🦄 Making great presentations more accessible.
This project aims to enhances multilingual accessibility and discoverability while maintaining the integrity of original content. Detailed transcriptions and keyframes preserve the nuances and technical insights that make each session compelling.
Overview
📖 AWS re:Invent 2025 - Maximizing customer outcomes: AWS Partner Led Customer Success (PEX106)
In this video, AWS addresses how customer success is the solution to closing the AI investment value gap, as 75% of companies investing in AI aren't seeing tangible business outcomes. The presentation introduces a framework for systematically tracking and delivering value through post-launch stages: Onboard, Activate, Consume & Adopt, and Scale. Real customer examples are shared, including Gimbo with Flexa achieving 95% faster decision-making, MediaCorp with Crayon reducing content production time from 20 to 3 seconds, and Tower Insurance with Deloitte cutting average handling time by 3 minutes across 630 agents. Cynthia discusses learnings from a partner-led customer success pilot, highlighting the importance of value realization plans, industry benchmarks, and outcome-based metrics. AWS announces a value realization toolkit launching in Q1 on Partner Central and invites partners to engage via AWS-PSS@amazon.com.
; This article is entirely auto-generated while preserving the original presentation content as much as possible. Please note that there may be typos or inaccuracies.
Main Part
The AI Investment Value Gap: Why Customer Success Is the Solution
Hello everyone. We are standing at the edge of the most exciting transformation in business history. AI is creating unprecedented opportunities. Our customers are investing billions, and innovation is moving at a pace we've never seen before. But here's the reality: AI is also collapsing the competitive advantage. Small players are becoming disruptive threats, and small partners are building capabilities overnight to become massive players. Our customers have almost an unlimited number of places to go for their solutions. So what does this all mean? How do you really differentiate? The answer is very clear. You have to unlock measurable, tangible business outcomes for our customers. That's the only way to do it. That's exactly what we're here to solve together today.
Let's dive in. So here's what we're going to solve together today. I'm going to start by outlining why customer success is the solution to closing the AI investment value gap. I will then outline the framework we are using to systematically track and deliver value. I will bring it to life with a few customer examples, and then Cynthia, my colleague, will come on stage and share learnings from a partner-led customer success pilot that we've been running and also outline where we need your input.
Let's dive in. The numbers actually tell a very interesting story. The 80% deployment number that you're seeing here represents almost a doubling of the adoption rate in less than a year, highlighting the unprecedented pace at which AI is getting integrated into business operations. Just on agentic AI alone, we expect the spend to be $155 billion by 2030. To put that number in context, this was zero just two years ago. Of all the Fortune 100 companies that are spending a whopping $400 billion on technology, 74% of CEOs and boardrooms are prioritizing AI. This level of commitment represents a tidal wave of investment we've never seen before.
But here's the reality: 3 out of 4 companies that are investing in AI are not seeing tangible business outcomes. We expect 40% of all AI pilots to be canceled by 2027, according to Gartner, and 3 out of 4 companies are also not reporting any specific bottom line impact. So what's the solution to addressing this gap? We have seen that when partner teams and AWS teams come together, outline the business case and the business value up front, and take a very structured and systematic approach to driving adoption and value realization, magic happens. We also see that pilots become production at 2x the success rate. So customer success is the solution that we're seeing to closing the AI investment value gap.
In light of all this momentum and in light of all the collapsing competitive landscape, a golden thread appears, and that's customer success. When IDC asked partners what is the factor that determines their future success, customer success activities rose to the top, higher than AI-enabled transformation, higher than industry-specific capabilities, and higher than coastal activities. The same research also said that about 70% of our partners are excited and see massive potential in deploying AI solutions. So where do we need the help? They need the help in providing a demonstrable path to ROI and proving value during delivery, and that's where customer success comes in.
And we also know that repeatedly IDC has said that 95% of organizations, which means almost every organization, purchases additional products and services as a result of positive customer success experiences. So what does this mean? Customer success is not just about retention. It's a sustainable growth driver for you in this crowded space that is only becoming even more crowded. Investing in customer success will help you win bigger deals, longer deals, and boost your revenue. So let me paint the picture of how we are thinking about the post-launch customer journey. We all know that often selecting a technology provider is much more difficult than even selling technology. Our customers are navigating massive complexity, so once they have chosen us, it is our responsibility to deliver on that trust.
So we start with a clear handoff between sales and customer success, where it's a set of documented use cases we are driving, and there is a clear 30, 60, 90 day plan. In Activate, what we do is we look for those quick wins to build confidence while laying solid technical foundations. We want to get that first AI application running, the first task automated, and when it comes to consume and adopt, it becomes very interesting.
That is where you start seeing value. We are not just going to measure which features our customers are using and how many users are adopting. We are going to trace it back to tangible value. Are your developer teams actually becoming more productive? Are your support teams resolving more tickets? Is your CSAT steadily improving? That's the business impact we keep looking for.
Eventually, if we get all this right, we take what really works and we scale it to the entire enterprise. We see this pattern with customers where they take high volume, easy to automate use cases up front, put wins on the board, build the foundation, build the confidence, then they move on to higher complexity workflows. Now let me show a few customer examples to tell you how this works.
Real-World Success Stories: How Partners Drive Measurable Business Outcomes Through Structured Customer Success
Gimbo is a Brazilian retail and supply chain company managing logistics and product catalogs for 300 of Brazil's largest 500 enterprises. That's massive scale. So what was the problem? The problem was quite complicated. A lot of their decisions were actually being made just by gut and individual experience and not with data. Catalog updates were manual and time consuming. Decision making was painfully slow, and if you were not a technical user, you had little to no access to the data that you need to do your job.
So Flexa, an AWS premier partner for generative AI in Latin America, came in and said, let's build Flexa AI, a generative AI platform extending Amazon Q Business, which democratizes access to data. They started with the activate phase where they integrated multiple data sources across Gimbo and provided natural language querying capabilities to their employees. During the adoption phase, they really focused on driving deep enablement and got about 60% to 90% of the entire sales organization really using the tool and gaining from it.
The results were spectacular. They saw decision making time come down by about 95%, so decisions were getting made at 2x the speed. A lot of individual decisions that used to take hours came down to 30 seconds. What they also saw was significant productivity savings which they think they're going to reinvest, and they're forecasting 20% sales growth by just reinvesting those productivity gains. The important thread here is that now Flexa has built an extremely solid foundation for Gimbo to embark on a much longer AI journey.
The second example is about MediaCorp, which is one of the largest media conglomerates in Singapore. Imagine handling 5 million images, 150,000 hours of video, 16,000 hours of audio, and millions of text files. Every single piece of content requires a ton of metadata, transcriptions, sentiment analysis, and brand logos so the viewers can actually find what they're looking for very easily.
So what was the problem? They actually put down a lot of AI processing workloads, but they did not know how their AI processing pipelines were working. Content creation was slow. The quality of content was not very consistent, and the editorial teams could not produce content at the pace that viewers really needed it at. Viewers were finding it very difficult to find the content they needed. Basically, the operation was flying a little blind.
So Crayon, which is a top partner in Asia for us, came in and said, let's put eyes on everything for you. They implemented CloudWatch and instrumented the AI pipelines in a very comprehensive way, completely end to end, and enabled traces, logs, metrics, real-time dashboards, health checks, the whole gamut. What they also did was go into each of the different dashboards and users and make sure that there was deep enablement and usage was actually happening.
The results were fantastic. They were able to reduce the content production time by bringing it down from 20 seconds to 3 seconds, all the while improving the accuracy of AI from 75% to 85%, and they're actually now slated to cross 90% in accuracy, which is phenomenal. In addition to all of this, they brought down the infrastructure costs by 85% while still maintaining the quality and a 99% uptime. Incredible results when you really take a very structured approach to make sure there's a clear business case up front, there's adoption, there's value realization, and sustained effort across all of these throughout the journey.
The last and very exciting example that I have for you is about Tower Insurance in New Zealand. Tower Insurance has about 630 agents all answering thousands of customer calls on a daily basis, and they were using age-old contact center solutions. It just wasn't cutting it.
Customer calls were being answered too slowly, and customers were getting frustrated. The agent experience was also quite frustrating because agents could not find the information they needed, and the quality of responses from the agents was slow, inconsistent, and the operation wasn't as efficient as it could be.
Deloitte came together with Tower and said, let's transform this whole operation. They deployed Amazon Connect and implemented five different AI capabilities, ranging from email automation to self-service bots for routine requests, as well as Amazon Q to help agents find answers to their questions very easily. They conducted deep enablement with the agents and supervisors, ensuring that the agents understood how all of these solutions worked together for them. They rearchitected their workflows on a daily basis and provided change management and enablement to make sure the agents understood what their future workflows would look like and could act on it.
They were able to drive 60% employee adoption, and the results were tremendous. They brought down the average handling time from 18 minutes or more to 15 minutes or more. You may ask, "It's just 3 minutes," but multiply that by 630 agents and across thousands of calls on a daily basis. Most importantly, Deloitte has now given Tower a fundamental set of capabilities that they can use for their AI transformation, and Tower is just getting started on their AI journey.
So what does this all mean? You need customer success to really translate your AI investments into tangible value. On this slide, you're seeing a set of our partners who have embarked on a journey of experimentation with us through a partner-led customer success pilot. They're testing new approaches and really telling us what's working, thoroughly informing how we approach customers, do customer success well, and unlock outcomes together. I want to take this opportunity to thank every single partner on this slide for their commitment to this journey and more importantly, the commitment to what matters the most to us: our customers.
Partner-Led Customer Success Pilot: Key Learnings and the Path Forward
Now I will hand it to Cynthia to share what we've learned from the pilot and where we need your input. Thank you, Bhargs, and I want to echo Bhargs in thanking everybody who's been a part of the pilot. I'm going to share some of the learnings we have from it. I would love to see this slide have many more partners on it next year, so at the end I'll have an email address that if you're interested in learning more, you can email us.
Moving to what we have learned, we investigated five areas. I will go through each area and the top-level findings for each of them. For the first area, regarding the progression of the stages that Bhargs talked about, we wanted to understand how to measure customer progression through these post-launch stages. What we found is that many of us are still stuck in the need to do simple cloud opex measurements. It's actually a big mindset shift to move to doing business metrics, ROI metrics, and potentially revenue-generating new deals. What we are going to do for our partners is work to find industry benchmarks about the right types of metrics based on the industry use case and domain that you can use as your starting point with your customer to build out a really strong business plan we call a value realization plan.
The second area is understanding what level of effort it took for partners to move a customer across each of these stages. The first thing I learned is that putting it on a slide is not nearly as easy as reality, so there's not one simple answer across all of these. This is partly because various partners are at different levels of maturity around customer success and business value realization, so we're going to have to tailor enablement to where you are. It's also because what needs to be done can totally change depending on the customer. For example, Flexa worked with Gimbo, and those were people who did not really have a generative AI understanding. What they ended up doing was having them download a consumer chatbot on their phone, take it home, and use it with their kids to help with their homework so they could understand AI. It's a great solution for that scenario, but if you have another scenario where you're doing QDev and working with developers, you don't need to do that. You can just go straight to a hackathon and they'll understand it. So we're going to be creating frameworks with very detailed scenarios of what we recommend for you to do, and that's going to be part of an enablement plan we put together to help partners come along this journey.
The third area is how we think about business model. I think we all know this. We're at a point right now where customers are really starting to look at pricing based on outcomes and partner CFOs are not really interested in something that doesn't have predictability yet. To help us all work through this, what we're working on is developing telemetry.
With benchmarks of best practices so we can help you with predictability about knowing if you're doing well. The goal is to lower average handling time, and you can generally commit to a 20% improvement over several months. We're going to be putting that all out through telemetry and automation to help you.
The fourth area is around what types of frameworks, playbooks, and tools are needed. We learned from all of our partners that having the language of a framework at the top and a playbook for each was super helpful for the conversation. But the most important was the one that comes first: the business plan I talked about, the value realization plan. It allows you and the business people at your client and the technology people at your client to sit in a room, hash it out, and all agree up front on what you are aiming for and how you're going to measure it. That is going to be an anchor on any kind of program that we put out so we're all basing it on something you can use with your customer.
The final area we looked at was governance. I thought we'd be getting feedback that governance should be really light. What we instead heard was that governance should be part of the rhythm of business that you each have with AWS and also with your customer. It's super important to have a regular cycle revisiting the value realization to make sure you're on track, and it's super important to keep AWS updated. What you did call out though is that we need to explain the roles and responsibilities. There are a lot of different people at AWS, a lot of different people at a partner, and a lot of different people at a customer. We're working on internally all of that roles and responsibilities starting with the PDMs so you'll know and you'll be speaking the same language as the people you work with at AWS.
We have a lot more deeper learnings that we can share later, but those are the five areas we looked into and what we're focused on. What comes next is the most important thing: the value realization toolkit. Look for it sometime in Q1 to start being posted on Partner Central for you to be able to access it. Then as we get more information around adoption patterns and best practices working both through you and also with our direct services teams, we're going to try and automate that. We might have to share it manually first, then we'll put it out as telemetry. And then finally we're going to start doing domain level assets as we expand into more and more domains and perhaps industry and use case level assets next year as well.
With that, the final thing is this email here: AWS-PSS at amazon.com. I have a team of what we call partner success specialists. These are people who work with partners aligned with the PDMs, so we're not going around the PDM, but we're going to work with partners to help them translate the way they do customer success to the way we do it, or for some partners who are just getting into it, to help them set up the practice. If you email this alias, we'll put you on the list of people who are interested in learning more and we'll get back to you. We're also working on posting the IDC data that was shared at the start on our external site so we'll be able to send that back to you as well.
So one final thing: please take the survey in the mobile app and feel free to email us. Thank you.
; This article is entirely auto-generated using Amazon Bedrock.
Top comments (0)