine was getting married and
15:49
um they were visiting uh LA where I live and a cousin and her fiance um where
15:54
they said, “Oh, we’re going to a friend’s house. They have this big kind
15:57
of like ‘9s CD folder looking thing, right? It is full of Magic cards. We’re
16:01
going to go play a big Magic game.” Like, “Oh, you’re into Magic.” I was
16:03
like, “Do you know TCG player?” Just like off chance. I thought it’d be cool.
16:05
Like I work with them. They’re like, “Do we know them?” Yeah. We use their we
16:09
literally like use a third party site called like TCG tracker to like track
16:12
how much money like our cars are worth. And so that’s like a big a big piece
16:17
there, right? Is um keeping um integrity on on uh uh the price of the price of
16:24
cards, right?
16:25
>> Yeah, it definitely is. And it it looks at all the transaction. It’s it’s based
16:28
on, you know, you know, it’s based on a bunch of different stuff. And we it’s
16:32
really just what we kind of hang our hat on. you know, and
16:38
it fluctuates almost it’s quicker because, you know, our marketplace is
16:42
open 24/7 and people are able to change their prices uh at at any time based on,
16:48
you know, what is something worth, you know, and that’s something’s worth is
16:52
what people will pay for. So, that’s kind of it helps us also identify a lot
16:57
of fraud, too. So,
16:58
>> yeah. Yeah.
16:59
>> It’s one of those it’s one of those red flags that pop up that really does help
17:03
a lot.
17:04
>> Um, okay, cool. One more sort of uh uh drill down question on this is um
17:11
between buyers and sellers, what is like the different risk that you face if an
17:16
account gets taken over? If a buyer account gets taken over versus seller
17:19
account gets taken over, what are the risks there for for you guys?
17:22
>> It’s actually it’s it’s very very huge risk. It’s a it’s um for buyer side,
17:27
it’s obviously you got your PII, you got your your personal information that’s
17:31
out there. You got addresses, you got obviously um credit card information if
17:37
it is stored. Um it is, you know, it’s tokenized, but so they can’t really get
17:42
the whole number, but they can definitely utilize it on our platform to
17:46
if if they are inclined to use that. Um they can do kind of a little bit of a
17:56
it’s kind of they could basically be negative towards our sellers. They can
18:00
make they can make the buyer side look worse by feedback or messaging all that
18:07
stuff. Um the seller side I think is more scare is scary to me. Um the seller
18:13
side they get um again everything’s kind of obviously tokenized because it’s more
18:19
security but they can essentially
18:23
manipulate market price. They can sell fake items to themselves.
18:28
Um, so that you know, so you’re using stolen credit cards to purchase from
18:32
yourself. But they can also kind of hide in the weeds a little bit, be good until
18:38
not, and basically change the payment information of how these sellers get
18:43
paid and and technically kind of hide and get that money that way. And now we
18:50
have alerts and all kinds of bells and whistles that that jump up when that
18:54
occurs. But, you know, one of the newer things that that SIFT is offering is
18:59
something I’m really excited about is the atto beyond login
19:03
um because that will help protect any kind of change in someone’s
19:10
information. So, if a if a shipping address gets changed allow if you want
19:15
to change your payment information to a FA. You know, obviously the dream is to
19:19
and it’s going to take a lot more work is to centralize all that PII so you
19:22
don’t have to do this in all different kinds of update events, but to just kind
19:26
of put everything behind a wall, but that’s that’s the newer thing that we’re
19:30
going to start incorporating into our into our uh workflow basically.
19:36
>> I I was going to ask you toward the end about your like sort of dream dream
19:40
setup and um like what the challenges are internally to get there and
19:43
externally with partners, but um you did mention that. So, um, maybe we can stick
19:48
on that for a sec. Um, so I know you talked about like, yeah, having having
19:52
all the sensitive information behind behind a 2FA wall. Um, but then also,
19:58
um, I know you have a pretty layered strategy for just to talk a little bit
20:01
more about the details of where things are at now for you. You have a pretty
20:05
like layered approach to atto and, um, you sift, you have some other uh, some
20:10
other things in play there. Can you talk to us about just some of the tactics
20:14
that you’ve used? um uh recently and and how that’s going and and actually can
20:18
you back up before that um getting into like the tactics. Can you just tell us
20:22
briefly like when did you start seeing atto pick up and and come to a point
20:26
where kind of like you said initially about the payment side like fraud
20:29
happens and then now you have this big use case where where did you see atto
20:32
start popping up? How did you identify it? And then what were some of like the
20:35
initial things you you did to to address that? And then let’s maybe like can we
20:39
unpack like what your setup looks like now?
20:41
>> Yeah. Yeah, no problem. Yeah, the the obviously for just like everybody else,
20:45
ATO started when COVID started. Um, nobody left their house, so all they did
20:49
was like walk around and steal people’s identity. Um, so that was that’s the big
20:54
one. That’s that’s the huge kind of like when it did uptick that I feel like it’s
20:58
always been there, but not in the quantity that we saw with with the
21:03
during CO. I mean, that and fake seller accounts, those were the two biggest
21:07
things during CO. Um that was the biggest jumping point for us as is as is
21:12
as for for for for seller and ATL was those two was COVID. Um the bigger kind
21:19
of things to kind of like identify um was the the way that we identified it
21:24
prior to you know atto service that we have in place with civ um was sheer was
21:31
CX you know CX would receive multitude of tickets saying hey I you know someone
21:38
got into my account or hey I didn’t authorize this purchase you know all
21:42
this stuff and and and we had to work with IT uh IT security.
21:48
Um they they they’re really great at looking at the the trends of and and
21:56
getting a better idea of login uh via like cookies and IP, you know,
22:02
just in a in a wider shot instead of an individual level um to say, hey, we have
22:08
an uptick of login or uptick of, you know, at the from this particular IP,
22:13
you know. So it’s it’s one of those things that helps us kind of you know
22:17
identify a lot of that stuff. So you know we we saw an up you know a huge
22:22
uptick and it’s kind of calm down a lot. We did in incorporate 2FA um at the
22:27
login based on certain rules. Uh another thing that we’ve
22:31
done uh which is very helpful and I recommend this to everybody is you
22:35
really got to focus on your dormant accounts. Uh people who create accounts
22:39
and just forget about it or leave or change you know just never come back. um
22:43
you really need to kind of what we do is we kind of systematically reset
22:48
passwords for older uh login uh dormant accounts um which has definitely helped
22:54
decrease our atto um because a lot of those things they
22:59
just forget um that they even have an account with
23:03
us um and then they that that account is ended up on a list and they just try and
23:09
all these different places and and then hopefully They, you know,
23:14
they’re just trying to get information or or try to,
23:18
you know, defraud somebody.
23:20
>> Yeah. Yeah. And and I think you’re also using capture as well.
23:24
>> Yeah, we we that’s another thing that IT security helped with. They do capture
23:28
which we actually integrated with SIFT as well. Um, we we in incorporated their
23:34
actual H capture score into our CIF workflows, which is kind of nice to tr
23:39
help trigger 2FAS on high-risk login. Um so and and I think this is a good uh
23:47
dovetail into another topic which is about sharing your positive results
23:52
internally and that’s something you’ve told me about um and you talked about at
23:56
the top actually about with um the the finance team on on the payment side um
24:01
but yeah you have a you have a really interesting approach I think and because
24:05
you came from the finance side in that world um and you kind of know what what
24:08
folks there are looking at um how do you go back to finance ance or or leadership
24:14
um operations teams and share the success you’re having with things like
24:20
ATO like um you know uh fake fake um fake merchandise that impact your your
24:26
your um market price authenticity or integrity. Those are the kind of things
24:31
that are not quite as quantitative as like here’s how many chargebacks we have
24:34
and here’s the dollar value. So how do you summarize that or how do you get the
24:38
attention of leadership that and can you talk us kind of through that? I think
24:41
you have kind
24:42
>> the yeah the the the key thing is to identify metrics. Um I I am really in
24:50
into the weeds with data. Um I learned that when I was in finance. Um you know
24:56
I I learned how the sausage is made basically you know from from start to
25:01
finish from you know from someone checking out to a seller sending it out
25:07
and getting paid. you know I know the whole step and every every stage of it
25:11
going through. Um so what you do is you create metrics on all I I I I gather on
25:19
all transactions, all data, all chargebacks and you find those key
25:24
components. Um a lot of what I was kind of taught a long time ago is to kind of
25:28
work with the the finance team and to understand the uh the line item on the
25:32
the P&L that says where are we losing the most? And from there, you can kind
25:37
of back into what you’re what you need to focus on to help result to fix that
25:42
leak. Basically, is it is it are you having too high of chargebacks? Are you
25:45
having too high of refunds? You given too much credits? You given too too much
25:49
adjustments? Um, all that is in the you know, all that is in the dollars. But,
25:55
so what I do is I gather all these metrics
25:58
and I kind of I I monitor them. I It helps me with trends. It helps me with
26:05
with making changes and sift all that stuff. I do it on a daily basis. Um, and
26:10
then I do it on a monthly cadence and then I do it on a week, you know, and
26:13
then quarterly. So, what I did what I do is I kind of gather all that
26:16
information, wins and losses. Doesn’t matter. You you can’t be afraid to tell
26:20
the losses because I was and then I realized, well, it’s not my fault. You
26:25
know, I didn’t make I didn’t do the chargebacks, but it is a hole that I
26:29
identified. it is something that I you know my my job is to make sure that it
26:33
gets decreased you know or increased security or whatever. Um, so what I do
26:38
is I set up cadences, um, meetings with leadership,
26:45
mostly the CFO and the VP of customer experience. Um, and I kind of walk them
26:51
through it, you know, not just to say, “Hey, our number is this.
26:56
Our I go deeper. I go that step further because they already have the number
27:01
because they do the the accounting. I go, it’s this.” because of that we had
27:06
an uptick in this um we you know here’s what I’m doing to help you know calm
27:13
that tide or it’s been decreasing since the end of quarter you know since I
27:17
identified it you know for example we had a an issue this quarter which was
27:22
billing and shipping mismatch so newly created accounts you could tell they’re
27:25
stolen credit cards because they’re using billing and shipping mismatch
27:28
that’s egregious basically you know you got New York and California you you
27:34
know, or you could see that they’re buying similar product line and type of
27:39
stuff. So, I was able to say, hey, our chargebacks were up because of this, but
27:44
it was identified early enough that here’s how I could show you that it’s
27:48
been decreasing since then. Um, or I could say, you know what, this was an
27:52
issue. We may need to get development work in place or something that helps
27:57
with that, you know, more better validation. I always a lot of my stuff
28:01
is I just I look at three buckets every day. Every day of my life I don’t want
28:06
to say I look at three buckets. Validation, prevention, and detection.
28:09
Um and I kind of like that’s how I categorize how I want to like prioritize
28:15
things. I I know I’m rhyming here and I don’t mean to.
28:17
>> Oh no. Well, actually that’s what I was kind of driving. I know you have that
28:20
that uh process you have of validation, detection, prevention and um I thought
28:26
why don’t we can can we can we go into those three items quickly and like what
28:29
do those mean to you and and how do you use those to kind of surface these
28:32
metrics and and
28:35
>> yeah detection is the easiest one it’s sift you know it’s utilizing sift and
28:39
the data to understand what is the issue you know and I kind of put what I want
28:44
to work on my goals for the year and how I want to work on that Um the the
28:51
prevention part is a bigger kind of issue. It’s not a lot of times it’s not
28:55
even technical. It’s it’s reviewing your policies that you have in place.
29:00
You know, getting making sure that they’re that they’re not vague, too
29:05
vague that there’s wiggle room. Um, the one of the issues that I’ve that we have
29:10
at TCG Player is our users are gamers and they will find a way to game the
29:18
system. They want to win. So they know how if they see weakness in a in a
29:23
policy, they will they they won’t do anything wrong, they will follow the
29:28
policy as it is written. So, you need to kind of what I’ve been working on is to
29:34
better verbalize the policy where there’s less
29:38
wiggle room and I’m able to, you know, I like to say tap the sign like, “All
29:43
right, this is what we this is how it is.”
29:46
>> We’re talking about your best customers here.
29:48
>> Yeah. Yeah.
29:49
>> Because they they’ll, you know, a lot of times, especially with like refund
29:53
policy is a big one. You know, refund abuse is a huge, you know, get right
29:58
now. It’s a really that’s kind of the hot topic right now. If you go to any
30:01
any fraud seminars, they’re going to talk about ATO and they’re going to talk
30:04
about refund about abuse. Um,
30:08
but it’s something you need to look at. You need to review it with all parties
30:12
involved. You just can’t make up the rules yourself because as fraud, you
30:15
know, protectors, we will be as stringent and nobody can do anything
30:20
ever. So you need to work with marketing and you need to work with CX, you need
30:24
to work with all the departments to understand what is the best route that
30:28
will that will benefit everybody involved.
30:32
Um and when it comes to the validation um as online marketplace um is my
30:38
biggest pet peeve of everything uh is that everybody’s faceless. All you see
30:44
is card not present. All you see is sellers just creating account with an
30:50
email. Um, we do get we do collect a lot of the stuff like W9’s and stuff like
30:55
that, but it’s that validation step that that definitely is in my biggest uh road
31:02
map because I want to give a face to the
31:05
faceless. You know, if you’re going to do wrong on our site, I want to be able
31:08
to look you square in the eyes and be able to catch you if need be
31:13
>> without jumping through too many hoops to get set up and start selling. Right.
31:17
>> Right.
31:18
>> Right. That’s where we work with all the other teams.
31:20
>> Yeah. Yeah. Uh that’s great. Um well, I know I want to make sure we save a
31:25
little bit of time towards the end for for some Q&A. We have uh 12 minutes
31:30
total left here, but maybe I can wrap up by just asking you, you know, we talked
31:34
about your dream setup. So, I know um you’ve really talked about like the last
31:40
year or so about um seller identity. So, like knowing more about the sellers,
31:45
you’ve talked about the atto uh sorry, you’ve talked about putting two two uh
31:50
2FA wall uh in front of like sensitive information. So, when account if
31:54
accounts are compromised, you have that that layer of protection. Um is there
31:59
anything else? Um you know I I know you’ve really been focused on the seller
32:02
side. If you had like one thing sift sift feature use case or otherwise or
32:08
other you know um internal etc. What would be something that you really are
32:12
looking forward to like putting in place uh in the next year on the seller side?
32:18
>> On the seller side um it’s going to be controls. Um, again, you know, what I
32:25
love about, you know, how we’ve progressed and how
32:29
we’re moving forward is is the seller experience. We focus on buyer, like I
32:34
said, 2018 buyer, boom, great. you know, sellers, not necessarily just fraudulent
32:40
sellers, but just giving the sellers what they need
32:44
that will benefit their growth as as as because them being successful means that
32:49
we’ll be successful, but subtrause giving them what they want by giving us
32:56
what we need. So, we give them better reporting, better, you know, guidance,
33:01
better stuff. Not only do they grow, but we can kind of build into those those
33:06
reportings, those those actions, experiences, ways to identify people who
33:12
may be having trouble, you know, because there’s I always like to say to when I
33:16
train people, there’s difference between fraud and bad. You know, if you’re a
33:19
fraud seller, you’re you’re very blatant about what you’re doing. But if you’re a
33:22
bad seller, you’re also hurting our name. You’re also hurting that trust.
33:26
you know, having a lot of negative feedback, having all that stuff them
33:30
because them having a bad having the buyer having a bad experience from a bad
33:34
seller makes them not want to come back to us. So, giving them those stuff will
33:38
help us be able to interject a little bit easier or or maybe stop them before
33:44
it gets out of hand. Mhm.
33:46
>> So that’s like the biggest kind of dream is having better controls and better um
33:51
experience for our sellers which will give better experience for our buyers.
33:55
>> Awesome. Um cool. Well, Joe, thanks so much for for walking through all that
34:01
with us. I I know we have a few minutes and we we have a couple of questions. Um
34:05
so maybe we can jump over there. Before we do that, real quick, um Joe has been
34:10
a super sifter and he’s been a um sift user for a long time and so he just got
34:15
uh in the mail from from Sift a very cool thing. That’s a super Joe the
34:20
bobblehead. So uh Joe is a collector and my favorite quote is Joe likes to say
34:25
he’s always shaking the bobblehead always shaking his head no. So
34:29
>> helps me write my emails a little bit better.
34:31
>> Yeah. Um
34:33
>> make those decisions. So, uh, good pitch for for the sifters community, uh, which
34:38
Joe is is involved on, and then, uh, for the Super Zifter program. And then, uh,
34:43
Eli, do we want to hand it back to you and and see if there’s any questions
34:45
from the audience?
34:46
>> Sounds great. Yeah, we’ve got a couple that, uh, came in while while we’re
34:49
going through this. Uh, so I’ll just kind of go in in order here. Um, first
34:54
one is uh regarding how the flow of revenue is affected by fraud on the
34:59
platform. Could you share any experiences you’ve had in communicating
35:02
that impact to your leaders or decision makers and how are you convincing them
35:07
to address it?
35:09
>> Yeah. Um, yeah. What what what really helped I mean again getting that
35:14
experience from the that side that finance side and being in those meetings
35:18
and all that it helps me understand what you know what they’re looking for and
35:23
what they’re doing so to help with you know the biggest like for example uh
35:28
refunds refunds are a big issue for us for revenue you know because it it
35:32
counts towards you know P&L and and and counts towards how we do in GMV
35:38
um so to understand that is I help them understand the breakout
35:45
of and I and again they know that number all they see is one number they see X
35:53
you know X refunds what I go into and is I will let them know who’s initiating
35:59
those refunds I’m just not afraid to explain the bad you know they know it’s
36:04
bad it just helps them understand why it’s bad so to kind of talk to and it
36:11
doesn’t have to be the CFO directly, but it can also be anyone who’s in that
36:17
who’s closing because then they will be able to help translate it to the CFO if
36:21
need be. So, it’s it’s all about understanding the why we have refunds,
36:27
who’s processing them. Um, you know, obviously one of my perfect world is to
36:32
have better reason codes and better controls over those refunds. So to see
36:36
to to repeatedly say what it is is, you know, the same as kind of
36:43
banging your head with a hammer. But um so what I try to do is I try to make it
36:49
better by helping them understand it by getting better reason
36:55
codes. I’m sorry that was a little janky, but it’s what it is. Um because
37:01
the unfortunate part on our side is that our marketplace is open 247
37:07
and TC player you know as agents are not. So sellers can start doing refunds
37:13
at all times. So we have to explain it in the back end of why are they doing it
37:18
and can we catch them you know oh this this we had a increase in missing
37:22
packages or we had an increase in fraud you know because fraud is a big one. Um,
37:28
but not all refunds are bad refunds. Now, repetitive refunds, that’s bad. And
37:34
that’s where the policy abuse comes in play. And it’s one of the things that,
37:37
you know, I’m working with CIFF to help better understand. And it does help
37:40
their score a little bit when they get refunds, but it’d be no one’s no one’s
37:44
fraudulent on the first refund. Um, Eli, I saw uh there’s a really
37:50
interesting question. I was hoping I could jump into
37:54
>> um and just quick time. We do have six minutes, but Joe, this is I thought this
37:58
was a fascinating question. Um, Robert asks, uh, how do you combat marketplace
38:02
manipulation? Big topic for you. Um, without
38:05
intervening with the quote freedom of the market,
38:10
you know what’s great about that is our users.
38:14
Our users are the first ones to let us know. Um, Reddit posts, all these
38:19
different people, they will they are policing it better than sometimes we
38:24
are. We get it. It’s you can list whatever you want for whatever you get.
38:27
You know, that’s great. That’s if you want to list a single card, you know, if
38:31
you want to list, you know, this 25cent card for a dollar, more than welcome to
38:38
do that. Listing is not the market price.
38:42
You know, list market price is based on if this sells for a dollar. So, if this
38:48
sells for a dollar and then you start seeing another one sell for a dollar and
38:52
then you see another one sell for a dollar. It’s only a quarter card. Um,
38:56
then that’s where we kind of step in and saying, “Hey, are those legitimate
39:00
orders?” And that’s where like Sift comes in play and that’s where our
39:03
chargebacks come in play. That’s where all that stuff. If it’s legitimately
39:07
selling for a dollar, that’s what it goes for. But when you
39:12
get cards like the Black Lotus, the Black Lotus is one of the most expensive
39:17
Magic cards. It sells for about 10 grand,
39:20
you know, heavily played. But if someone’s selling that for $500,
39:25
that right there is more of that’s a the giant red flag for all of us. You know,
39:32
the flow of marketplace for price for standard cards and and normal cards, you
39:37
know, cuz that’s eb and flow there. They can go up. I you know I I once bought a
39:40
car for 50 cents and then I sold it for 20 you know that just the just because
39:44
of the value of that card exploded like that.
39:48
But when it comes to you know it’s more about seeing what
39:54
either one seller is doing or all sellers. That’s how you can kind of
39:57
identify that a little bit better. And it’s also based on how this the card
40:03
sold or the product sold. listing again if you want to list a card for what you
40:08
can list it for whatever you want does not mean it’s going to sell
40:13
you know being closer to you know now if you sell it really low it will generate
40:20
flags and then sometimes you’ll see that is an error they listed a a $100 card
40:26
for $10 because they misentered the the decimal point um that’s one of those
40:31
controls that I was talking about you know I want if someone lists that before
40:36
it goes live into the world. I want a red flag that pops up that says, “Did
40:40
you mean to list this card?” You know, stuff like that is where you kind of
40:44
like help with the experience and also be able to identify fraudulent activity.
40:51
Um um we will never stop anyone. We will nudge if something looks really weird,
40:58
but we’ll we can’t we it’s it’s live as soon as I hit that button. So, and if
41:03
they don’t sell it for the underpric or it’s not legit, we will shut them down
41:08
and that and that will be stricken from the market price.
41:14
>> Um, Joe, looks like we have uh thanks for thanks for that answer. That was
41:17
really that was really fascinating. It looks like we only have two minutes. So,
41:21
um running short on time for uh the questions. Um but I think Eli can uh
41:27
send the questions out to you and we can we can follow up with them offline. Um
41:31
there are a few other ones, but um the last last one it just came in. Just
41:34
maybe a quick one. You mentioned triangle schemes. What do you do for
41:38
that? What do you look for? And
41:41
>> yeah. Yeah, triangles. Triangle schemes. They were fun. They were actually a lot
41:46
of it was really fun and and and it’s just a weird especially since we
41:51
got bought by eBay and most of these sellers are from eBay and they purchase
41:55
from us. What you see from that is you will see completely different billing
42:01
and shipping, newly created accounts, guest account, guest purchases. Um,
42:07
newly created um mostly it’s sealed product. It’s very very niche. Um, and
42:15
what you’ll have there is billing, shipping different, it’s going to a
42:20
specific spot. They are buying it from a legitimate person. These are all
42:24
legitimate transactions. But what will happen is um when you’re
42:29
doing the review or if it does pop into a review because of that billing
42:33
shipping and based on our trends um you’ll see connections and sometimes
42:38
you’ll see them connected to a legitimate buyer who’s been with us with
42:42
us for 10 12 years um you know from the beginning. So they
42:48
they like oh that’s okay that’s weird that they would get that bought like
42:51
that. So what will happen is because of you we’ll
42:58
obviously we’ll identify it quicker with a chargeback when the chargeback comes
43:02
in it’ll block them and any associated accounts and then when this this regular
43:07
buyer will come on our platform he would say why can’t I purchase otherwise it
43:10
keep getting declined we’ll say oh did you purchase this recently so the
43:15
triangle scams they are fun that we haven’t seen them in a while since
43:19
especially since we got purchased by eBay we have not seen them in a M uh but
43:23
they were kind of fun because you were able to get learn more about the the
43:27
activity of the buyer, the legitimate buyer because you would you would write
43:31
to them like, “Hey, did you purchase this from another market?” “Oh yeah, I
43:34
bought from eBay last week. I got it. Everything was seem fine.” Well, you may
43:37
want to let that seller know because we have no control over that seller at all.
43:41
Even with this acquisition, we have no control over that. So, we just let them
43:45
know like, “Hey, you may want to report that seller because they are not legit.
43:51
But they were fun. They They were a lot of fun. Not Not as fun as call having to
43:55
personally call a angry mother whose son stole her credit card and we have to
44:01
work with them to have them return all the cards.
44:05
>> You and the Apple uh uh game store.
44:08
>> Yeah.
44:10
>> All right. Well, thanks. Appreciate your time.
44:13
>> Yeah. Joe,
44:14
>> I I could talk for hours on fraud, honestly.
44:17
>> Oh, that’s I really appreciate it. and Jesse, thank you so much for hosting.
44:21
Thank you all so much for attending. I know there are several questions we
44:24
didn’t get to. We’ll get those questions over to Joe. Uh get you answers. Keep a
44:28
keep an eye out in your inbox for a message from uh the SIFT team. Uh and
44:32
we’ll also include everyone who attended and everyone who registered uh we’ll
44:35
send out a copy of the recording of this uh probably tomorrow, but uh for sure by
44:41
end of day Friday. Um thank you again to Joe, thank you to Jesse. And thank you
44:46
all so much. Have a great rest of your week. Cheers.
44:50
Yeah, thank you.