Personalize omnichannel customer experience with Amazon Connect

Ok. Hello everyone and welcome to our breakout session. We are so excited to have the opportunity to talk to all of you today about how you can personalize the omni channel customer experience with Amazon Connect.

I am Hannah Bloking and I am a Solution Architect Manager with Amazon Connect. Later in the session, I'm going to be inviting Mike Nolan onto the stage with me. He is our worldwide IVR specialist ESSAY and he will be showing you a live demonstration of what a superior customer experience looks like with Amazon Connect.

After that, I'm going to invite two CX leaders at Adobe to the stage, Turing Gupta and Preti Sahu are going to share Adobe story with Amazon Connect.

So there are a lot of reasons why I'm really excited to be here today. But one of the reasons I'm most excited about is that I have the opportunity to talk about my favorite part of the contact center. It's the one part of the contact center that's so critical and it's the one person you really can't automate and that is the customer.

So just by a quick show of hands, how many of you are from the contact center space? Wow. A lot of you. Ok. Wonderful. Thank you for everything that you do. It's so important for your customers by a quick show of hands who isn't here from the contact center space? Don't be shy. Ok. Amazing.

So even if you've never spent a day in a contact center in your entire life, I have got grades for news for you. You are still an expert and do you wanna know why we're all experts because we are all customers?

So if you don't believe me, I think you will shortly, I'm gonna try to prove it to you. Will you all do a little bit of a social experiment with me? Thank you. Please everybody stand up. Ok, I promise. Nothing. Nothing bad. This is good. Ok, so you have all been in Vegas now for a couple of days. So presumably you've had one or two customer experiences. By this point, you've been out in taxis, you've been at restaurants, you've been at casinos.

So sit down if you've had a negative customer experience in the last 24 hours, 0, 48 hours. Let's extend that to a week drop in like for like two weeks. Ok. If you are still there, if you're still standing, you are very gracious or very lucky. But thank you very much everybody. You have just proven my point that you are all experts in customer experience now.

Um delivering a great customer experience it's really hard. And one reason that makes it really hard is because customer expectations are constantly evolving. And on top of that, customer expectations vary by things like age group demographic. In fact, 59% of customers think customer service is more important now than it was before the covid-19 pandemic. And on the flip side of that, 91% of baby boomers and 74% of gen z think that it's more important than ever.

But why is it so hard? There's actually a lot of reasons, but I wanna talk about some key themes that we hear about from our customers.

So number one, the customer's preferred channel isn't available. This is really important where you have to think about the whole design of the customer experience. End to end everything from a customer engaging with your company on your website through digital channels all the way through the phone and while digital channels are critical, 75% of inbound interactions in the us remain voice. So if customers do need to talk to somebody, it's important that they can get it from a person.

Number two, unfriendly out of date service options when websites and self-service options are archaic and difficult to use. Like press one for sales, press two for support. It drives away customers and it leads to abandonment. 36% of customers will abandon when they're frustrated by too many options or long menus.

Number three, long handle times who here likes to wait for a long time, put your hand down. Ok? So anything that doesn't save a customer time is fried in to delivering a great customer experience. It's critical. You get customers to the right resolution and you get it to them fast.

So fragmented technologies across the contact center, it makes it really challenging and sometimes impossible to provide a seamless consistent view of a customer creates a fragmented view. And this is what contributes to those long handle times, long hold times and multiple contacts resolve the same issue.

Remember those high customer expectations, we talked about just a minute ago, 70% of consumers under 25 prefer automated experiences. Ok. And last challenge we'll discuss today. Number four disconnected interactions. This prevents companies from personalizing the customer experience. This goes from everything from self-service all the way through when agents are engaging with customers so that they have everything that they need to provide a personalized experience to customers.

And this is so important because 75% of customers would switch to a competitor if it was more convenient to do business with them.

So at Amazon Connect, we don't just help our customers meet customer expectations. We work backwards with our customers to ensure we give you the capabilities, you need to exceed them.

So we continue to evolve rapidly based on what customers say they need and what their most top of mind. Business challenges are at Amazon Connect. We have tens of thousands of customers supporting 10 million contact center interactions a day across all industries and sizes and we're responding to their feedback and their needs. And we're working backwards from these asks to deliver what's needed for our customers to exceed their customer expectations.

So now let's talk about what it means to deliver a superior customer experience. It means you give your customers choice so they can engage with you on the channel of their choice. You design an end to end experience that's modern and intuitive so that customers can get self-service or get assistance when they need it the most.

It means leveraging smart a I powered automation to reduce all of those friction points that cause long wait and long resolution times. And you leverage every single customer touch point to personalize every single service experience.

So when you do this, you can improve customer satisfaction, reduce your customer's effort and decrease handle times by personalizing self-service experiences, which has traditionally been a challenge. You also give yourself the ability to make sure that all of your high-value engagements are prioritized for live assistance.

So what makes Amazon Connect unique is our ability to deliver seamless experiences across all channels. This means with Connect, you build once and you deploy across multiple channels in just a simple to use UI. And because every single touch point is managed and Connect, you can use that data to build dynamic and personalized experiences across all of those channels with AI and machine learning.

At the heart of every single interaction you can use Gen AI to generate humanlike conversations and comprehensive FAQs. And what does that really mean? Well, I mean, customers can self serve and feel like they're still getting a live interaction from the best agent in your contact center.

When your customers do need live help and they engage with you over the phone, they get reliable, high quality voice no matter where they are in the world.

And now I am going to invite my colleague Mike Nolan up to show you what an example of a superior omni channel customer experience looks like.

Mike: Hey, what's going on re invent how we doing? All right. Awesome. So like Hannah mentioned, my name is Mike Nolan and I have the honor and privilege of giving a demonstration of the omnichannel capabilities of Amazon Connect.

Before we dive into the demo, I want to introduce you to our protagonist. Meet Nikki. It could be you sure, meet Sheila. Um meet, meet Nikki.

Nikki has been working really hard in her current role for the last six months and has recently secured a promotion along with that promotion comes with an increased compensation package that means more, take home pay at the end of every pay period. In addition to that, Nikki's also going to be traveling a lot more for work. So she's going to be booking things like hotels, flights, cars and other travel expenses.

So Nikki's three weeks into her job and she's been traveling and booking a lot of uh trips, um and she goes to book another trip, but her card gets declined. Now, let's take a look at how Nikki and her bank. Any company financial resolves the issue.

Nicky's first notified of the declined transaction via a push notification through the through the bank's mobile application. Niki goes to the her financial any company financials website and sees the opportunity to chat in the lower right hand corner.

Nicky's greeted with a personal, a personalized experience from the virtual assistant, greeting her by name and also recognizing why she may be contacting us today. Of course, Nikki is contacting us about her declined transaction and selects as such.

The virtual assistant explains the situation to nie that she is over her credit limit but also has a resolution for her and she's that she's eligible for a $5000 credit limit increase. Nicky's interested in applying the credit limit increase to her card and the bot fulfills her request.

Now, we've successfully unblocked Niki from booking her trip, but Any Company Financial has also identified Niki as a good candidate for their traveler rewards credit card. So while we have Nikki's attention, we ask if Nikki if this would be something Nikki is interested in being that she's traveling a lot more for work. She's she's inclined to, to listen more and wants to talk to an agent to learn more.

So we're going to slide over to our agent experience. And on the left-hand side, you can see the entire transcription of the digital assistant and Niki and the agents able to dive right in and start helping Nikki wants to just learn more about the traveler reward credit card. And on the right hand side, you see Amazon Q has generated a response based off of what Nikki is asking. The response is generated from the traveler reward card. FAQ page is noted at the bottom.

The agent reviews the response and decides that it's an appropriate response to respond to Niki with and selects to do so.

Micky is interested in the card and wants to proceed with upgrading her card to the traveler reward card. The agent uses agent guides to walk him through the process of upgrading her cards, her card. He confirms her basic information, confirms the new card and the delivery address, reviews and confirms all of nicky's information and submits nikki is now upgraded to the traveler reward card and should be receiving it within 5 to 7 days.

In this experience, we were able to resolve nikki's issue by offering a personalized experience and allows her to unblock her from booking travel. We also identified an upsell opportunity to see if nikki would be interested in the traveler reward card

In this customer journey, we saw personalized service experiences, an intuitive, easy to use self-service in allowing the virtual assistant to resolve Niki's issue with just two simple interactions. This leaves Niki feeling delighted and she's thinking about where to spend all those traveler award points.

Now, three days later, Niki receives her first paycheck with the increased compensation. She wants to use the additional funds to start investing, but she's new to investing and is feeling a little overwhelmed and that the industry is filled with a lot of jargon and she just wants to make sure she's doing the right thing for her future.

So she starts by navigating to Any Bank's, Any Company Financial's, wealth management website and sees the opportunity to start a call in the bottom right hand corner of the page. Niki is connected with an agent and the agent identifies that this is Niki's first time investing and sees that as a good opportunity to escalate the conversation to a video chat in order to build a relationship with Niki and earn trust. Niki agrees and they start chatting through video.

So we'll switch to the agent experience here. After Niki and the agent are conversing, they find that the best course of action is to send Nie an informational guide to help her get started through their digital collateral on their website. The agent uses step by step guides to identify the channel in which she wants the information sent being SMS.

He enters in Niki's phone number and creates a follow-up task for seven days so that somebody can reach back out to Nie in seven days and see if she has any other questions or help close the sale. The agent identifies the proper informational packet to send Niki, confirms all of her information, and submits it.

We were able to send Niki the information that she needed and in the background, create an Amazon Connect task to follow up with Niki within seven days. In this experience, we provided digital artifacts to help Niki with her wealth management education. And we created a follow-up task with Niki to help close the sale or see if Niki has any additional questions.

During that interaction, we saw a customer channel of choice in the agent recognizing that this was an opportunity to build trust and a relationship with Niki. The customer in turn, Niki, feels like a valued customer and she's more likely to invest in their brand.

Two days later, Niki receives her traveler reward card. Any Company Financial reaches out to Niki telling her that her card has been delivered and asks if she has any questions. Niki's card has been delivered, but unfortunately, it was delivered damaged. So the virtual assistant recognizes that they need to order a new card for Niki and places the order with expedited shipping.

But the envelope was torn and Niki's a little uncomfortable and feeling a little worried that somebody might have seen her data, her PII. And I mean, who's been the victim of identity theft or a stolen credit card or anything like that, right? By show of hands, right? It's becoming more and more prevalent and when that happens, it's a violation you feel intruded on and all you want to be able to do is talk to somebody at your financial institution and ensure that you're secure. And that's where we find Niki.

The virtual assistant recognizes this as a security concern and escalates to an agent. The agent is able to join the SMS conversation right from Niki's virtual assistant conversation. Again, Niki's a little worried about people seeing her information and the agent offers to give her a call on her phone in which she is happy to oblige.

So on the agent side, you see that the chat ongoing with Niki, but the agent's gonna go to the upper left hand corner and select, make a call to give Niki a call. He'll enter Niki's phone number in the number pad area and click the blue call button to initiate the call.

The agent and Niki are now connected through voice and they're able to identify that the next appropriate action would be to put a security alert on Niki's account in the event of any fraudsters trying to impersonate her.

In this journey, we saw proactive outreach in reaching out to Niki to see if she received her card okay and to see if she had any additional questions. We were able to quickly and efficiently reorder a card for Niki, but she still felt unsure about the security aspect. So we escalated to an agent and put a security alert on her account.

In this experience, we saw smart AI powered automation and a customer channel of choice being that escalation to a phone channel. This leaves Niki feeling relieved and secure.

In this demonstration, we saw omnichannel customer experience and agent empowerment and productivity tools. On the customer side, we saw the channels available to reach your customer where it's convenient for them or to use channels as a way to escalate and have a more intimate conversation to earn trust and build a relationship.

We saw proactive outreach to reach out to Niki to see if she had any additional questions with her new card. We saw task management being used to follow up with Niki in seven days to close the sale or to help Niki with her investment strategy.

Finally, throughout all of this, we saw AI being able to quickly and efficiently resolve Niki's issues. On the agent side, we saw AI able to assist the agent in their responses, expediting the time that an agent has with a customer. We saw step by step guides being used to solve common customer issues that allows agents to have the same level of service regardless if they've been with the company for two weeks or 10 years. And we saw this all within a channel consistent experience in a single pane of glass agent workspace.

Now, I'm super excited to also say that we highlighted two channels that we just went generally available with yesterday - SMS and in-app calling. I'm gonna invite Hannah back on stage to learn a little bit more about those channels.

Christy: In our business, it's extremely important for our business partners to be able to open up new contact centers quickly as our customer demand change or specific customers need us to be available in certain time zone and certain countries. Prior to Amazon Connect, it used to take us easily three plus months to set up a new contact center. Thanks to the new platform that we have put in place, now we are able to open up contact centers in three weeks or less. We have zero dependencies on the MPLS and other telecom infrastructure challenges that we used to run into. And our business partners are extremely happy with the agility we have been able to provide them.

Mitesh: I'll talk about a very classic example that most of the people in the contact center industry can relate to, which is the IVR routes and the transfers between the sales and the support teams. So we had the same problem and our sales team was getting frustrated because they were getting a lot of support calls and that was taking them away from the revenue generation cost that they needed to do. Plus they had to transfer the call back to the support team which resulted in a broken experience for our customers as well.

So with Amazon Connect, we were able to drill down with the analytics on the IVR itself to see where the problem lies. Like what's prompting the customer to choose the wrong options or are we not understanding what they are saying on the NLP IVR as well? So we were able to fine tune the IVR both DTMF based and the NLP IVRs to see where the problem lies and how we can adjust and, and we did multiple rounds of testing on that and we saw actually significant reduction in the transfers that were happening previously. So the capability to fine tune do the A/B testing and make those results counting was one of the key things for us as well from Amazon.

Christy: The first example that comes in mind is like the translation or the geo fluency services that we call, which allows our customers to interact with us in their native language and the agents who are not that native language speakers can interact with them on the messaging side, not on the phone side though on the messaging side. So as an example, if you are a customer from Germany and I am only an English speaking agent, I can still support you on chat because whatever you are typing is coming back to me in English and whatever I am responding back is going back to you in German.

So that kind of helped us with providing a more holistic end to end 24 by seven support for our customers across the globe because we as like 68 countries. If we are operating, there's some place which is already always online. So based on follow the sun model as well, we are able to support our customers across all time zones, even if their local time zone was not available. And we see significant value from that being done on the European languages. And now we are trying it with the Japanese language, which is a more tougher language from a translation perspective. But we are seeing early success on that.

Mitesh: When the pandemic hit in 2020 we had to send every agent to work from home. The we had our voice on the legacy chat voice platform on prem systems. And we started seeing voice quality issues especially in the Jpac region and the way the Japanese language is spoken, we started seeing a lot of issues in the Japanese customer interactions and we were seeing a lot of complaints come across.

So when we started this migration for Amazon Connect, we were able to use the tools provided by Amazon and the third party tools to see the entire path of the call where the blockages are happening, which points could be causing those issues and we were able to resolve before we went live. So by seeing and improving on those blockaging points. we were able to provide a more, you know, I would say a pretty decent experience on the voice side and we got some feedback from a Japanese agent and which is a very tough feedback from them. Like it was as good as talking to you here when they were talking to the customers.

So, you know, we take that compliment seriously because that was a lot of effort had gone into that tough customers.

Mitesh: With the first step of doing the heavy lifting on the chat already done and then the migration to voice for omni channel on the same platform was much more easier for us because all the integrations were built in the agents were already in the system. The queues, the routing profiles were already in the system. It was just a matter of enhancing it to take in the omni channel experience and the configurations versus just the chat configurations which were earlier there. So I would say it was seamless from a technology perspective, but a lot of learnings were needed by the WFM team and the operations team on how to fine tune the omni channel cues to make it work the right way.

Mitesh: One key thing which we had was like the number of phone numbers across the globe and and that was becoming unmanageable. All the agents wanted a separate DID for them to get the callbacks for their customers. And we had toll free numbers across anyways for all those countries. So with Amazon's capability, the routing capability, we were able to bring in something called as last agent routing and that helped us to get away from that DID model.

So if you are looking at, let's say 2000 agents each having their own DID and if you are able to reduce that to a single number across regions, you are maybe looking down to maybe 50 or 60 numbers replacing those 2000 numbers. And that makes the system more simplified, it results in cost savings for us. And plus the challenges if an agent leaves tomorrow, what happens to the calls that they were supposed to get from their previous customers. So all of that gets taken care of with this last agent routing and helps in reducing the cost on the telecom and the manageability on the number side as well.

Mitesh: The other key thing I would call out here is the the pay as you use model of Amazon. Because previously with our legacy on prem system, we were using per agent license. So every agent was either a chat agent or a voice agent. And if we had to make a voice call back to the customer, because we were not able to solve the issue on chat, we had to, you know, give it to somebody who had a voice license. So he, he would the that particular agent will call the customer back and that introduces a time lag and plus made the experience more friction. But with Amazon's pay as you go model, every agent is now an omni channel agent. It's more of a training thing that they can use the system the right way versus any licensing or a cost innovation on that.

Christy: I think the next big one we are looking at briefly touched upon it. Our enterprise customers raises probably 80% of their support issues through web cases and we do plan to integrate this web ticket routing using the same omni channel routing that's in place for phone and chat channels because we strongly believe that will give a much better experience for our enterprise customers than that is in place today.

We are also looking at expanding our AI/ML capabilities. We are introducing conversation summary for our agent so that whenever there is a transfer from a board to an agent or agent to agent transfer, we are quickly to provide a summary of the past conversations for the agent to respond. Rather than you know, putting a customer on hold for a few minutes to read through the case notes.

There are other agent assist capabilities. We are introducing some of it right now. You saw a glimpse of it in the Connect announcement. We have developed something very similar that's going into production shortly.

We will also be evaluating the web and the video calling capabilities just to see if that's going to give us a differentiated experience and more personalized experience for our customers and that's on our road map as well.