WEBVTT
1
00:00:09.600 --> 00:00:13.600
My name is Abhi Tiwari and I lead our global product team here at Stripe.
2
00:00:14.480 --> 00:00:16.100
Today we're going to talk about global expansion.
3
00:00:17.000 --> 00:00:19.520
Think about what it used to take to sell internationally.
4
00:00:20.380 --> 00:00:24.020
You would typically start in your home market, let's say the US.
5
00:00:24.860 --> 00:00:28.080
You would spend time building your products, looking for customers,
6
00:00:28.480 --> 00:00:30.840
and really focus on winning in your domestic place.
7
00:00:31.820 --> 00:00:36.200
Then you might decide that you want to go to the next market and you'll pick
8
00:00:36.840 --> 00:00:38.940
the UK because the language is the same.
9
00:00:40.300 --> 00:00:43.360
Then you would have to go set up a legal entity.
10
00:00:44.020 --> 00:00:47.260
You'd have to hire a country manager, set up a bank account,
11
00:00:48.340 --> 00:00:52.140
start winning customers locally, and maybe six months later,
12
00:00:52.560 --> 00:00:54.280
you might get your first transaction.
13
00:00:55.580 --> 00:00:57.580
And then when you decide to go to your next market,
14
00:00:57.940 --> 00:01:02.440
you would have to do this whole thing all over again. Now,
15
00:01:03.040 --> 00:01:07.480
let's look at what's happening today. In the last five years,
16
00:01:07.680 --> 00:01:11.060
the number of businesses on Stripe that are selling in more than a hundred
17
00:01:11.100 --> 00:01:12.920
countries has quadrupled.
18
00:01:14.200 --> 00:01:18.840
And much of this is down to AI companies like OpenAI, Anthropic,
19
00:01:19.280 --> 00:01:23.400
Perplexity, and Midjourney. But it's not just AI.
20
00:01:23.980 --> 00:01:28.100
We see companies of all shapes and sizes going along this trend.
21
00:01:29.160 --> 00:01:30.540
For example, Halara,
22
00:01:31.300 --> 00:01:36.000
a B2C athleisure brand that saw tremendous traction on social media.
23
00:01:36.860 --> 00:01:41.180
They partnered with Stripe to expand into seven new countries with each country
24
00:01:41.320 --> 00:01:44.060
taking only three days to launch.
25
00:01:45.180 --> 00:01:49.660
What's replaced the old playbook is a fundamentally different starting point.
26
00:01:51.340 --> 00:01:54.100
Your domestic market isn't just your home country anymore,
27
00:01:54.580 --> 00:01:56.000
it's the entire internet.
28
00:01:58.060 --> 00:02:02.760
But converting reach into revenue requires solving three important
29
00:02:02.780 --> 00:02:07.480
problems. First, you have to start with localizing the buyer experience.
30
00:02:08.620 --> 00:02:13.460
And then you need to move money across borders efficiently. And finally,
31
00:02:13.600 --> 00:02:16.400
you need to manage tax and regulatory compliance.
32
00:02:17.800 --> 00:02:22.260
Let's start with localizing the buyer experience. These are table stakes.
33
00:02:22.480 --> 00:02:24.840
These are the basics of going into a new country,
34
00:02:25.320 --> 00:02:27.700
and everyone here knows that this is important.
35
00:02:29.980 --> 00:02:34.600
We start by talking about currencies and showing a local currency to a buyer is
36
00:02:34.640 --> 00:02:35.720
extremely important.
37
00:02:36.460 --> 00:02:41.400
We see that 76% of buyers choose to pay in their local currency when
38
00:02:41.480 --> 00:02:45.620
presented that option. Then think about payment methods.
39
00:02:46.620 --> 00:02:49.560
When we offer an additional payment method beyond cards,
40
00:02:49.980 --> 00:02:52.760
we see an increase in cross-border revenue of 12%.
41
00:02:55.520 --> 00:02:59.300
Now, this sounds super simple, but in practice, it rarely is.
42
00:03:00.080 --> 00:03:02.320
Exchange rates fluctuate constantly.
43
00:03:03.060 --> 00:03:07.680
Payment methods don't work with certain types of transactions and currencies,
44
00:03:08.120 --> 00:03:12.580
and your checkout has to determine in real time what to show each buyer.
45
00:03:13.940 --> 00:03:18.940
These choices can really impact whether your buyer feels like they're getting
46
00:03:19.080 --> 00:03:22.820
a local experience. Pricing is a great example.
47
00:03:24.140 --> 00:03:26.100
A simple currency conversion is never enough,
48
00:03:26.920 --> 00:03:29.720
and we have to think about how the number comes across to the buyer.
49
00:03:30.720 --> 00:03:34.820
Let's say you're pricing something in 99.99 US dollars,
50
00:03:35.580 --> 00:03:37.800
and a Japanese buyer comes to look at this.
51
00:03:38.620 --> 00:03:41.640
A straight conversion might make this look something like this,
52
00:03:44.780 --> 00:03:49.280
15,956 yen. Now, this is accurate, but in Japan,
53
00:03:49.440 --> 00:03:50.320
this is not the norm.
54
00:03:51.080 --> 00:03:55.940
Most prices often end in 80 or 90 yen below a major
55
00:03:56.020 --> 00:04:01.020
threshold like 100. So instead of feeling intentional,
56
00:04:01.280 --> 00:04:03.220
a straight conversion feels random.
57
00:04:05.020 --> 00:04:08.460
And the same principle applies to payment methods.
Now, here in the US,
58
00:04:08.760 --> 00:04:11.900
cards are extremely popular. But in Japan,
59
00:04:12.880 --> 00:04:17.760
PayPay is a popular payment method used by up to 72 million users.
60
00:04:18.800 --> 00:04:20.920
And if PayPay isn't supported in your checkout,
61
00:04:21.440 --> 00:04:25.840
you are creating friction at the very moment that your buyer is ready to pay.
62
00:04:26.400 --> 00:04:28.120
And that's often when you lose the sale.
63
00:04:30.100 --> 00:04:34.860
This is the challenge of selling globally. Pricing norms, payment methods,
64
00:04:35.240 --> 00:04:37.260
and buyer expectations all vary.
65
00:04:38.520 --> 00:04:43.520
Businesses that get this right can truly offer a localized experience leading to
66
00:04:43.740 --> 00:04:44.573
higher conversion.
67
00:04:46.440 --> 00:04:50.740
This used to be months of work going market by market by market. And today,
68
00:04:51.060 --> 00:04:54.880
it no longer takes a dedicated localization team to make this happen.
69
00:04:56.540 --> 00:05:01.520
Stripe's Optimized Checkout Suite can deliver an optimized, localized,
70
00:05:01.660 --> 00:05:04.520
frictionless experience to improve your conversion.
71
00:05:05.400 --> 00:05:06.860
This includes Adaptive Pricing,
72
00:05:07.680 --> 00:05:11.640
where you can present prices in over 150 different currencies.
73
00:05:12.720 --> 00:05:16.360
And we've seen that Adaptive Pricing increases cross-border revenue by
74
00:05:17.970 --> 00:05:22.960
17.8%. Stripe handles all of the FX and allows you to expand
75
00:05:23.100 --> 00:05:24.520
and create a localized currency.
76
00:05:27.180 --> 00:05:29.400
And to help prices feel more local,
77
00:05:29.480 --> 00:05:32.320
like I talked about with our Japanese example,
78
00:05:32.740 --> 00:05:36.800
we are going to introduce the ability to round prices automatically into a much
79
00:05:36.900 --> 00:05:41.640
more localized state. But as we discuss, this is only part of the experience.
80
00:05:43.380 --> 00:05:46.580
You need to offer payment methods that buyers want to pay with.
81
00:05:47.500 --> 00:05:51.980
The Optimized Checkout Suite also supports over 125 payment methods globally,
82
00:05:52.560 --> 00:05:57.280
and our AI models constantly optimize the exact right payment method to show
83
00:05:57.780 --> 00:05:59.600
to a buyer at the exact right time.
84
00:06:01.300 --> 00:06:03.460
So instead of starting from scratch with every market,
85
00:06:03.880 --> 00:06:08.820
you can now go live with a localized checkout experience globally in days,
86
00:06:08.900 --> 00:06:10.720
not months.
Now,
87
00:06:11.360 --> 00:06:14.260
money coming in should be a really good wins,
88
00:06:14.820 --> 00:06:18.980
but this is where complexity starts in terms of how you move money across
89
00:06:19.060 --> 00:06:20.540
borders. Let me show you why.
90
00:06:21.320 --> 00:06:26.060
Let's say that you're a company in the US selling to a buyer in Spain.
91
00:06:26.060 --> 00:06:28.200
You've localized your currency into the euro.
92
00:06:28.840 --> 00:06:31.400
You're also offering Bizum as a payment method,
93
00:06:32.640 --> 00:06:34.720
which is popular with buyers in Spain.
94
00:06:36.740 --> 00:06:41.060
Let's say you also have a bank account in the UK to manage your pounds.
95
00:06:42.300 --> 00:06:44.520
There's an operational reality that exists here
96
00:06:46.080 --> 00:06:50.060
when you have bank accounts across multiple countries. Revenue is coming in,
97
00:06:50.220 --> 00:06:53.780
but your treasury is fragmented across multiple currencies and rails.
98
00:06:56.080 --> 00:06:58.300
If you want to centralize and manage your funds in the US,
99
00:06:58.360 --> 00:07:02.740
you now have to pay foreign exchange to bring it back into your US bank account.
100
00:07:04.440 --> 00:07:05.273
Then comes payouts.
101
00:07:05.400 --> 00:07:09.740
Let's say that you have a supplier or a vendor in France or Germany,
102
00:07:10.200 --> 00:07:14.360
and now you need to pay foreign exchange again to get this money to them.
103
00:07:15.040 --> 00:07:19.300
And then you layer on settlement times and delays and cutoffs,
104
00:07:19.660 --> 00:07:23.520
and this whole process becomes expensive and delayed.
105
00:07:25.080 --> 00:07:29.040
So that's money lost on the way in and money lost on the way out.
106
00:07:30.680 --> 00:07:35.280
Instead, here's what money management can look like with Stripe. With Stripe,
107
00:07:35.660 --> 00:07:37.220
businesses can settle, hold,
108
00:07:37.420 --> 00:07:40.760
and pay out funds within a single system across multiple
109
00:07:40.800 --> 00:07:43.060
currencies.
110
00:07:43.060 --> 00:07:45.900
If you're collecting euros in Spain or pounds in the UK,
111
00:07:46.340 --> 00:07:50.360
you can manage it all in a single place while settling everything back into your
112
00:07:50.380 --> 00:07:51.420
domestic country.
113
00:07:54.420 --> 00:07:56.640
As you heard in the keynote today, this coverage is expanding.
114
00:07:57.240 --> 00:07:58.160
By the end of this year,
115
00:07:58.280 --> 00:08:02.120
Stripe Treasury will support storage in 15 different currencies,
116
00:08:02.780 --> 00:08:06.220
and you can keep more of your operation tied to the currency that you prefer.
117
00:08:08.380 --> 00:08:12.120
And from the same account, you can now move funds to vendors, customers,
118
00:08:12.480 --> 00:08:15.100
or any third party in more than 150 countries.
119
00:08:16.140 --> 00:08:20.520
Part of what makes all this possible is stablecoins, which are creating faster,
120
00:08:20.720 --> 00:08:23.460
more flexible ways to move money across borders.
121
00:08:25.960 --> 00:08:29.520
For a long time, global businesses had to work across fragmented systems,
122
00:08:30.100 --> 00:08:33.300
and the result was added cost, complexity, delays.
123
00:08:33.920 --> 00:08:37.540
And now you have access to a global infrastructure that allows you to operate
124
00:08:37.580 --> 00:08:38.413
like a local.
125
00:08:40.500 --> 00:08:44.480
Even when money is moving efficiently and you've localized your currency and
126
00:08:44.500 --> 00:08:46.280
you're offering the right payment methods,
127
00:08:46.980 --> 00:08:49.220
there's a third problem that quickly creeps up,
128
00:08:49.660 --> 00:08:51.360
which is tax and regulatory compliance.
129
00:08:52.320 --> 00:08:54.580
It is not the first priority for most businesses,
130
00:08:55.040 --> 00:08:58.400
but it very quickly adds to your to-do list as your business starts to grow.
131
00:08:59.920 --> 00:09:02.440
Let's say that you're on top of your sales tax in the US.
132
00:09:02.760 --> 00:09:07.000
You have gone and figured out these 16,000 combinations of
133
00:09:07.240 --> 00:09:11.560
jurisdictions and tax rules, and you have a very efficient system in place.
134
00:09:12.420 --> 00:09:14.040
But when you expand to India,
135
00:09:14.600 --> 00:09:18.840
India's GST is a completely different system altogether with its own set of
136
00:09:18.920 --> 00:09:21.480
rules and regulations. In some cases,
137
00:09:21.840 --> 00:09:26.080
e-invoicing requires you to report transactions to the government in real time.
138
00:09:27.720 --> 00:09:32.120
Disputes is another place where there are differences. In the US,
139
00:09:32.400 --> 00:09:34.360
you have 30 days to respond to a dispute,
140
00:09:34.680 --> 00:09:39.000
which is usually done through established chargeback processes. In India,
141
00:09:39.320 --> 00:09:42.320
you are required to respond to a dispute in 48 hours,
142
00:09:43.000 --> 00:09:45.280
and that doesn't just stop with taxes and disputes.
143
00:09:48.280 --> 00:09:50.840
Other differences include fraud patterns,
144
00:09:51.000 --> 00:09:53.760
which can vary across different types of markets,
145
00:09:54.400 --> 00:09:58.280
and that compliance just becomes a limiting factor in your global expansion.
146
00:09:58.840 --> 00:10:02.760
The rules don't just multiply, they vary dramatically from market to market.
147
00:10:03.160 --> 00:10:08.080
So how do you stay on top of all this? There are two ways to tackle compliance,
148
00:10:08.400 --> 00:10:12.040
and it really sort of comes down to a very critical decision that you need to
149
00:10:12.120 --> 00:10:15.600
make, which is, who is the merchant of record in my transaction?
150
00:10:17.480 --> 00:10:21.000
That usually is the entity that's responsible for tax, disputes,
151
00:10:21.440 --> 00:10:24.400
and refunds and customer support in a particular market.
152
00:10:25.280 --> 00:10:27.880
And that choice does not have to look the same everywhere.
153
00:10:28.800 --> 00:10:30.920
More businesses are taking a blended approach.
154
00:10:31.320 --> 00:10:34.240
In large markets where you have a significant presence,
155
00:10:34.680 --> 00:10:36.880
businesses choose to be their own merchant of record.
156
00:10:37.240 --> 00:10:39.920
And in emerging markets where you're potentially taking a bet,
157
00:10:40.320 --> 00:10:45.280
business can offload that merchant of record responsibility to someone else.
If
158
00:10:45.360 --> 00:10:49.120
you want to say the merchant of record, tax is often the first place to start.
159
00:10:50.560 --> 00:10:55.480
Stripe Tax can help monitor your obligations across all
160
00:10:55.710 --> 00:10:59.360
geos, whether it's sales tax in the US or GST in India.
161
00:11:00.000 --> 00:11:04.920
More than 67,000 businesses use Stripe Tax today to automate their collection,
162
00:11:05.320 --> 00:11:06.040
registration,
163
00:11:06.040 --> 00:11:10.600
and filing for growing businesses that could remove very significant complexity.
164
00:11:12.080 --> 00:11:16.280
Retell AI use Stripe Tax to automate compliance across more than a hundred
165
00:11:16.360 --> 00:11:21.080
countries, but tax, like I said, is only part of the problem.
166
00:11:22.480 --> 00:11:25.180
We talked about how disputes vary, how fraud can vary,
167
00:11:25.760 --> 00:11:27.460
and if you don't want to take that on yourself,
168
00:11:28.520 --> 00:11:31.900
Stripe Managed Payments offers a different model. Here,
169
00:11:32.260 --> 00:11:35.600
Stripe can become your merchant of record for eligible digital goods
170
00:11:35.640 --> 00:11:39.660
transactions. We can handle tax compliance, fraud prevention,
171
00:11:39.820 --> 00:11:42.160
dispute management, as well as customer support.
172
00:11:43.680 --> 00:11:47.220
So instead of learning regulations on a country-by-country basis,
173
00:11:47.680 --> 00:11:51.180
you can now offload all of this to a different merchant of record.
174
00:11:53.960 --> 00:11:55.180
And we see this in our data.
175
00:11:56.340 --> 00:12:00.680
77% of merchant record transactions today involve a
176
00:12:01.160 --> 00:12:05.720
service and percent of merchants using Stripe Managed Payments today have a
177
00:12:05.800 --> 00:12:06.700
cross-border transaction.
178
00:12:07.580 --> 00:12:12.080
And most merchants who start with Stripe Managed Payments accept a cross-border
179
00:12:12.120 --> 00:12:16.460
payment on their very first day. Compliance is on everybody's to-do list,
180
00:12:16.860 --> 00:12:20.100
and the reality is that with every market, the burden grows.
181
00:12:20.320 --> 00:12:24.820
And the good news is that businesses now have many tools to help solve this for
182
00:12:24.840 --> 00:12:25.673
them.
183
00:12:26.540 --> 00:12:31.140
So whether you own compliance with [Stripe] tax or offload it to a merchant of
184
00:12:31.200 --> 00:12:32.720
record like Managed Payments,
185
00:12:33.300 --> 00:12:37.940
both approaches help to reduce your operational load. Now,
186
00:12:38.300 --> 00:12:40.660
the question used to be when to go global.
187
00:12:41.960 --> 00:12:43.760
Now it has quickly turned into,
188
00:12:44.160 --> 00:12:47.020
how do I quickly turn global demand into revenue?
189
00:12:47.380 --> 00:12:51.840
And that comes down to three things, localizing the experience,
190
00:12:52.720 --> 00:12:56.540
moving money efficiently, and managing your tax and regulatory burdens.
191
00:12:57.540 --> 00:13:01.120
These used to be separate problems with their own solutions and teams that you
192
00:13:01.220 --> 00:13:04.480
used to have to put together. And now with Stripe,
193
00:13:05.460 --> 00:13:08.580
this infrastructure comes together into a single operating system.
194
00:13:10.380 --> 00:13:13.240
The companies you're about to hear from now have been through this very journey
195
00:13:13.280 --> 00:13:17.260
themselves. So please, welcome to the stage, Mark, Raman, Peter,
196
00:13:17.360 --> 00:13:18.193
and my colleague, Conor.
197
00:13:21.380 --> 00:13:24.500
Thank you, Abhi. Good afternoon, everyone. My name's Conor.
198
00:13:24.560 --> 00:13:26.980
I lead the EMEA business at Stripe.
199
00:13:27.060 --> 00:13:29.700
Delighted to be here with you today to talk a bit more about global.
200
00:13:30.380 --> 00:13:34.040
Hopefully you've gotten it from the entire messaging over the course of the last
201
00:13:34.120 --> 00:13:34.740
two days,
202
00:13:34.740 --> 00:13:39.320
but going global by default is a key cornerstone of our messaging at Sessions
203
00:13:39.440 --> 00:13:40.900
this year. And of course,
204
00:13:40.980 --> 00:13:45.540
we want to turn this now from a conceptual into real-world talk by hearing from
205
00:13:45.620 --> 00:13:48.740
leaders who are doing this every day in their day-to-day lives.
206
00:13:48.820 --> 00:13:51.540
So thank you guys for joining us. Mark, I might start with you.
207
00:13:51.960 --> 00:13:55.900
Maybe just give a bit of an insight in where you are today at OpenAI in your
208
00:13:56.080 --> 00:13:57.260
globalization journey.
209
00:13:58.360 --> 00:14:03.240
Yeah. I think today we are effectively a very global company.
210
00:14:03.440 --> 00:14:04.620
I think from day one,
211
00:14:05.220 --> 00:14:09.120
our stance was that to be true to our mission,
212
00:14:09.200 --> 00:14:13.340
we want to provide intelligence to effectively anyone in the world.
213
00:14:13.900 --> 00:14:17.640
And so we've been on the journey to take products basically from zero to one,
214
00:14:18.160 --> 00:14:19.180
starting with free tier,
215
00:14:19.460 --> 00:14:23.800
then moving into a paid tier and effectively being everywhere we can all at
216
00:14:23.880 --> 00:14:26.300
once. Yeah, that's where we are today.
217
00:14:26.840 --> 00:14:28.300
Raman, great. Thank you, Mark. Raman,
218
00:14:28.380 --> 00:14:29.980
where are you from a Perplexity standpoint?
219
00:14:30.980 --> 00:14:35.980
We are as global as you can get. We, what, about three years ago,
220
00:14:36.060 --> 00:14:40.880
Perplexity launched with essentially a tweet merging search and LLMs and
221
00:14:41.160 --> 00:14:44.020
a little bit of a prayer. And luckily it took off globally.
222
00:14:44.160 --> 00:14:47.960
So the initial launch, the product adoption had no borders to it.
223
00:14:48.680 --> 00:14:49.580
On the monetization side,
224
00:14:49.700 --> 00:14:54.400
we had to move very fast to meet that adoption across every single country.
225
00:14:54.470 --> 00:14:58.770
And so we are now fully rolled out with Stripe across almost every single
226
00:14:58.790 --> 00:15:03.010
country where we are served and localized with payment methods.
227
00:15:03.270 --> 00:15:06.450
Yeah. Awesome. Awesome. Peter, Xero,
228
00:15:06.550 --> 00:15:11.090
obviously slightly different age of company than the two here on the left-hand
229
00:15:11.130 --> 00:15:13.000
side, but you guys are around since 2006.
230
00:15:13.770 --> 00:15:18.730
Maybe tell the story about what was going global like back in 2006 for Xero
231
00:15:19.150 --> 00:15:21.530
versus maybe where you are today on your journey.
232
00:15:21.810 --> 00:15:26.510
Yeah. Yeah. So we're celebrating our 20th anniversary this year.
233
00:15:26.570 --> 00:15:29.730
And if you go back to 2000 ... Thank you. Yeah.
234
00:15:30.170 --> 00:15:33.610
If you go back to 2006, I wasn't at the company at the time,
235
00:15:33.670 --> 00:15:36.210
but it was a very different world back then.
236
00:15:36.690 --> 00:15:40.870
Cloud was still a novelty and particularly accounting and business software in
237
00:15:40.890 --> 00:15:42.910
the cloud, that was definitely an oldie.
238
00:15:43.510 --> 00:15:48.310
And so launched in New Zealand, quickly saw demand in Australia.
239
00:15:48.450 --> 00:15:53.360
So moved to Australia in 2008 through 2010 to 2015,
240
00:15:53.510 --> 00:15:55.150
scaled out to the US and the UK.
241
00:15:55.650 --> 00:15:57.490
And then if you skip forward where we are today,
242
00:15:58.490 --> 00:16:02.550
we have offices all around the world. We're in the US, the UK, Australia,
243
00:16:03.250 --> 00:16:05.910
Canada, Hong Kong, Singapore, South Africa.
244
00:16:06.830 --> 00:16:11.210
We're serving five million small businesses and operating in 180 countries
245
00:16:11.430 --> 00:16:15.510
around the world. And a lot of that was driven by the demand.
246
00:16:16.550 --> 00:16:18.270
There wasn't a lot of providers, as I said,
247
00:16:18.730 --> 00:16:20.750
at that time that we're doing what we're doing,
248
00:16:20.810 --> 00:16:24.990
and there's still only a few players.
And we have a really strong distribution
249
00:16:25.070 --> 00:16:27.110
network that drives that global flywheel.
250
00:16:27.630 --> 00:16:30.630
We have local advisors in each of our markets,
251
00:16:30.690 --> 00:16:33.870
which is accountants and bookkeepers, and they're like the heart of it.
252
00:16:33.930 --> 00:16:36.230
They recommend Xero to their customers.
253
00:16:36.770 --> 00:16:40.430
Then we have distribution partners with Stripe and Monzo,
254
00:16:40.630 --> 00:16:42.330
so Stripe with the Atlas program.
255
00:16:42.510 --> 00:16:45.350
So you get payments and accounting with Monzo in the UK,
256
00:16:45.510 --> 00:16:47.090
you get banking and accounting.
257
00:16:47.930 --> 00:16:52.490
And all of that works together to continue to grow and drive our flywheel.
258
00:16:52.970 --> 00:16:57.610
Awesome. Awesome. Thank you very much. Raman, maybe go to you next.
259
00:16:57.670 --> 00:17:02.170
But we heard a lot this morning in the keynote as well and John's indexing the
260
00:17:02.230 --> 00:17:07.070
economy session about the long tail of globalization and that it's
261
00:17:07.150 --> 00:17:10.720
not enough just to go into maybe five of big global markets that there's such a
262
00:17:11.210 --> 00:17:14.850
long tail. So maybe one, is that true from a Perplexity standpoint?
263
00:17:14.910 --> 00:17:15.830
And then second of all,
264
00:17:16.210 --> 00:17:18.770
when you were looking at maybe some of the markets that you were looking to
265
00:17:18.830 --> 00:17:19.663
expand into,
266
00:17:19.870 --> 00:17:24.220
what surprised you or maybe created surface challenges that you didn't really
267
00:17:24.290 --> 00:17:25.830
expect from the outset?
268
00:17:26.410 --> 00:17:28.310
Yeah. I mean, I think there are a couple layers here.
269
00:17:28.510 --> 00:17:30.710
One is just on the competition side. Obviously,
270
00:17:31.210 --> 00:17:35.430
this is the fastest adopted technology in mankind history,
271
00:17:35.510 --> 00:17:39.770
and we have to move fast to get this as quickly as possible out to every single
272
00:17:39.790 --> 00:17:41.110
customer regardless of country.
273
00:17:42.480 --> 00:17:46.930
One of the challenges early on was part of our distribution strategy was landing
274
00:17:47.050 --> 00:17:50.290
in a global country with a partner to help us distribute.
275
00:17:50.350 --> 00:17:52.210
So this could be like a Telco, for example,
276
00:17:52.850 --> 00:17:54.430
and a Telco would distribute out Perplexity.
277
00:17:55.030 --> 00:17:58.490
The problem was is that there's a lot of friction in the actual checkout flow if
278
00:17:59.090 --> 00:18:01.750
it's not localized, if it doesn't have the right payment methods.
279
00:18:02.150 --> 00:18:04.850
So not only were we getting inbound levels of support tickets,
280
00:18:05.410 --> 00:18:08.770
but we also were getting friction on the partnership side of, "Hey,
281
00:18:09.150 --> 00:18:12.210
we want the local currency here. This is what our customers actually [expect]."
282
00:18:13.790 --> 00:18:16.690
And so that's where we actually spent so much time working with you guys,
283
00:18:16.750 --> 00:18:19.910
of course, on actually how do we localize these checkout pages?
284
00:18:19.910 --> 00:18:21.790
How do we localize the actual end-to-end experience,
285
00:18:21.870 --> 00:18:25.850
which just feels like a very straightforward flow to checkout.
286
00:18:25.910 --> 00:18:28.950
And then also it enables all these partnerships to actually come to life now
287
00:18:29.030 --> 00:18:29.863
faster.
288
00:18:30.030 --> 00:18:33.970
Was there any specific geography that maybe gave you challenges or gave you
289
00:18:34.050 --> 00:18:37.090
unexpected hurdles that you hadn't foreseen at the outset?
290
00:18:37.170 --> 00:18:38.003
I'm sure there was a few.
291
00:18:38.850 --> 00:18:39.890
Yeah, of course. I mean,
292
00:18:40.410 --> 00:18:45.390
the one that quickly comes to mind is geographies where the actual
293
00:18:45.470 --> 00:18:48.710
payment method, there's a clear preference outside of credit cards.
294
00:18:49.410 --> 00:18:50.650
Out of the box, we had credit cards,
295
00:18:50.730 --> 00:18:54.710
but it was tough for us to convince a partner to launch with us if we weren't
296
00:18:54.750 --> 00:18:56.070
serving the most common payment.
297
00:18:56.190 --> 00:18:57.230
We're talking like UPI-.
298
00:18:57.710 --> 00:18:58.970
Exactly. Pix in Brazil,
299
00:18:59.050 --> 00:19:03.050
I think is one that immediately comes to mind where as soon as we turned on Pix,
300
00:19:03.110 --> 00:19:05.170
you saw all the adoption start to flow right through it.
301
00:19:05.230 --> 00:19:10.230
Yeah, yeah, yeah. And Mark, for your perspective, maybe similar question,
302
00:19:10.290 --> 00:19:14.590
maybe talk about that. I know obviously mission, like when we spoke beforehand,
303
00:19:14.670 --> 00:19:19.110
you often anchored back to mission and the importance of having monetization
304
00:19:19.170 --> 00:19:21.230
aligned to the mission of OpenAI.
305
00:19:21.770 --> 00:19:26.130
Talk a bit about that and the importance of getting this technology into the
306
00:19:26.170 --> 00:19:29.810
hands of everyone and how that links into building a monetization strategy that
307
00:19:29.850 --> 00:19:31.090
enables that mission.
308
00:19:31.410 --> 00:19:32.230
Yeah, for sure.
309
00:19:32.230 --> 00:19:36.640
So we launched ChatGPT November 30 of 2022.
310
00:19:38.030 --> 00:19:41.320
At the end of the year, at the end of 2022,
311
00:19:41.870 --> 00:19:45.990
we were already at a hundred million MAUs. And at that point,
312
00:19:46.130 --> 00:19:50.390
we were really putting stress on the systems, primarily our compute,
313
00:19:50.810 --> 00:19:52.570
but our infrastructure in general.
314
00:19:53.190 --> 00:19:58.130
And it became a worse experience for some of our most engaged
315
00:19:58.210 --> 00:20:00.990
users. They were getting slower responses.
316
00:20:01.150 --> 00:20:03.510
The infrastructure was just not as reliable.
317
00:20:03.950 --> 00:20:08.510
So it effectively became incumbent on us to figure out how we could
318
00:20:08.630 --> 00:20:11.890
serve our users and especially our most engaged users.
319
00:20:12.570 --> 00:20:17.250
And having a paid model became a way for us to really segment that population
320
00:20:18.010 --> 00:20:21.490
in a place that they could then pay for getting what they wanted,
321
00:20:21.930 --> 00:20:23.690
but also as a result of that,
322
00:20:23.930 --> 00:20:28.890
for us to capture revenue that we can then use to subsidize the usage of our
323
00:20:28.910 --> 00:20:33.270
product across all of our free user base.
And so that's the playbook that we
324
00:20:33.330 --> 00:20:35.090
have taken. We've really expanded.
325
00:20:35.330 --> 00:20:40.150
And what we've faced is that it's really pushed us in terms of this
326
00:20:40.710 --> 00:20:42.890
mission of providing intelligence to everyone,
327
00:20:42.950 --> 00:20:46.770
but the tension of compute costs being very,
328
00:20:46.830 --> 00:20:50.870
very material on a marginal basis for us to think about what is the
329
00:20:51.910 --> 00:20:56.570
SKU structure that allows for us to achieve that mission and has forced us
330
00:20:56.630 --> 00:20:58.930
really to become much more complex, I think, on the SKU side.
331
00:20:58.990 --> 00:21:01.080
So we have a plus SKU at $20.
332
00:21:01.850 --> 00:21:05.320
We launched a pro SKU at the end of 2024 at $200,
333
00:21:06.330 --> 00:21:09.230
and we launched a go SKU, which is in the $5 range.
334
00:21:09.590 --> 00:21:12.650
We two weeks ago launched a $100 pro SKU.
335
00:21:13.310 --> 00:21:16.610
So these are the things that we need to think about when we're thinking about
336
00:21:16.650 --> 00:21:21.310
expanding globally is how do we match the SKU structure to the demand that
337
00:21:21.370 --> 00:21:25.430
exists in that market and do so in a way that people actually have the ability
338
00:21:25.490 --> 00:21:26.610
to pay and willingness to pay.
339
00:21:27.010 --> 00:21:29.290
Yeah. And when you look, maybe a follow-up question,
340
00:21:29.350 --> 00:21:31.390
when you look at the path forward,
341
00:21:31.510 --> 00:21:36.110
and obviously we had Sam here yesterday with Patrick as part of the
342
00:21:36.190 --> 00:21:37.450
afternoon session,
343
00:21:38.510 --> 00:21:43.050
what are the type of challenges you think you're going to face on a monetization
344
00:21:43.110 --> 00:21:46.850
standpoint as you continue to see the next evolution of OpenAI?
345
00:21:48.230 --> 00:21:50.470
Yeah, I think it's really getting very,
346
00:21:50.550 --> 00:21:55.470
very good at understanding what are the local nuances of a market
347
00:21:55.610 --> 00:22:00.590
that mandate us positioning our pricing and localization in a unique
348
00:22:00.670 --> 00:22:04.630
way to face those customers. For example,
349
00:22:05.330 --> 00:22:09.850
once you have a payment method live, that's really the beginning of the work.
350
00:22:10.570 --> 00:22:13.230
There's a lot of local nuance that you have to appreciate.
351
00:22:13.690 --> 00:22:17.710
How is this payment going to recur? Can you support retrials for this payment?
352
00:22:18.150 --> 00:22:21.450
Should you have a dunning experience if you're on a renewing subscription?
353
00:22:21.890 --> 00:22:25.370
Should that dunning length be the same as any other country?
354
00:22:26.330 --> 00:22:30.330
If the payment method requires you to do a step up for the user to sign a
355
00:22:30.390 --> 00:22:33.050
mandate, what should that mandate amount be?
356
00:22:33.770 --> 00:22:37.170
How is that mandate affecting the UX of a recurring payment?
357
00:22:37.570 --> 00:22:40.050
Should you have a higher mandate to allow for upgrade paths?
358
00:22:40.410 --> 00:22:41.243
There's just a ton...
359
00:22:42.010 --> 00:22:44.130
Not at all complex then. Not at all complex.
360
00:22:44.230 --> 00:22:46.930
There's a ton of nuance, and I think it's on us to just get very,
361
00:22:47.010 --> 00:22:49.650
very good at understanding that. To be quite frank,
362
00:22:50.210 --> 00:22:52.170
AI has really helped us do that.
363
00:22:53.230 --> 00:22:55.530
This has been a solved problem in many cases,
364
00:22:55.590 --> 00:22:59.930
and so getting the value of the knowledge that has been bestowed on other
365
00:22:59.990 --> 00:23:03.030
companies that have gone through this journey and for us to plug into that and
366
00:23:03.070 --> 00:23:06.330
really quickly understand what we need to do to scale is how we approach this.
367
00:23:06.610 --> 00:23:10.350
Awesome. Thank you for those insights. Peter, maybe turning to you,
368
00:23:11.130 --> 00:23:15.630
Abhi talked a bit about the tax compliance, regulation, et cetera.
369
00:23:15.770 --> 00:23:16.870
So I mean,
370
00:23:17.050 --> 00:23:20.230
one aspect of it is collecting the money and having a localized experience,
371
00:23:20.290 --> 00:23:24.530
but then the back office complexity that maybe has come with a truly global
372
00:23:24.930 --> 00:23:25.430
offering.
373
00:23:25.430 --> 00:23:29.270
Can you maybe talk about that through the lens of Xero or maybe your prior
374
00:23:29.330 --> 00:23:32.010
experience in terms of what you've seen there and some of the challenges you've
375
00:23:32.030 --> 00:23:35.930
encountered and any advice you have for the audience in terms of how to untangle
376
00:23:35.970 --> 00:23:37.090
that complex web.
377
00:23:37.990 --> 00:23:40.410
Yeah. So prior to Xero,
378
00:23:40.830 --> 00:23:44.350
I was leading product for global expansion for Cash App.
379
00:23:44.930 --> 00:23:47.450
And so Cash App's obviously huge in the US.
380
00:23:47.590 --> 00:23:52.130
We had a growing product in the UK and we're aiming to launch in,
381
00:23:52.430 --> 00:23:57.090
at the time, launch in three new markets within 12 months of one another.
382
00:23:58.950 --> 00:24:03.430
And the cost to build is obviously getting
383
00:24:03.490 --> 00:24:04.190
compressed.
384
00:24:04.190 --> 00:24:08.930
And product like eng capacity was a challenge to localize the
385
00:24:08.970 --> 00:24:11.690
product, especially a product as dense as Cash App.
386
00:24:12.250 --> 00:24:16.950
But if you look at what was outside of our control, there was two big things.
387
00:24:17.010 --> 00:24:17.590
So one,
388
00:24:17.590 --> 00:24:21.010
which was like Abhi was talking about setting up local banking relationships,
389
00:24:21.750 --> 00:24:26.090
and the second was engaging with regulators and getting licenses in each region
390
00:24:26.170 --> 00:24:29.210
as well. And so when you think about banks and regulators,
391
00:24:29.310 --> 00:24:33.570
they are not moving at the same pace as a fintech or like an AI company today.
392
00:24:33.690 --> 00:24:36.150
So each of those, they have a lead time.
393
00:24:36.730 --> 00:24:40.650
So you do not underestimate the lead time to get your local bank relationships
394
00:24:40.710 --> 00:24:44.990
set up or your licenses as well. Because you start to engage with regulators,
395
00:24:45.070 --> 00:24:47.410
as you start to engage with those bank partners,
396
00:24:47.830 --> 00:24:49.790
they're going to want to know who are your customers,
397
00:24:49.850 --> 00:24:51.390
how are they using your product,
398
00:24:51.490 --> 00:24:54.790
how are you keeping bad actors off your product? What are your AML controls?
399
00:24:54.850 --> 00:24:57.250
All of those things, they take time.
400
00:24:57.490 --> 00:25:00.490
And then as you start to scale out and go live,
401
00:25:01.030 --> 00:25:04.910
as you're holding funds as well, how do you manage cross-border flows?
402
00:25:05.270 --> 00:25:07.350
Who's managing FX hedging?
403
00:25:07.510 --> 00:25:11.190
All of that stuff starts to grow with complexity as the businesses scale.
404
00:25:11.950 --> 00:25:15.570
Yeah. Yeah. No, super interesting. Okay.
405
00:25:16.350 --> 00:25:19.570
And we're almost up on time here. So maybe just Raman,
406
00:25:19.630 --> 00:25:23.170
we'll go to you and the two-sided maybe question to close.
407
00:25:23.230 --> 00:25:28.210
So if you had to go back in time and give yourself advice a number of
408
00:25:28.310 --> 00:25:32.630
years ago on what you would've done differently as you embarked on this journey,
409
00:25:32.810 --> 00:25:35.750
what would that be? And then the second part to the question is,
410
00:25:36.230 --> 00:25:39.010
as you look over the next 12 months, I know we talked about this as well,
411
00:25:39.150 --> 00:25:43.390
what are your top priorities from a globalization standpoint as you look to the
412
00:25:43.410 --> 00:25:44.150
future?
413
00:25:44.150 --> 00:25:48.890
Yeah. Oh man, if I could go back. I mean, I think the first piece is,
414
00:25:49.170 --> 00:25:51.930
can we get that monetization platform up faster,
415
00:25:52.950 --> 00:25:56.410
and then can we actually globalize faster on the localization front?
416
00:25:57.830 --> 00:26:02.150
When we look at our checkout flow, there's a beautiful kink in the graph.
417
00:26:02.330 --> 00:26:04.550
You can tell when we launched localization,
418
00:26:04.790 --> 00:26:06.670
the earlier we could have pulled it up, the much better,
419
00:26:07.270 --> 00:26:09.910
especially just given that there's a lot of-.
420
00:26:10.150 --> 00:26:14.590
Was that because you didn't fully understand the value that localization
421
00:26:15.010 --> 00:26:19.050
would present it or was it just that you weren't aware of the technology that
422
00:26:19.070 --> 00:26:21.970
was there to enable localization sooner or was it a combination?
423
00:26:22.190 --> 00:26:25.550
I mean, I think it was also just a team at some point of like,
424
00:26:25.830 --> 00:26:28.970
do we have the right resourcing here to invest in this,
425
00:26:29.030 --> 00:26:32.410
especially when we knew that certain products were on the roadmap on the Stripe
426
00:26:32.450 --> 00:26:36.930
side, which is candidly, all of us up here,
427
00:26:37.950 --> 00:26:42.130
team-constrained in terms of what we can do and what's the next big initiative
428
00:26:42.210 --> 00:26:44.310
to work on. And so that-.
429
00:26:44.410 --> 00:26:45.243
Prioritization.
430
00:26:45.550 --> 00:26:46.930
It's the big piece of it. Yeah.
431
00:26:47.710 --> 00:26:51.550
So I think moving that up to be as early as possible is key.
432
00:26:52.530 --> 00:26:54.090
And then on the localization front more broadly,
433
00:26:55.170 --> 00:26:59.730
we are essentially now starting to really build more of a footprint in specific
434
00:26:59.770 --> 00:27:04.070
countries, build more partnerships on the B2B front. And so as we look ahead,
435
00:27:04.130 --> 00:27:07.910
it's now having enterprise catch up to where consumer is in terms of
436
00:27:07.950 --> 00:27:10.410
localization, in terms of a footprint globally,
437
00:27:10.470 --> 00:27:14.430
and starting to build those inroads in terms of how do we manage seats,
438
00:27:14.490 --> 00:27:15.750
how do we manage checkout flows?
439
00:27:15.810 --> 00:27:17.870
How do we manage usage-based billing all globally?
440
00:27:18.310 --> 00:27:23.010
And is the work done on the consumer front highly replicable onto the enterprise
441
00:27:23.030 --> 00:27:24.510
side? There's a lot of...
442
00:27:24.690 --> 00:27:24.990
It is. I mean,
443
00:27:24.990 --> 00:27:28.590
there are a lot of learnings you have to transfer over and then it's a little
444
00:27:28.610 --> 00:27:31.030
bit of a different system. But yeah, I mean,
445
00:27:31.090 --> 00:27:34.630
a lot of it is our roadmap's pretty clear because we've now done it on consumer
446
00:27:34.710 --> 00:27:35.450
already.
447
00:27:35.450 --> 00:27:38.030
Yeah. Okay. Makes sense. Mark, same question, same two set of question to you.
448
00:27:38.240 --> 00:27:41.450
Yeah. I think I echo Raman's sentiment getting
449
00:27:43.150 --> 00:27:45.650
localized earlier. We launched our go SKU,
450
00:27:45.730 --> 00:27:48.200
which is our $8 SKU in 2025,
451
00:27:48.750 --> 00:27:53.190
which is three years after we launched our paid-plus SKU and reflecting,
452
00:27:54.530 --> 00:27:56.150
I think we should have been a lot quicker.
453
00:27:57.650 --> 00:28:01.510
I think the reality was that we were really focused on expansion and getting
454
00:28:02.410 --> 00:28:05.270
users to really understand the value of this technology,
455
00:28:05.450 --> 00:28:08.830
especially in our free tier. So it wasn't as important, but if we did it again,
456
00:28:08.890 --> 00:28:13.010
it would be quicker.
I think the thing that I'm excited about in the future is
457
00:28:13.090 --> 00:28:17.150
how do we think about fully automating the process to
458
00:28:17.750 --> 00:28:18.590
expand globally?
459
00:28:19.750 --> 00:28:23.430
Just to talk our book on OpenAI with Codex,
460
00:28:24.490 --> 00:28:29.490
you really can automate the end-to-end production release of new features in
461
00:28:29.530 --> 00:28:33.810
a way that is very, very powerful. So I think the push to our teams are,
462
00:28:34.290 --> 00:28:38.970
how can you just sort of ask Codex to reference our Figma
463
00:28:39.350 --> 00:28:41.350
mock of our localized checkout page,
464
00:28:41.830 --> 00:28:44.090
say that we want to optimize it in this market,
465
00:28:44.550 --> 00:28:48.410
have it then use the connector on Stripe to then turn on payment method,
466
00:28:49.030 --> 00:28:53.410
then launch that into a layer in Statsig, have the experiment run,
467
00:28:54.210 --> 00:28:55.610
Codex review the experiment,
468
00:28:55.770 --> 00:28:58.810
show where it is optimized or not optimized and then just ship it.
469
00:28:59.250 --> 00:29:01.870
And I think we can do that in a very, very automated way.
470
00:29:02.370 --> 00:29:05.870
I think the things that prevented us from doing that as a sort of industry
471
00:29:05.950 --> 00:29:10.810
before is that you needed to have local integrations with local rails and
472
00:29:10.970 --> 00:29:11.803
payment methods,
473
00:29:12.170 --> 00:29:15.390
but now that's largely solved by companies at Stripe where you just have one
474
00:29:15.490 --> 00:29:18.170
central plugin point and you get access to all these.
475
00:29:18.170 --> 00:29:21.530
So I really do think we have all the ingredients together to get this fully
476
00:29:21.630 --> 00:29:25.450
automated and experiments running and show where it's incremental and then just
477
00:29:25.470 --> 00:29:27.470
shipping it and having the best experience for it.
478
00:29:27.530 --> 00:29:29.490
And a time horizon for that is, is that-.
479
00:29:30.030 --> 00:29:33.010
I think we can... If folks on my team are here,
480
00:29:33.070 --> 00:29:34.450
we should start doing this this week.
481
00:29:35.010 --> 00:29:35.843
Nice.
482
00:29:36.330 --> 00:29:39.930
But I think it's possible now. I think we just need to put our minds to it.
483
00:29:40.030 --> 00:29:44.850
Wow, that's exciting, exciting. Peter, your perspective, so again, looking back,
484
00:29:45.450 --> 00:29:48.710
what's the key learnings and what's priorities for Xero into the future?
485
00:29:49.010 --> 00:29:51.490
Yeah, exactly as Raman was saying.
486
00:29:51.550 --> 00:29:55.710
So key learning would just be get your prioritization framework now right from
487
00:29:55.730 --> 00:29:59.450
the start because there's going to be people that are shared across a bunch of
488
00:29:59.470 --> 00:30:02.710
different things, so legal compliance, data science,
489
00:30:02.930 --> 00:30:05.770
thinking about your existing markets and new markets as well.
490
00:30:05.830 --> 00:30:09.570
So how do you prioritize that capacity alongside all of your product and eng?
491
00:30:10.290 --> 00:30:13.790
And then what we're super excited about is like everyone,
492
00:30:13.870 --> 00:30:18.570
embedding more and more AI deeper into our customers' workflows to give them
493
00:30:18.650 --> 00:30:22.690
time back. And so that power of accounting plus payments plus AI,
494
00:30:22.850 --> 00:30:25.230
there's a massive opportunity that we're undertaking at the moment.
495
00:30:25.760 --> 00:30:30.060
Yeah. Awesome. Awesome. Well, thank you guys. Thank you, Peter, Raman, Mark.
496
00:30:30.180 --> 00:30:33.060
Thank you so much for your time. Thank you everyone.
497
00:30:33.120 --> 00:30:37.480
Hopefully this give you further insight into the criticality of global.
498
00:30:37.580 --> 00:30:42.220
And remember, don't wait. You can't wait to go global. It's a day one thing.
499
00:30:42.340 --> 00:30:45.520
So thank you so much everybody. Enjoy the rest of your Sessions experience.