WEBVTT
1
00:00:09.560 --> 00:00:11.160
Hello everyone and welcome.
2
00:00:11.700 --> 00:00:14.240
I'm Raymond Lee and I lead our key accounts here at Stripe.
3
00:00:14.680 --> 00:00:17.980
And today I'm joined by Corinne Holland, who leads our network cost strategy.
4
00:00:18.900 --> 00:00:21.940
Today we're going to be talking about the payments balancing act,
5
00:00:22.500 --> 00:00:27.200
focusing on three critical pillars of payments, authorization rates, fraud,
6
00:00:27.720 --> 00:00:30.380
and network costs. Now, before we begin,
7
00:00:30.700 --> 00:00:32.160
I want to get to know you a little bit better.
8
00:00:32.860 --> 00:00:35.100
Please raise your hand just a quick poll, super easy.
9
00:00:35.520 --> 00:00:38.860
Raise your hand if the primary metric that you're optimizing for is
10
00:00:39.000 --> 00:00:42.500
authorization rates. Anybody? Small handful of you,
11
00:00:42.560 --> 00:00:47.120
a couple people in the front. Anybody on fraud? Yeah,
12
00:00:47.180 --> 00:00:49.760
a couple more of you on fraud. How about cost and margin?
13
00:00:49.880 --> 00:00:52.660
Something we're all talking about. Wow.
14
00:00:53.280 --> 00:00:56.340
And is anybody here stuck trying to balance out all three of these?
15
00:00:57.540 --> 00:00:58.960
A bunch of you. That's awesome.
16
00:00:59.020 --> 00:01:01.920
And that's exactly what we're going to talk about today. Because on paper,
17
00:01:02.240 --> 00:01:06.300
these metrics seem really separate, but in reality, they're tightly connected.
18
00:01:07.760 --> 00:01:12.160
Picture this. It's Q4 and you go all in on auth rates,
19
00:01:12.480 --> 00:01:15.840
doing everything you can to improve them. And it's successful.
20
00:01:16.300 --> 00:01:20.620
Going from 87 to 91% in just one quarter.
21
00:01:21.400 --> 00:01:25.660
But when finance closes the books, the business had actually lost money.
22
00:01:26.920 --> 00:01:31.000
Retrying more declines caused fees to increase significantly.
23
00:01:32.140 --> 00:01:37.080
And approving suspicious transactions that used to be blocked caused disputes to
24
00:01:37.200 --> 00:01:39.660
spike and issuers took notice.
25
00:01:40.800 --> 00:01:42.900
So they started approving fewer transactions,
26
00:01:43.000 --> 00:01:47.180
which caused the exact opposite outcome that was being optimized for.
27
00:01:47.920 --> 00:01:48.753
Why?
28
00:01:48.940 --> 00:01:53.280
The business focused too much on one high level metric in
29
00:01:53.440 --> 00:01:54.273
isolation.
30
00:01:55.420 --> 00:01:57.800
And I hear versions of this story all the time.
31
00:01:58.180 --> 00:02:02.240
The payments leaders that I talk to almost always start with the same question.
32
00:02:03.200 --> 00:02:07.840
How do I know if my payments are optimized? They always want a number,
33
00:02:08.240 --> 00:02:11.400
a benchmark, some sort of metric to know if they're winning.
34
00:02:12.400 --> 00:02:15.260
But they usually ask for something like auth rates,
35
00:02:15.320 --> 00:02:18.620
but like in Corinne's example, if you focus too much on off rates,
36
00:02:19.200 --> 00:02:22.380
you risk over optimizing one area at the expense of others.
37
00:02:22.920 --> 00:02:24.560
You might accept too much fraud,
38
00:02:25.420 --> 00:02:28.580
or you might rack up retry fees that really start impacting your margin.
39
00:02:29.740 --> 00:02:32.260
And managing these trade-offs is genuinely hard.
40
00:02:32.960 --> 00:02:36.300
By the time you spot a metric that might be going in the wrong direction,
41
00:02:36.780 --> 00:02:38.900
a lot of the times the damage is already done.
42
00:02:39.660 --> 00:02:43.040
And fixing it means you're coordinating across separate teams who might all be
43
00:02:43.180 --> 00:02:46.640
optimizing for a different number.
And it's really,
44
00:02:46.740 --> 00:02:50.320
really hard to drive the right outcome by just focusing on a single metric.
45
00:02:50.980 --> 00:02:53.340
Today, we're going to show you what to optimize instead.
46
00:02:53.920 --> 00:02:57.600
And then we'll hear from Microsoft and Wayfair on how they make these trade-offs
47
00:02:57.760 --> 00:02:58.593
in practice.
48
00:03:00.120 --> 00:03:01.480
So over the next 10 minutes,
49
00:03:01.840 --> 00:03:06.240
we'll discuss framework for evaluating the trade-offs between auth, fraud,
50
00:03:06.640 --> 00:03:09.320
and network fees. For auth rates,
51
00:03:09.540 --> 00:03:14.500
instead of tracking that one high-level number, the key is to ask,
52
00:03:14.760 --> 00:03:19.060
"Is my payment strategy increasing the lifetime value of my best
53
00:03:19.120 --> 00:03:23.800
customers?" Valuable customers are the ones who come back repeatedly,
54
00:03:24.540 --> 00:03:27.660
and we want them to have the most frictionless experience.
55
00:03:28.600 --> 00:03:33.200
If those repeat customers are converting at a lower rate than first-time
56
00:03:33.220 --> 00:03:36.140
buyers, you have yourself a very big retention problem.
57
00:03:37.040 --> 00:03:39.820
A failed payment risks losing them to a competitor.
58
00:03:40.300 --> 00:03:45.000
So when you focus on auth rate overall without drilling in at the customer
59
00:03:45.060 --> 00:03:47.800
level, you might be optimizing for the wrong thing.
60
00:03:48.420 --> 00:03:52.920
So we recommend protecting that lifetime value of your best customers.
61
00:03:53.360 --> 00:03:55.320
And here's how. First,
62
00:03:55.900 --> 00:04:00.100
focus on building strong foundations for all customers.
63
00:04:00.820 --> 00:04:05.240
Use tools like network tokens and card account updater to
64
00:04:05.320 --> 00:04:08.700
minimize lost sales from things like reissued cards.
65
00:04:10.580 --> 00:04:15.540
Then segment customers and tailor your approach for
66
00:04:15.700 --> 00:04:18.140
each type. Third,
67
00:04:18.600 --> 00:04:21.060
reduce friction for the best customers.
68
00:04:21.640 --> 00:04:26.020
Review fraud rules so they distinguish between a loyal subscriber and an unknown
69
00:04:26.080 --> 00:04:26.913
buyer.
70
00:04:27.480 --> 00:04:31.180
And review your checkout flows to make sure you have things like the local
71
00:04:31.220 --> 00:04:33.680
payment methods that your customers prefer.
72
00:04:34.980 --> 00:04:39.900
Combining all of these principles will improve auth where it matters
73
00:04:39.960 --> 00:04:42.420
most, and that's the strategy working.
74
00:04:44.020 --> 00:04:48.200
And every decision that you make on auth impacts our second pillar, fraud.
75
00:04:49.060 --> 00:04:50.520
This might sound a little bit odd,
76
00:04:50.760 --> 00:04:52.880
but your goal isn't actually to have zero fraud,
77
00:04:53.480 --> 00:04:56.440
even though that's probably the dream scenario for all of your trust and safety
78
00:04:56.520 --> 00:04:57.353
teams.
79
00:04:57.760 --> 00:05:01.220
The goal is actually to have the right fraud tolerance that balances between
80
00:05:01.280 --> 00:05:04.520
blocks and acceptance. If you have no fraud,
81
00:05:04.940 --> 00:05:08.640
it likely means you're probably being too restrictive with your fraud rules and
82
00:05:08.660 --> 00:05:10.920
you're probably blocking some legitimate customers.
83
00:05:11.720 --> 00:05:15.100
And if you're too permissive, you're absorbing unnecessary dispute costs,
84
00:05:15.280 --> 00:05:17.420
and trust me, issuers are going to notice.
85
00:05:18.100 --> 00:05:21.980
Our recommendation is to calibrate your fraud tolerance based on your margin.
86
00:05:22.680 --> 00:05:26.500
If you're a high-margin business like SaaS or digital goods,
87
00:05:27.000 --> 00:05:30.380
costs are a little bit lower. And that means fraud is still going to hurt,
88
00:05:30.560 --> 00:05:35.440
but maybe not as much as blocking a legitimate buyer.
And if you're a business
89
00:05:35.520 --> 00:05:39.680
with lower margins like physical goods where you have cost of goods sold and
90
00:05:39.740 --> 00:05:41.480
shipping and fulfillment costs,
91
00:05:41.900 --> 00:05:45.780
you might want to tighten controls even if you lose a few buyers that you're not
92
00:05:45.880 --> 00:05:46.713
quite sure about.
93
00:05:47.920 --> 00:05:51.800
And your fraud tolerance also shouldn't be a fixed number that you just set and
94
00:05:51.840 --> 00:05:54.660
forget. You need to adjust it for your current goals.
95
00:05:55.320 --> 00:05:59.880
Think about high growth periods like Black Friday and Cyber Monday or Way Day
96
00:05:59.920 --> 00:06:00.753
for Wayfair.
97
00:06:00.980 --> 00:06:04.400
You might want to loosen your rules to focus on customer acquisition.
98
00:06:05.100 --> 00:06:07.820
And then in quieter periods, focus on profitability.
99
00:06:08.420 --> 00:06:11.640
Your business is always changing, and so should your fraud tolerance.
100
00:06:12.700 --> 00:06:14.620
The third pillar is network fees,
101
00:06:14.980 --> 00:06:18.700
and this can be the most challenging and complex of the three to optimize.
102
00:06:19.400 --> 00:06:23.160
Because there are so many different types of fees and different variables that
103
00:06:23.220 --> 00:06:25.800
drive them. You have interchange, you have scheme fees.
104
00:06:25.860 --> 00:06:27.480
We've simplified this to one line,
105
00:06:27.540 --> 00:06:31.860
but there can be multiple scheme fees per transaction. Fees for authorization,
106
00:06:32.380 --> 00:06:35.280
behavioral fees, and expensive cross-border fees.
107
00:06:35.840 --> 00:06:38.020
And all of these really add up fast,
108
00:06:38.160 --> 00:06:40.680
so the savings here can be really significant.
109
00:06:41.780 --> 00:06:43.900
Because there's so much complexity in this pillar,
110
00:06:44.060 --> 00:06:46.280
it can feel really overwhelming to optimize,
111
00:06:46.600 --> 00:06:51.160
but it's critical to develop framework that focuses on your key
112
00:06:51.240 --> 00:06:52.600
spots for optimization.
113
00:06:53.800 --> 00:06:57.360
Let's say you decide to disable retries to reduce network fees.
114
00:06:57.880 --> 00:07:01.560
This can lead to lost revenue because your auth rates go down.
115
00:07:02.640 --> 00:07:06.640
You want to avoid a situation like this where you lose more revenue than you're
116
00:07:06.740 --> 00:07:11.720
saving from the network fee reduction.
So review your volume and margins and
117
00:07:11.840 --> 00:07:14.160
choose the best optimizations to match.
118
00:07:14.840 --> 00:07:18.640
Two levers that are likely worth considering that can make a really big
119
00:07:18.680 --> 00:07:21.370
difference: first, local acquiring.
120
00:07:21.830 --> 00:07:24.590
If you have business operations in multiple geographies,
121
00:07:24.910 --> 00:07:28.990
but only process through one US entity, for example,
122
00:07:29.510 --> 00:07:33.970
consider processing UK-domiciled transactions through a UK account
123
00:07:34.590 --> 00:07:37.410
and Canadian-issued transactions through a Canadian account.
124
00:07:37.950 --> 00:07:40.910
This can reduce those high cross-border fees I was mentioning.
125
00:07:42.250 --> 00:07:45.250
The second lever to consider is least cost routing.
126
00:07:45.950 --> 00:07:49.550
Automatically send transactions to the network with the lowest cost.
127
00:07:50.390 --> 00:07:53.650
Both of these levers can make a big dent in network costs,
128
00:07:53.970 --> 00:07:57.390
but they require serious engineering effort to build and maintain.
129
00:07:58.250 --> 00:08:00.890
Whether it's worth it depends on your business.
130
00:08:02.590 --> 00:08:07.050
For retailers processing billions on thin margins, every basis point matters.
131
00:08:07.630 --> 00:08:11.170
But a SaaS business with high margins and high recurring payments,
132
00:08:11.590 --> 00:08:16.330
that's just different math. Instead of investing eng time for cost optimization,
133
00:08:16.710 --> 00:08:18.850
focus on reducing that involuntary churn.
134
00:08:20.810 --> 00:08:24.870
So that's the framework. There's three pillars: auth rates, fraud,
135
00:08:25.290 --> 00:08:27.650
and network costs, each with their own trade-offs.
136
00:08:28.110 --> 00:08:30.990
And the takeaway here isn't to optimize all three equally.
137
00:08:31.530 --> 00:08:33.180
It's to know how to balance them together.
138
00:08:34.130 --> 00:08:37.110
Set your strategy intentionally and revisit it as it changes.
139
00:08:37.470 --> 00:08:40.210
There's no single number that's going to tell you that you're winning.
140
00:08:40.670 --> 00:08:44.130
The right answer actually depends on your margins, your customers,
141
00:08:44.510 --> 00:08:46.150
and your business goals. Now,
142
00:08:46.670 --> 00:08:48.670
let's make this concrete with a real world example.
143
00:08:49.650 --> 00:08:51.930
You just heard Corinne talk about lease cost routing,
144
00:08:52.170 --> 00:08:55.670
and PINless debit is a popular example here in the United States,
145
00:08:56.090 --> 00:08:59.330
and it is a perfect case study in the trade-offs that we were talking about.
146
00:09:00.490 --> 00:09:02.110
At first, it can be compelling.
147
00:09:02.850 --> 00:09:05.810
You get lower costs on some debit transactions.
148
00:09:06.550 --> 00:09:10.570
And there's pretty broad support-a lot of issuers do accept these transactions.
149
00:09:11.550 --> 00:09:15.490
And so that means the savings can apply across a pretty meaningful portion of
150
00:09:15.530 --> 00:09:19.350
your payment volume.
But there are some real trade-offs here.
151
00:09:20.230 --> 00:09:22.210
First off, it can lower your auth rates,
152
00:09:22.470 --> 00:09:25.730
and we've seen this to be especially true on those really high-value
153
00:09:25.750 --> 00:09:26.583
transactions.
154
00:09:27.250 --> 00:09:31.370
You can also lose advanced network features like real-time card account updater
155
00:09:31.630 --> 00:09:34.990
or multicapture, sometimes features that are critical for your business,
156
00:09:36.070 --> 00:09:39.730
and you'll almost certainly see increased ops and engineering costs.
157
00:09:40.270 --> 00:09:43.150
You now have to manage routing across multiple networks,
158
00:09:43.270 --> 00:09:45.150
deal with different chargeback processes,
159
00:09:45.610 --> 00:09:49.250
and you have to monitor performance often at a BIN level across multiple
160
00:09:49.290 --> 00:09:52.310
networks. So is it worth it?
161
00:09:52.850 --> 00:09:55.510
You probably have guessed the answer by now that it depends.
162
00:09:56.230 --> 00:10:00.730
It'll depend on your cost sensitivity, your comfort with lower auth rates,
163
00:10:01.130 --> 00:10:04.050
your engineering and ops capacity, and your business model.
164
00:10:04.670 --> 00:10:07.430
So at the end of the day, there's no universal right answer,
165
00:10:08.010 --> 00:10:09.870
only the right answer for your business,
166
00:10:10.470 --> 00:10:11.930
and that's true well beyond PINless debit.
167
00:10:13.590 --> 00:10:18.370
So picture two retailers of similar size that operate in a comparable
168
00:10:18.410 --> 00:10:22.810
market. They're peers with the same goal to maximize profits,
169
00:10:23.150 --> 00:10:27.250
but they take two very different approaches. First,
170
00:10:27.310 --> 00:10:31.690
we have Raymond's Refined Rugs that has low margins on high-end carpets.
171
00:10:32.010 --> 00:10:36.070
So when a $2,000 rug is shipped to a fraudster, it really hurts.
172
00:10:36.590 --> 00:10:41.190
So his business prioritizes reducing cost and fraud and is willing to
173
00:10:41.270 --> 00:10:43.110
accept lower auth rates.
174
00:10:43.830 --> 00:10:47.130
But Corinne's Comfy Chairs? She has a very different philosophy.
175
00:10:48.090 --> 00:10:51.010
She sells chairs at a high volume with decent margins,
176
00:10:51.410 --> 00:10:54.130
so a fraudulent order on $50 stings,
177
00:10:54.230 --> 00:10:58.150
but probably not as much as turning away a legitimate buyer who was never going
178
00:10:58.170 --> 00:10:59.003
to come back.
179
00:10:59.350 --> 00:11:03.210
She prioritizes maximizing auth rates while accepting slightly higher disputes
180
00:11:03.310 --> 00:11:04.143
and fees.
181
00:11:05.330 --> 00:11:09.290
Raymond is protecting margin and I'm chasing every sale.
182
00:11:09.790 --> 00:11:13.290
I feel like there's a joke here about how our hypothetical businesses might
183
00:11:13.350 --> 00:11:14.710
reflect our real personalities,
184
00:11:14.790 --> 00:11:17.430
so I think we're going to do some deep dives into that over lunch later.
185
00:11:17.890 --> 00:11:20.750
But which is right? We both are.
186
00:11:21.130 --> 00:11:23.470
Raymond does have lower auth rates and less fraud.
187
00:11:24.070 --> 00:11:27.710
Even though I have more fraud, my higher auth rates make up for the difference.
188
00:11:28.250 --> 00:11:31.250
Both are profitable and performing well. Why?
189
00:11:31.690 --> 00:11:34.070
Because our strategies align to our business models.
190
00:11:34.930 --> 00:11:38.130
And if I tried to match Corinne's auth rates while they sound great,
191
00:11:38.510 --> 00:11:41.790
I'd actually increase my fraud and undermine my entire strategy.
192
00:11:43.270 --> 00:11:44.730
And the inverse is true too.
193
00:11:45.050 --> 00:11:48.670
The key takeaway here is that you aren't trying to be another business.
194
00:11:48.950 --> 00:11:51.370
You want to be the best version of your own business.
195
00:11:52.610 --> 00:11:54.790
Even if your peers' numbers look better,
196
00:11:55.190 --> 00:11:57.450
their strategy might be wrong for your business.
197
00:11:57.690 --> 00:12:01.790
And I feel like this is the same life lesson I tell my middle schooler at home.
198
00:12:02.270 --> 00:12:06.750
When it comes to payment optimization, just like life, worry about yourself,
199
00:12:07.030 --> 00:12:08.870
not your peers. My poor children.
200
00:12:10.910 --> 00:12:15.530
And if you're thinking like, this feels like it's a lot of manual work,
201
00:12:16.110 --> 00:12:20.150
you're right, and it is. Balancing these trade-offs is legitimately hard,
202
00:12:20.450 --> 00:12:22.490
and this is where having the right tools can help.
203
00:12:23.270 --> 00:12:25.230
Across all three of these pillars,
204
00:12:25.630 --> 00:12:28.950
Stripe provides AI-powered tools to help you get optimized.
205
00:12:29.770 --> 00:12:31.230
We have Authorization Boost,
206
00:12:31.470 --> 00:12:35.630
which intelligently adjusts transaction messages in real time to approve your
207
00:12:35.710 --> 00:12:39.150
authorization rates while simultaneously lowering your costs.
208
00:12:39.890 --> 00:12:43.650
And something new in Auth Boost is you can now A/B test it against your current
209
00:12:43.710 --> 00:12:46.830
stack so you can see the performance before you commit.
210
00:12:47.990 --> 00:12:49.970
We also have Radar, our fraud solution,
211
00:12:50.130 --> 00:12:52.830
which allows you to set personalized risk preferences.
212
00:12:53.450 --> 00:12:57.570
You can now specifically decide how you want to balance out fraud versus
213
00:12:57.610 --> 00:13:02.290
revenue. What's great about Radar is it also learns with your business.
214
00:13:02.650 --> 00:13:05.490
It'll automatically adapt as your business grows and changes.
215
00:13:06.750 --> 00:13:10.570
And the point here ultimately is not to hand over your optimization plan to
216
00:13:10.630 --> 00:13:14.600
someone else.
It's to reduce the manual balancing your team has to do so you can
217
00:13:14.760 --> 00:13:18.200
focus on strategy and not tuning. Now,
218
00:13:18.480 --> 00:13:21.760
let's hear from two businesses that apply this framework at scale today.
219
00:13:21.960 --> 00:13:23.960
Please join me in welcoming Ana Leite,
220
00:13:24.040 --> 00:13:27.200
director of global payments from Microsoft and Curtis Crawford,
221
00:13:27.240 --> 00:13:29.960
director and GM of fintech and loyalty at Wayfair. Welcome.
222
00:13:30.160 --> 00:13:30.600
Thank you.
223
00:13:30.600 --> 00:13:31.433
Thank you.
224
00:13:31.480 --> 00:13:32.520
Well, I know you a little bit,
225
00:13:32.680 --> 00:13:36.040
but hopefully we can get the crowd to know you a little bit more first and maybe
226
00:13:36.080 --> 00:13:37.080
Ana will start with you.
227
00:13:37.520 --> 00:13:40.960
Would love to hear how you got into payments and what your payments world looks
228
00:13:40.960 --> 00:13:41.793
like today.
229
00:13:42.040 --> 00:13:44.640
Yeah. Ana Leite, nice to meet you. Thank you for coming.
230
00:13:44.800 --> 00:13:49.720
I know lunch is going on. I work in payments since college.
231
00:13:50.000 --> 00:13:53.520
So my internship in college was already developing credit cards,
232
00:13:53.600 --> 00:13:55.360
apparently it was my mission on earth.
233
00:13:56.040 --> 00:14:00.760
And I have worked for banks and issuer side, loyalty programs,
234
00:14:01.240 --> 00:14:04.320
and also for merchants for American Express.
235
00:14:04.440 --> 00:14:08.200
So I've done a lot in the industry, but that's mainly what I have done.
236
00:14:08.760 --> 00:14:11.760
And today I work... For the past eight years,
237
00:14:11.840 --> 00:14:15.680
I work for Microsoft in the payments group.
238
00:14:16.240 --> 00:14:20.120
And it's interesting to tell you the complexity of what we do there.
239
00:14:20.320 --> 00:14:25.280
So we manage one payments platform that supports all lines of businesses that
240
00:14:25.440 --> 00:14:29.240
we have, almost all, but it's going to be B2B, B2C,
241
00:14:29.840 --> 00:14:33.960
different industries like advertising, software, physical goods,
242
00:14:34.520 --> 00:14:36.920
gaming, cloud, and so on.
243
00:14:38.080 --> 00:14:42.000
We have different business models. So onetime sales,
244
00:14:42.240 --> 00:14:46.600
recurring sales, prepaid, postpaid. It's a little complex.
245
00:14:46.800 --> 00:14:51.480
And for all of us who like payments, it's kind of a fun thing,
246
00:14:51.680 --> 00:14:52.513
a fun job to have.
247
00:14:53.080 --> 00:14:53.960
Yeah, a little bit of everything.
248
00:14:55.120 --> 00:14:58.720
Yeah. And that with about 40 providers in one platform.
249
00:14:59.120 --> 00:15:03.720
And we are in every market where an American company can make business,
250
00:15:04.200 --> 00:15:06.720
can do business. So that's it. That's me.
251
00:15:07.040 --> 00:15:10.360
Wow. Excited to learn from you today. Curtis, how about you?
252
00:15:10.400 --> 00:15:13.080
How did you get into payments and what does payments look like for you today?
253
00:15:13.480 --> 00:15:16.120
So my first job in payments was 25 years ago,
254
00:15:16.560 --> 00:15:21.360
and it was a marketing and financing company that worked on the back of credit
255
00:15:21.440 --> 00:15:21.880
cards.
256
00:15:21.880 --> 00:15:26.200
We would give airline miles to customers who went into certain restaurants.
257
00:15:26.480 --> 00:15:30.720
There was actually like a paper book. It's been a minute. Okay.
258
00:15:30.880 --> 00:15:35.280
So what does my payment world look like now? Between now and then, let's see,
259
00:15:35.400 --> 00:15:39.760
I worked for a credit card processor for a few years. I am at Wayfair now.
260
00:15:39.800 --> 00:15:41.720
Before that, I was at Amazon for six years.
261
00:15:42.560 --> 00:15:45.960
I spent two in Seattle working on their co-branded credit card,
262
00:15:46.120 --> 00:15:46.953
went over to Europe,
263
00:15:47.000 --> 00:15:49.670
which was really awesome to be in payments in Europe for a few years,
264
00:15:49.730 --> 00:15:53.210
and then came to Wayfair to look after their payments org,
265
00:15:53.270 --> 00:15:56.230
which is payment processing, payment fraud,
266
00:15:56.670 --> 00:16:00.910
and then our stack of payment products, things like gift cards, financing,
267
00:16:01.030 --> 00:16:02.180
and our co-branded credit card.
268
00:16:03.110 --> 00:16:07.030
Awesome. That's a great huge breadth of experience across both of you. And so,
269
00:16:07.630 --> 00:16:10.950
we just talked about three critical pillars of payments: auth rates, fraud,
270
00:16:11.010 --> 00:16:11.730
and costs,
271
00:16:11.730 --> 00:16:14.470
and I'm sure this shows up in your world in a lot of different places.
272
00:16:14.790 --> 00:16:16.090
Maybe again, we'll start with Ana.
273
00:16:16.490 --> 00:16:19.010
How do you think about balancing those together and the trade-offs?
274
00:16:19.430 --> 00:16:21.230
Well, that's the beauty of it, right?
275
00:16:22.090 --> 00:16:24.830
I always joke that if you don't want to have fraud, close your business.
276
00:16:25.270 --> 00:16:28.170
Zero risk, bureau fraud, no problem at all. Easy. Right? Easy.
277
00:16:28.810 --> 00:16:32.670
But once you start selling the first sale, you are already at risk.
278
00:16:32.730 --> 00:16:37.130
So we need to always make sure that we are fortunate to be in an industry where
279
00:16:37.150 --> 00:16:42.110
we do not have to worry about cost as much as in other industries.
280
00:16:42.530 --> 00:16:45.290
Although it is an important, for sure, pillar for us,
281
00:16:45.690 --> 00:16:50.650
we can rely more on focusing on customer experience and authorization rate,
282
00:16:51.070 --> 00:16:55.990
but sometimes the same customer is consuming very different
283
00:16:56.010 --> 00:16:59.190
products from us like Cloud, Office, and Gaming.
284
00:16:59.990 --> 00:17:02.110
And as we treat fraud, for example,
285
00:17:02.310 --> 00:17:05.890
one thing is if you block him from buying something for his game. The other,
286
00:17:06.730 --> 00:17:10.370
if you block his cloud service or his office,
287
00:17:10.430 --> 00:17:14.210
now this person doesn't work and it's kind of complicated, right?
288
00:17:15.090 --> 00:17:16.250
It's beyond frustration.
289
00:17:16.690 --> 00:17:20.890
So that's where using all the tools and everything that you guys were talking
290
00:17:20.950 --> 00:17:25.510
about becomes fun because you need to balance them also
291
00:17:25.630 --> 00:17:27.830
across the different business lines that we have.
292
00:17:28.150 --> 00:17:28.330
Yeah.
293
00:17:28.330 --> 00:17:31.750
And so would you say that one of those metrics between auth cost and fraud is
294
00:17:31.790 --> 00:17:33.130
more important than the others?
295
00:17:33.990 --> 00:17:36.850
Yeah. Auth for sure because as I said,
296
00:17:36.910 --> 00:17:41.670
customer experience in some of our products are vital for customers to work
297
00:17:42.430 --> 00:17:46.110
and you cannot really make a mistake there, but at the same time,
298
00:17:46.670 --> 00:17:49.650
you don't want to open your business for fraud that much.
299
00:17:49.730 --> 00:17:54.470
So we do have our own risk platform that comes
300
00:17:54.550 --> 00:17:59.210
before our providers and that we use all the data that we know,
301
00:17:59.470 --> 00:18:03.330
all the information that is in our ecosystem to make decisions.
302
00:18:03.670 --> 00:18:06.250
But once the transactions are out of my ecosystem,
303
00:18:06.330 --> 00:18:10.630
then I leverage what the providers see because they see more than Microsoft,
304
00:18:10.730 --> 00:18:13.850
right? They know other sides of that one customer.
305
00:18:14.210 --> 00:18:18.890
So also balancing these two things is part of the game. Yeah.
306
00:18:19.070 --> 00:18:21.070
Yeah. Sounds a little complex sometimes.
307
00:18:21.530 --> 00:18:22.230
It's fun.
308
00:18:22.230 --> 00:18:26.670
Yeah. Curtis, how about at Wayfair? How do you balance out between auth rates,
309
00:18:26.730 --> 00:18:27.563
fraud, and costs?
310
00:18:27.910 --> 00:18:31.630
Sure. So we actually zoom out. Our uber KPI is called total cost of payments,
311
00:18:32.270 --> 00:18:34.510
and it's a full P&L. So you hear a lot,
312
00:18:34.690 --> 00:18:39.390
manage your payment stack to grow business value. And so yes,
313
00:18:39.630 --> 00:18:42.310
our total cost of payments, there's line items for the cost.
314
00:18:42.370 --> 00:18:45.210
You're going to see what are your vendor fees, all of that,
315
00:18:45.350 --> 00:18:48.330
but there's a lot of revenue lines. So for example,
316
00:18:48.390 --> 00:18:50.890
can you drive new customer acquisitions? At Wayfair,
317
00:18:50.950 --> 00:18:54.790
we have a relatively high AOV, so we do a lot of financing on the platforms.
318
00:18:54.850 --> 00:18:59.090
We work with our financing partners to one market on our site.
319
00:18:59.150 --> 00:19:03.310
So we are doing marketing campaigns through our payment methods.
320
00:19:03.770 --> 00:19:06.770
Those drive us new customers. We take credit for that as a payments org.
321
00:19:07.890 --> 00:19:12.030
We manage the... When you think about customer lifetime, loyalty,
322
00:19:12.750 --> 00:19:17.070
Wayfair, we're always doing a lot of paid marketing. We find that folks,
323
00:19:17.130 --> 00:19:18.890
when we build a relationship with them,
324
00:19:19.010 --> 00:19:22.130
they come to us first so we can skip the paid marketing.
So that's another
325
00:19:22.170 --> 00:19:26.130
benefit. So we, again, it's total cost of payments,
326
00:19:26.190 --> 00:19:29.810
but it's a full P&L. And then over time, that P&L should get better.
327
00:19:30.530 --> 00:19:33.630
And we really try to be active about managing our payment method mix,
328
00:19:34.170 --> 00:19:35.003
not in a way that's...
329
00:19:35.430 --> 00:19:37.470
We don't want to steer folks away from something they love,
330
00:19:37.530 --> 00:19:38.590
but we do want to be helpful.
331
00:19:39.190 --> 00:19:41.810
We do want to let folks know if we're running special promotions,
332
00:19:41.870 --> 00:19:45.390
things like that. And then in terms of the trade off,
333
00:19:46.610 --> 00:19:48.610
fraud is one of those bits. Ideally,
334
00:19:48.710 --> 00:19:50.610
authorization rates just get better over time,
335
00:19:50.670 --> 00:19:51.970
and maybe there's some blips here and there.
336
00:19:52.390 --> 00:19:56.130
I think fraud is probably the exception where the range of outcomes is wider,
337
00:19:57.410 --> 00:20:00.210
but you have to have tolerance for that. If there's an attack,
338
00:20:00.270 --> 00:20:03.310
then you need to clamp down and kind of make the fraudsters go away for a while.
339
00:20:03.650 --> 00:20:06.510
And maybe that changes the overall picture for a little bit,
340
00:20:06.570 --> 00:20:09.410
but that's the right move to make. And so again,
341
00:20:09.470 --> 00:20:11.710
if you're just managing quarter by quarter over time,
342
00:20:13.030 --> 00:20:15.710
ideally things get better over the long horizon.
343
00:20:16.250 --> 00:20:19.110
Yeah. And I love the similarities and differences between your answers there,
344
00:20:19.210 --> 00:20:19.890
right? Ana,
345
00:20:19.890 --> 00:20:24.350
it sounds like you guys are hyper-focused on customer experiences and Curtis at
346
00:20:24.390 --> 00:20:28.010
Wayfair, you are trying to come up with a metric that balances out all three.
347
00:20:28.410 --> 00:20:29.570
But as you both mentioned,
348
00:20:29.770 --> 00:20:31.570
things happen like fraud attacks and we have to adjust.
349
00:20:31.630 --> 00:20:33.670
So it's really interesting to see your different,
350
00:20:33.730 --> 00:20:34.930
but like similar approaches there.
351
00:20:35.370 --> 00:20:39.050
Yeah. People ask this like, "Oh, what's the number?" You want to have what,
352
00:20:39.140 --> 00:20:43.230
95 approval rate? Well, if I'm being attacked, fraud attack, no,
353
00:20:43.570 --> 00:20:46.450
I want to have a very low approval rate because means they are not being
354
00:20:46.470 --> 00:20:49.130
successful. So that's exactly what you're mentioning.
355
00:20:49.190 --> 00:20:50.630
So there is no magic number.
356
00:20:50.910 --> 00:20:52.470
Right. Better is good.
357
00:20:53.050 --> 00:20:57.790
Better is good. Harder to find better. Ana, question for you.
358
00:20:58.610 --> 00:21:02.170
Is there a time you can think about at Microsoft or otherwise where you're
359
00:21:02.190 --> 00:21:03.330
trying to improve one metric,
360
00:21:03.430 --> 00:21:05.310
but it's actually causing pressure in another area?
361
00:21:06.230 --> 00:21:06.890
Yeah.
362
00:21:06.890 --> 00:21:11.810
Especially when new technologies come like tokens or tokens not so
363
00:21:11.890 --> 00:21:12.130
much,
364
00:21:12.130 --> 00:21:16.670
but 3DS authentication and customers are going to now learn that they have to go
365
00:21:16.830 --> 00:21:17.850
through authentication.
366
00:21:18.530 --> 00:21:22.770
When PSD2 came in Europe and many merchants here probably faced the same,
367
00:21:23.110 --> 00:21:27.730
banks were still adjusting and then declining more transactions
368
00:21:27.790 --> 00:21:32.150
because of liability shift when we were sending authenticated transactions.
369
00:21:32.610 --> 00:21:35.450
So there are moments where you have to balance that.
370
00:21:35.510 --> 00:21:36.930
And also we are doing that right now,
371
00:21:36.990 --> 00:21:41.130
expanding 3DS in Latin America and you need to be able to
372
00:21:41.950 --> 00:21:46.870
tweak how much you send and work with the banks to make sure that their
373
00:21:46.990 --> 00:21:48.050
portion of the business,
374
00:21:48.110 --> 00:21:52.850
their side is being addressed properly so they don't impact my customer and then
375
00:21:52.950 --> 00:21:54.070
my authorization rate.
376
00:21:54.570 --> 00:21:59.190
And I may be spending more because I'm retrying without the
377
00:21:59.250 --> 00:22:04.110
authentication or to another provider or whatever my strategy
378
00:22:04.250 --> 00:22:08.120
is. So definitely cost then suffers a little,
379
00:22:08.640 --> 00:22:10.760
authorization suffers a little, but then
380
00:22:13.080 --> 00:22:15.730
the goal is reaching stability again as fast as we can.
381
00:22:16.180 --> 00:22:19.020
Yeah. And you touched on something really interesting there where you kind of
382
00:22:19.060 --> 00:22:21.300
mentioned 3DS in different regions.
383
00:22:21.420 --> 00:22:23.480
And part of this is like the ecosystem changing too,
384
00:22:23.900 --> 00:22:26.620
where like your result today is probably going to look different than results
385
00:22:26.680 --> 00:22:29.440
later if customers are getting used to the flows or if we're going to use to
386
00:22:29.460 --> 00:22:31.620
them. So how do you deal with all those changes?
387
00:22:32.060 --> 00:22:36.060
Yeah. I was like years ago when I was working for an issuer,
388
00:22:36.120 --> 00:22:38.940
we were implementing in Brazil chips on the cards.
389
00:22:39.320 --> 00:22:42.490
It was the first market to do it in the world.
390
00:22:43.080 --> 00:22:46.760
And now you need to sit all the banks together and decide what the customer
391
00:22:46.800 --> 00:22:51.040
experience will be because if it's not consistent, then the fraudsters,
392
00:22:51.460 --> 00:22:54.660
it becomes amazing because whatever is in front of the customer,
393
00:22:54.720 --> 00:22:57.730
the customer will buy it, right? We will think it's right.
394
00:22:58.200 --> 00:22:59.660
So we have to do that.
395
00:23:00.040 --> 00:23:04.160
And it's same with 3DS that we are doing and other changes.
396
00:23:04.220 --> 00:23:07.120
If we are to make other changes on customer experience,
397
00:23:07.180 --> 00:23:09.160
there is this learning curve and there is this
398
00:23:09.600 --> 00:23:12.800
teaching-the-customer-what-to-expect portion of it.
399
00:23:13.840 --> 00:23:14.360
Yeah. Yeah.
400
00:23:14.360 --> 00:23:17.800
There's definitely a piece of training your customers for free close as well and
401
00:23:17.820 --> 00:23:18.653
start to-.
402
00:23:18.680 --> 00:23:21.990
And aligning with the other. So for example, agentic commerce now,
403
00:23:22.400 --> 00:23:25.600
we are all talking about agentic commerce. We are always super excited.
404
00:23:25.730 --> 00:23:28.700
We made it 20 minutes, I think. We did good.
405
00:23:28.920 --> 00:23:30.220
Yeah. But you know what I mean?
406
00:23:31.800 --> 00:23:36.460
If we don't have an experience that it's clear for our customers what to expect
407
00:23:36.500 --> 00:23:41.500
when these agents working for the fraudsters, it's like heaven,
408
00:23:41.600 --> 00:23:41.880
right?
409
00:23:41.880 --> 00:23:46.720
Whatever they put out there is going to work and then how our risk systems are
410
00:23:46.740 --> 00:23:48.900
going to adapt. So all that is part of ... Yeah,
411
00:23:49.240 --> 00:23:50.770
I think I answered your question.
412
00:23:50.780 --> 00:23:54.200
Yes. And Curtis,
413
00:23:54.940 --> 00:23:58.520
I know unfortunately our key metrics don't always go up into the right like the
414
00:23:58.540 --> 00:24:02.060
way we want it to. And when it maybe waivers in the wrong direction,
415
00:24:02.200 --> 00:24:05.180
we all wish we could have that magical dashboard that just tells you, "Hey,
416
00:24:05.240 --> 00:24:06.320
this is exactly what's going on.
417
00:24:06.400 --> 00:24:08.720
Here's a core driver," but that's not always what happens.
418
00:24:08.820 --> 00:24:12.200
What are some of the second and third and four things that you look at when
419
00:24:12.220 --> 00:24:15.480
you're really trying to dive deep into what's going on with a certain metric?
420
00:24:15.820 --> 00:24:19.640
Sure. So Wayfair operates in four countries.
421
00:24:20.500 --> 00:24:24.460
We have five different brands. We have Parigold, Joss & Main, AllModern,
422
00:24:24.540 --> 00:24:27.600
Birch Lane, B2B, we have B2C.
423
00:24:28.840 --> 00:24:33.140
We have physical retail, we have a growing physical retail footprint. So yes,
424
00:24:33.200 --> 00:24:34.600
we look at the high-level metrics,
425
00:24:34.680 --> 00:24:37.100
but there can be a lot going on underneath those.
426
00:24:37.460 --> 00:24:40.760
So we really never take our eye off what's going on underneath.
427
00:24:41.220 --> 00:24:44.740
That's where the alerting is, that's where our analysts are focused,
428
00:24:44.840 --> 00:24:46.420
that's where we're pointing AI to say,
429
00:24:47.160 --> 00:24:50.060
"Is there anything going on here that might not even be bubbling up to kind of
430
00:24:50.080 --> 00:24:52.540
those headline numbers?" And then even from there,
431
00:24:52.640 --> 00:24:55.380
you can start looking at things like device type.
432
00:24:56.420 --> 00:24:57.880
Which PSP did you route it through?
433
00:24:58.340 --> 00:25:01.140
Do you have any systems where you can go actually see what happened on your
434
00:25:01.180 --> 00:25:02.540
website if there was error, right?
435
00:25:02.880 --> 00:25:05.800
Can I play back what that customer experience was?
436
00:25:06.000 --> 00:25:10.620
So we are always operating at a very granular level to watch for issues,
437
00:25:11.560 --> 00:25:15.320
watch for changes and trend, and that's kind of where the analysis starts.
438
00:25:16.800 --> 00:25:16.940
Yeah,
439
00:25:16.940 --> 00:25:21.060
that makes a lot of sense where you have to start somewhere and then really,
440
00:25:21.120 --> 00:25:22.680
really, if I'm pulling out a theme here,
441
00:25:22.740 --> 00:25:26.060
it's like really segmentation where you're looking at your customers and there's
442
00:25:26.100 --> 00:25:27.480
like almost infinite ways to segment.
443
00:25:27.540 --> 00:25:31.360
So it becomes a challenge to prioritize some time.
444
00:25:32.440 --> 00:25:34.360
Ana, you mentioned agentic.
445
00:25:34.840 --> 00:25:38.080
I want to talk a little bit about agentic because it's something that a lot of
446
00:25:38.140 --> 00:25:39.140
merchants are thinking about.
447
00:25:39.500 --> 00:25:43.080
And it's not just a brand new channel for a particular merchant like Microsoft
448
00:25:43.100 --> 00:25:44.960
Wafer, but it's kind of brand new for everybody.
449
00:25:45.480 --> 00:25:48.720
So it is really hard to tell what's good for off rates,
450
00:25:48.780 --> 00:25:50.720
fraud and cost or total cost of payments.
451
00:25:51.680 --> 00:25:54.060
How do you think about with a brand new channel like this,
452
00:25:54.300 --> 00:25:55.440
what's good and what's not,
453
00:25:55.500 --> 00:25:58.540
and could it be better or are we actually lucky and doing really well?
454
00:25:58.600 --> 00:25:59.460
How do you think about that?
455
00:26:00.480 --> 00:26:04.940
Yeah, I think it's ... I'm super excited, honestly. And being at Microsoft,
456
00:26:05.040 --> 00:26:07.200
I see this with two different hats,
457
00:26:07.440 --> 00:26:11.560
one hat of being a merchant that is going to receive requests,
458
00:26:11.840 --> 00:26:14.380
purchases coming from an agent.
459
00:26:15.300 --> 00:26:18.840
And there is also the side of Microsoft where we are the browser,
460
00:26:18.920 --> 00:26:20.100
we are Copilot,
461
00:26:20.200 --> 00:26:24.120
we are also generating those sales for other merchants.
462
00:26:24.560 --> 00:26:27.900
And my team is involved in both to some level.
463
00:26:29.040 --> 00:26:33.820
And we've been also collaborating with Stripe and other players to
464
00:26:33.900 --> 00:26:35.120
understand, okay,
465
00:26:35.180 --> 00:26:39.100
we may use still metrics like authorization and fraud,
466
00:26:39.480 --> 00:26:42.540
but what else should we be measuring here?
467
00:26:43.040 --> 00:26:48.000
And how can we send like a signal like the issuing banks
468
00:26:48.120 --> 00:26:53.080
want to know if this transaction is coming originated from a human in a
469
00:26:53.600 --> 00:26:58.000
traditional mode or using an agent or maybe in the future
470
00:26:58.020 --> 00:27:01.780
initiated by an agent when the owner of that agent is sleeping.
471
00:27:02.420 --> 00:27:06.700
So how is this going to work? And then comes the conversations with the schemes.
472
00:27:06.760 --> 00:27:08.160
What about liability shift?
473
00:27:08.220 --> 00:27:11.220
If something goes wrong with a transaction done with an agent,
474
00:27:11.660 --> 00:27:14.120
who should we blame? Right?
475
00:27:14.180 --> 00:27:15.560
Yeah, that's a question I get all the time.
476
00:27:15.680 --> 00:27:19.420
All the time. And I wish I had all the answers. I would be a millionaire by now,
477
00:27:20.100 --> 00:27:24.640
but none of us has all the answers because all this thing is being born and we
478
00:27:24.760 --> 00:27:29.000
are in one of those moments where we are all here lucky to be living through,
479
00:27:29.100 --> 00:27:33.260
in my view, to be living through this moment because this is big,
480
00:27:33.340 --> 00:27:34.740
big change ahead of us.
481
00:27:35.000 --> 00:27:37.560
Yeah. Definitely an exciting time for payments. Curtis,
482
00:27:37.740 --> 00:27:40.380
how's Wayfair thinking about agenttic and what's good and what's not on the
483
00:27:40.600 --> 00:27:41.560
payments balancing act?
484
00:27:41.940 --> 00:27:42.280
Sure.
485
00:27:42.280 --> 00:27:46.980
So from the moment something or someone hits the purchase now button,
486
00:27:47.580 --> 00:27:51.200
your KPIs and your measurement might look similar. What was my auth rate?
487
00:27:51.460 --> 00:27:53.360
What's the return rate? All the bits.
488
00:27:53.840 --> 00:27:57.240
The upper funnel gets really interesting and it hasn't quite come into focus
489
00:27:57.280 --> 00:28:00.240
yet, right? So if you're thinking about end-to-end conversion,
490
00:28:00.320 --> 00:28:02.980
someone like Wayfair might think, how many impressions did I have?
491
00:28:03.340 --> 00:28:05.620
How many folks went to a product page, then the cart,
492
00:28:05.680 --> 00:28:07.440
how many folks started the checkout journey?
493
00:28:08.000 --> 00:28:09.680
All of that's going to look a lot different.
494
00:28:10.000 --> 00:28:14.420
And then in terms of just traffic on your site, is more good? Is it bad?
495
00:28:14.500 --> 00:28:17.520
Is it pointless? That really hasn't come into focus yet.
496
00:28:18.560 --> 00:28:21.060
That's where I think really helpful partners come into play.
497
00:28:21.540 --> 00:28:24.260
That's where events like this come into play where you build your network and
498
00:28:24.340 --> 00:28:25.240
say, "What are you seeing?
499
00:28:25.300 --> 00:28:29.220
How are you thinking about it?" Some of our partners have great dashboards
500
00:28:29.560 --> 00:28:32.920
beyond what we're capable of doing to just see how much traffic's out there to
501
00:28:33.000 --> 00:28:35.300
let us know how many conversions are going, right?
502
00:28:35.380 --> 00:28:39.240
If somebody else is just doing more sales from you than you at this point,
503
00:28:39.680 --> 00:28:42.820
is there some experience that you don't have? Is there a discoverability issue?
504
00:28:42.880 --> 00:28:47.280
So we're still solving that upper funnel bit and kind of that beginning of the
505
00:28:47.320 --> 00:28:50.480
journey, but towards the end, it gets a bit more familiar.
506
00:28:51.560 --> 00:28:52.460
Yeah, that makes a lot of sense.
507
00:28:52.520 --> 00:28:54.920
And I think we're all on this journey together and hopefully this is an area
508
00:28:55.540 --> 00:28:58.960
where Stripe can help both of you and show you what we're seeing as well.
509
00:28:59.920 --> 00:29:02.940
But I think that's unfortunately all the time we have today.
510
00:29:03.280 --> 00:29:05.600
If anyone wants to learn more about the payments balancing act,
511
00:29:05.800 --> 00:29:08.420
you can come visit us at our payments booth. Next we haul,
512
00:29:08.500 --> 00:29:12.980
we'll have experts there from our product teams or sales teams all day to talk
513
00:29:13.000 --> 00:29:16.840
to you about how you can think about balancing out your auth rates, fraud,
514
00:29:16.860 --> 00:29:18.620
and costs. Thank you.
515
00:29:18.880 --> 00:29:19.300
Thank you.
516
00:29:19.300 --> 00:29:19.540
Thank you.