WEBVTT

00:00:00.000 --> 00:00:02.836
I do think industry ISVs
are so well positioned

00:00:02.836 --> 00:00:07.008
to really deliver true
value with agentic AI.

00:00:07.008 --> 00:00:09.761
Mary, John, there
are so many companies who

00:00:09.761 --> 00:00:13.975
test these agentic AI
systems and solutions

00:00:13.975 --> 00:00:16.728
but fewer who actually implement them

00:00:16.728 --> 00:00:18.563
in their business successfully.

00:00:18.563 --> 00:00:20.524
What separates those who are doing it

00:00:20.524 --> 00:00:22.234
and having measurable results

00:00:22.234 --> 00:00:24.696
from those who are just not even trying?

00:00:24.696 --> 00:00:28.784
Sure. Well, at AWS I
work with industry ISVs.

00:00:28.784 --> 00:00:31.495
I do think industry ISVs
are so well positioned

00:00:31.495 --> 00:00:35.417
to really deliver true
value with agentic AI.

00:00:35.417 --> 00:00:38.379
These are ISVs who have longstanding,

00:00:38.379 --> 00:00:40.840
deep customer relationships
that are sticky.

00:00:40.840 --> 00:00:43.385
They, you know, have proprietary data

00:00:43.385 --> 00:00:45.554
often collected over decades.

00:00:45.554 --> 00:00:46.931
And most importantly,

00:00:46.931 --> 00:00:48.683
they really deeply understand

00:00:48.683 --> 00:00:50.810
these line of business workloads

00:00:50.810 --> 00:00:53.981
and the regulatory environments
in which they operate.

00:00:53.981 --> 00:00:57.151
So I think this, you know, customer trust

00:00:57.151 --> 00:01:00.447
and credibility combined
with domain expertise

00:01:00.447 --> 00:01:03.575
really set them up to leverage agentic AI

00:01:03.575 --> 00:01:05.620
to deliver true business value.

00:01:05.620 --> 00:01:07.038
And we're ironically seeing this

00:01:07.038 --> 00:01:09.332
in some of our most regulated industries

00:01:09.332 --> 00:01:11.502
like financial service and health care,

00:01:11.502 --> 00:01:14.380
where adoption is actually
moving the fastest

00:01:14.380 --> 00:01:16.341
In the most regulated industries.

00:01:16.341 --> 00:01:17.509
Yes.

00:01:17.509 --> 00:01:18.468
John?

00:01:18.468 --> 00:01:23.474
What we've also seen is more
broadly across industries,

00:01:24.684 --> 00:01:28.313
agentic technologies that
drive value, drive value with

00:01:28.313 --> 00:01:29.732
both the end user

00:01:29.732 --> 00:01:31.818
and the overall enterprise.

00:01:31.818 --> 00:01:35.030
There are so many agentic technologies

00:01:35.030 --> 00:01:38.534
that are almost moments
of user delight, right?

00:01:38.534 --> 00:01:40.328
Where users say like, Wow,
this is really intuitive.

00:01:40.328 --> 00:01:41.746
This is really easy to use.

00:01:41.746 --> 00:01:44.291
And that's been a big part
of their broad adoption, of

00:01:44.291 --> 00:01:45.584
the fast adoption.

00:01:46.877 --> 00:01:49.297
Enterprises have big goals as well.

00:01:49.297 --> 00:01:50.548
So they need to get value too.

00:01:50.548 --> 00:01:52.092
So when you hear disappointment sometimes

00:01:52.092 --> 00:01:54.178
it's because the enterprise
doesn't feel like

00:01:54.178 --> 00:01:55.554
they're getting a lot of value out of it,

00:01:55.554 --> 00:01:57.431
even though the end users are.

00:01:57.431 --> 00:02:00.852
Occasionally you hear the
enterprise has a top-down mandate,

00:02:00.852 --> 00:02:04.356
but the end users are not
comfortable with the technology

00:02:04.356 --> 00:02:07.610
or the implications that it
might have for their jobs.

00:02:07.610 --> 00:02:09.446
And so I would argue
you need to have both,

00:02:09.446 --> 00:02:11.615
you need to have end-user
delight and value,

00:02:11.615 --> 00:02:14.535
and you need to have
enterprise integration

00:02:14.535 --> 00:02:16.246
into the ways of working

00:02:16.246 --> 00:02:18.832
that drives value at the
enterprise level as well.

00:02:18.832 --> 00:02:21.418
And I think that's a really
important point, John.

00:02:21.418 --> 00:02:23.337
There is a disconnect right now

00:02:23.337 --> 00:02:26.508
between some of the
mandates and the commitments

00:02:26.508 --> 00:02:28.010
that we're seeing executives

00:02:28.010 --> 00:02:30.596
and CEOs making to their boards

00:02:30.596 --> 00:02:34.476
to deliver a return against
these large AI investments.

00:02:34.476 --> 00:02:36.520
But there is still a bit of a gap

00:02:36.520 --> 00:02:39.106
between what teams are delivering today

00:02:39.106 --> 00:02:41.651
and that CEO-level commitment.

00:02:41.651 --> 00:02:43.278
And this is actually a white space

00:02:43.278 --> 00:02:47.199
where ISVs have a real
opportunity to deliver value.

00:02:47.199 --> 00:02:50.119
The prototypes we're seeing
that are moving from production

00:02:50.119 --> 00:02:52.038
or from prototype into production

00:02:52.038 --> 00:02:54.959
are actually those that
respond to customer priorities

00:02:54.959 --> 00:02:57.754
or even come from customer-driven asks.

00:02:57.754 --> 00:03:00.256
If we can talk about the pricing models,

00:03:00.256 --> 00:03:04.970
you know, the pricing models
that use AI agentic solutions.

00:03:04.970 --> 00:03:07.974
Which ones are resonating with customers?

00:03:07.974 --> 00:03:09.851
Well, one of the truisms in pricing-

00:03:09.851 --> 00:03:11.562
and I work a lot with software companies

00:03:11.562 --> 00:03:13.606
that deliver all sorts of innovation

00:03:13.606 --> 00:03:15.191
but especially AI innovation-

00:03:15.191 --> 00:03:17.319
one of the truisms has been that

00:03:17.319 --> 00:03:19.488
user-based subscription models

00:03:19.488 --> 00:03:21.657
are not going be as good a fit

00:03:21.657 --> 00:03:24.285
for a lot of what AI is delivering.

00:03:24.285 --> 00:03:25.161
And so the question is,

00:03:25.161 --> 00:03:28.415
Well, what are the
alternative pricing models?

00:03:28.415 --> 00:03:30.459
What we're seeing is

00:03:30.459 --> 00:03:34.339
there are a lot of conversations
about pricing to outcome,

00:03:34.339 --> 00:03:35.716
but when you really press

00:03:35.716 --> 00:03:38.511
in terms of what resonates
with end customers,

00:03:38.511 --> 00:03:39.762
in the current state,

00:03:39.762 --> 00:03:42.182
it's some hybrid of the
current subscription

00:03:42.182 --> 00:03:43.934
and usage models.

00:03:43.934 --> 00:03:45.936
And in the future, it's a conversation

00:03:45.936 --> 00:03:48.230
that starts getting closer
and closer to outcomes.

00:03:48.230 --> 00:03:53.195
So Mary, what should
businesses be prepared for

00:03:53.236 --> 00:03:56.323
as they move toward this
outcome-based model?

00:03:57.408 --> 00:03:59.869
Outcome-based pricing really does require

00:03:59.869 --> 00:04:02.998
a very deep understanding of a use case.

00:04:02.998 --> 00:04:04.375
So I think it really comes back

00:04:04.375 --> 00:04:08.004
to very deeply understanding
the customer problem

00:04:08.004 --> 00:04:11.258
and their metrics and
their measures of success.

00:04:11.258 --> 00:04:13.803
Over time, this will
likely need to evolve,

00:04:13.803 --> 00:04:16.139
so there's something of like a telemetry

00:04:16.139 --> 00:04:19.518
that helps to measure and standardize

00:04:19.518 --> 00:04:23.564
how outcomes are defined
and attributed and measured

00:04:23.564 --> 00:04:26.526
and really operationalized at scale.

00:04:26.526 --> 00:04:28.654
Operationalized at scale.

00:04:28.654 --> 00:04:30.448
That's going to be the key.
It will.

00:04:30.448 --> 00:04:31.574
Thank you so much.