As enterprises invest in AI proof-of-concepts for customer support, software companies are getting squeezed with acquisition costs rising. Will these once glorious companies ever get their glam back?
After chatting with GPT-4o, I am quite certain that most customer support jobs are on the way out.
Who wouldn’t want an AI agent, available 24/7, who is helpful, friendly, and best yet….no need to navigate the dreaded phone tree?
Companies will find AI agents, not only a great deal cheaper, but better than human. The biggest challenge will probably be integrating the technology in existing software.
Have you used the Voice Mode feature, JK? That still continues to blow my mind. Lots of promise and scope for AI in this area. I am very curious however on how AI is going to traverse the empathy side of the communication. Having worked with Customer Support department in prior projects, I know that they measure Customer Satisfaction (CSAT). One of the questions they use in their CSAT surveys is about empathy. for e.g. "Was the agent able to empathize with your situation?".. Not all companies use this question. I know Applecare uses it.
This empathy aspect will only be known after AI has a more meaningful scale of deployment in the enterprise. I do feel like GPT4-o was able to really push the boundaries of providing empathetic customer support...
“I am very curious however on how AI is going to traverse the empathy side of the communication.”
In fairness it’s a low bar. I had a credit card stolen when we were traveling in europe earlier this year. The customer service rep kept repeating in this monotonous almost robotic tone every 30 or so seconds
“I’m sorry. I know it’s inconvenient when you are traveling”
Why yes it is, but it’s even worse when you are on a call that should take 3 minutes, but really takes 45 minutes with 32 minutes of fake empathy apologetic filler.
And fwiw microsoft did a demo years ago of an ai bot making a hair appointment. It sounded eerily real and human.
I think this is something they will solve. And in any case I would bet fewer mistakes from reps who don’t know anything would more than make up for a questionably empathetic tone.
That said there was a software company several years back that has a suite of tools to analyze customer mood/mental state. it’s pretty easy to envision such analysis in real time that instructs the agent to turn on empathy.
I think the biggest issue will be oversight. There are QA departments to check calls of live operators. Someone will need to figure out the tooling to do those checks on what bots are actually saying.
In any case it should be better for consumers. I worked at a call center briefly out of college, and one of the favorite tricks for some groups was to pick up a call and hang up immediately. It reduced the dropped call ratio, and increased calls per hour which were both important metrics. i have to believe our AI overlords will provide a better experience.
Im rethinking what I said after reading both comments from JK and Bret.. The empathy metric of customer satisfaction, in all probability, is outdated especially given that CS went for a toss after the pandemic. Before the pandemic it maybe made sense and few companies who were really geared towards giving meaningful customer support would track these metrics. Now - Im not so sure - so yes - it would be (to borrow Bret's words) a low bar, especially after all the conversational skill GPT4o demonstrated.
Bret to your point on oversight - I can tell for sure that there used to emphasis even before on what human agents would be talking to customers. This was done by certain tools that Five9 and other systems would provide where call/chat transcripts were provided on demand... But the volume of those transcripts were intense, so my product team had worked on a few projects with the CS teams to help them get more insight. Salesforce, Five9 did not have the kind of capabilities were looking for.
Im assuming now all of that has changed and its easier for QA's to derive insight from multiple chat/call transcripts at the same time even the interactions held by AI bots. There are still going to be issues inherent within the scope of what we see today with LLMs but it should eb solved quickly
Yes. I think one of the reasons why I also brought that up was it may get tough to (ahem😁😁) really expalin issues to an AI bot to get refunds/returns from lets say Amazon or Walmart once these are implemented. Its a just a theory.
Very interesting read. Customer support everywhere is broken and needs a big fix. Hope AI can provide that. Although, Im not full sure about how capable AI is for prime time. I watched the video demo and I can already see some minor errors and the conversation getting extended for no reason. Still, Im optimistic ChatGPT will solve this really quickly.
Thank you Sully, glad you enjoyed the post. I fully agree that customer support is fundamentally broken and perhaps a big part of that is how it is treated as a "cost center" as the customer has already been acquired. I agree that we are still a couple if not few years away from AI taking prime time in customer support, as there still needs to be a lot of "hand holding" by human beings in the interim period, especially as the effectiveness of these AI customer agents depend a lot on the quality of data. But, I am optimistic long-term when it comes to genAI elevating customer satisfaction as a whole.
Very interesting and enlightening perspective on the matter.
I was pondering about a couple of other dimensions...no real answers at this time...will have to see how things play out over the medium to long term.
1. How will AI in customer service change the growing services export industry in countries like India, Mexico and Brasil? I would think that hiring in these industries in these countries will drop as AI driven automation rises.
2. Will AI result in more nearshoring and onshoring into the US which at this time is driving almost all the AI innovation and will likely be the owners of these AI platforms?
Such an interesting time to be investing and tracking these business developments!
Indeed it is an interesting time. There is just so much to keep track of!
1. My view: AI is disrupting any kind of IT and BPO off-shoring project. With the Philipinnes, the impact was immediately felt and lawmakers last year itself have tried to take measures. This was from last year: https://www.pna.gov.ph/articles/1200928.
But I suspect its not going to be something that will sustain because those that have figured out the right mix of AI capabilities in CS & CRM departments while also assessing the customer satisfaction and ROI from the project will migrate CS workflows onto the AI platform. The others are still possibly working on PoCs or may be still assessing how to deploy AI but may move to reduce CS costs further, in that case, they actually will end up offshoring more CS workloads to Manilla or Mumbai etc. So in the interim period they wont be hit as much. Im was also trying to track any possible data such as bankruptcies etc of BPO Outsourcing companies in Manilla but cant find much.
2. AI could possibly result in near/onshoring not just in the US since US has currently outgrown any other hub around the world. Im not fully certain of this yet, because there exist many other AI hubs as well around the world like Paris, Israel, India etc... For example, I ve read MSFT and GOOG are offshoring many jobs to India and Brazil, even some AI jobs.
Excellent article—you are right on. I’d like to add one comment that colors the entire CS industry. Companies have two, or more, accounting lines in their business—COGS (cost of goods sold—including DevOps for software companies—I’m lumping CAC and COR here together) and operations/overhead expenses OpEx (internet, phones, computers, etc). They have to decide where customer service/support goes. Most decide it is OpEx and therefore as you describe. A mostly fixed cost regardless of the number of units sold and doesn’t drive revenue. This accounting report methodology alone makes CS vulnerable to cost cutting as cost is the only metric that matters in OpEx. If customers purchase because there is high quality customer service, there is no method to measure the revenue increase—so no credit to the excellent customer service team for their revenue growth impact. They are just an overhead expense. Unless they have an executive who knows this and is a tenacious blankety-blank and coordinates with sales to promote the customer support team’s excellence. The trend is those executives are fading away and customers are losing out.
The best strategy is CS as part of GPM—gross profit margin—and therefore as revenue after support to service customers. CS then has to become part of development and QA/QC—higher quality products will lead to higher revenue—post CAC customer acquisition costs. This will not come from just an AI agent—a company needs well trained support staff to lead the development work integration. AI will be a powerful tool but not a complete substitute for a wholistic support process which generates revenue through high customer experience on many fronts.
The customer support software companies salesforce, five9s, etc. have pushed for integration and automation for the past decade and this part is moving quickly towards AI which puts them at some risk—as you point out.
Thanks very much Dean! I actually did not know that organizations could also report the CS cost center under COGS. But your reasoning does make sense if they sell CS in the premium attach rate to the core product purchase. So, for example, Apple sells Applecare as a premium service. In your experience does that get recorded under COGS line item?
You're correct in asserting that customer support was ripe for AI intervention. It will be interesting to see if other companies follow Klarna's example and, if so, how well that plays out over the long-term. For me, there's nothing like interacting with a human -- even if that "human" is a highly-sophisticated chatbot. Utlimately, however, I don't care who (or what) I talk to so long as I get my problem taken care of quickly (which includes another KPI - First Call Resolution).
Thanks Paul, I agree with you on the matter that I don't necessarily care whether I talk to a human or a bot as long as my issues are resolved. And given that genAI models are "goal seeking" in nature, I believe they might actually be more effective over time to drive better customer support outcomes. Having said that, it's issues that fall outside of defined parameters for which the AI agent may not have sufficient context to produce a response that could be a hurdle, but my thinking is short-term, there will be a lot of "hand holding" with human supervision, while over time we see roles such as chatbot developers, conversation designers emerge (something that the Intercom report also hinted at).
Having worked in this space as a marketer, I do hope there is room for skilled human agents. I've seen firsthand some of the crap those folks have to put up with and I consider them heros. I've also seen the empathy that an agent expressed when dealing with a customer that no AI could even begin to emulate. So much so, that sitting there, I started to cry. I gave that lady a hub when it was over and said, "You are amazing." She was.
CS is truly the department full of unsung heroes. It is a pity that the work is often considered "tactical" and not attributed directly to revenue growth, when delightful customer care experiences actually result in repeat purchases and deepening adoption.
Right on, Paul! I will be curious to see how I can make the case for a return/refund to Amazon for a situation that AI may not be trained for. Lets say my Fitbit strap broke on the 31st day after purchase and I was trying to make a return/refund but AI may not understand that because it falls outside the 30-day grace period. Sometimes, I can convince human CS agents to take care of that for me.
Also you are right about FCR.. Just curious Paul - do you have any experience in working with CS/CRM folks in your work? Not many outsiders know about FCR as a CS metric😉
I worked as a digital marketing manager for an outsourced BPO company for a couple of years, and have some other experience writing about topics in the space.
Hey Amrita and Uttam, a great post, as always. I'm wondering whether sooner or later there'll be a company that'll make it a mission to help other companies streamline customer support? Just brainstorming :)
There are tons of those companies already who help streamlining customer support operations. Salesforce, Five9 and others were from 10 years ago and promised a certain level of streamlining of operations. Those streamlining benefits failed as we moved through the pandemic which is why there still is elevated levels of perception of decreased value - longer wait times, less empathetic human CS agents, more confusing ways to speak/chat with agents etc..
There are startups like Dialpad and Intercom which have started taking a lot of market share from legacy providers and they provide almost end-end solutions to in manage CS operations....
...The question now is that enterprises and basically, anyone, looking to add customer support for their customers aren't looking for just streamlining anymore... they're looking for productivty and the subsequent cost benefits.. Like in Klaarna's case - they were supposedly able to reduce their CS costs by $40M and yet increase the volume of interactions with their customers.
Thank you Michael, it will be interesting to watch how traditional SaaS business models adapt in the coming quarters and years, as companies do more with less.
Customer support is usually one of the sore points of all companies: they assist you and are around you before you become a customer, once inside their network you are just a number to manage (and a burden). AI could set new and better standards, good point.
Fully agree with you Mirko. I think AI agents can actually get better at resolving customer issues faster and more effectively. Interestingly in the Intercom's report, support teams were asked what new roles they anticipate being created by AI and the answers were Chatbot developers, Chatbot analytics and conversation designers.
Yes. Very true. I've worked with Customer Support teams in the past so Im very aware about the business and the metrics they follow. While researching for this post, I do see that there is a new range metrics, standards and KPI's that are being used such as number of tickets managed by bot, bot handled time etc... After all we had to endure through the pandemic with 1 hour long wait times.. im excited to see the things that AI could do.
Thank you for explaining the customer support operation in details. I certainly feel the threat for the customer support software companies as their business model is upended. I hope they will come up with ways to reassure the customer's value. AI is eating software! But our life can be better with it than without!
Thanks Marianne. I think it is most software companies whose business models will be under pressure especially with their "seat-based " pricing model, leading to lower ACV and thinner margins. Having said that, I think large, well funded and fully-integrated software companies are probably better positioned than point solutions in this decade given increasing vendor consolidation.
As for businesses, many of them are taking Klarna's route to build AI proof-of-concepts across use cases, such as customer support and deploying it over the course of multiple business cycles while assessing the ROI on these projects. The irony with customer support has been that it is treated as "cost-center", as the customer has already been acquired and especially after the pandemic, the quality of CS has degraded across organizations as they looked to optimize their cost structure. So, I am quite optimistic that given the goal-seeking nature of genAI models, we can actually arrive at a point where we have "cost-effective" agentic customer support that actually improves the overall customer experience.
Thank you for alerting us, Allan. Somehow the grammar check tool completely skipped the 2-minute summary. We've made the changes for you. Apologize that you had to go through the experience.
Amazing read and insights! Regarding BPOs and the potential threat from chatbots, they have a window of time to adapt. By embracing AI and automation as enablers rather than threats, they can thrive, and some BPOs already began this transition years ago. And as AI continues to advance, they will shift towards offering high-value-added services while improving margins.
After chatting with GPT-4o, I am quite certain that most customer support jobs are on the way out.
Who wouldn’t want an AI agent, available 24/7, who is helpful, friendly, and best yet….no need to navigate the dreaded phone tree?
Companies will find AI agents, not only a great deal cheaper, but better than human. The biggest challenge will probably be integrating the technology in existing software.
Have you used the Voice Mode feature, JK? That still continues to blow my mind. Lots of promise and scope for AI in this area. I am very curious however on how AI is going to traverse the empathy side of the communication. Having worked with Customer Support department in prior projects, I know that they measure Customer Satisfaction (CSAT). One of the questions they use in their CSAT surveys is about empathy. for e.g. "Was the agent able to empathize with your situation?".. Not all companies use this question. I know Applecare uses it.
This empathy aspect will only be known after AI has a more meaningful scale of deployment in the enterprise. I do feel like GPT4-o was able to really push the boundaries of providing empathetic customer support...
“I am very curious however on how AI is going to traverse the empathy side of the communication.”
In fairness it’s a low bar. I had a credit card stolen when we were traveling in europe earlier this year. The customer service rep kept repeating in this monotonous almost robotic tone every 30 or so seconds
“I’m sorry. I know it’s inconvenient when you are traveling”
Why yes it is, but it’s even worse when you are on a call that should take 3 minutes, but really takes 45 minutes with 32 minutes of fake empathy apologetic filler.
And fwiw microsoft did a demo years ago of an ai bot making a hair appointment. It sounded eerily real and human.
I think this is something they will solve. And in any case I would bet fewer mistakes from reps who don’t know anything would more than make up for a questionably empathetic tone.
That said there was a software company several years back that has a suite of tools to analyze customer mood/mental state. it’s pretty easy to envision such analysis in real time that instructs the agent to turn on empathy.
I think the biggest issue will be oversight. There are QA departments to check calls of live operators. Someone will need to figure out the tooling to do those checks on what bots are actually saying.
In any case it should be better for consumers. I worked at a call center briefly out of college, and one of the favorite tricks for some groups was to pick up a call and hang up immediately. It reduced the dropped call ratio, and increased calls per hour which were both important metrics. i have to believe our AI overlords will provide a better experience.
Im rethinking what I said after reading both comments from JK and Bret.. The empathy metric of customer satisfaction, in all probability, is outdated especially given that CS went for a toss after the pandemic. Before the pandemic it maybe made sense and few companies who were really geared towards giving meaningful customer support would track these metrics. Now - Im not so sure - so yes - it would be (to borrow Bret's words) a low bar, especially after all the conversational skill GPT4o demonstrated.
Bret to your point on oversight - I can tell for sure that there used to emphasis even before on what human agents would be talking to customers. This was done by certain tools that Five9 and other systems would provide where call/chat transcripts were provided on demand... But the volume of those transcripts were intense, so my product team had worked on a few projects with the CS teams to help them get more insight. Salesforce, Five9 did not have the kind of capabilities were looking for.
Im assuming now all of that has changed and its easier for QA's to derive insight from multiple chat/call transcripts at the same time even the interactions held by AI bots. There are still going to be issues inherent within the scope of what we see today with LLMs but it should eb solved quickly
I don't think the empathy really matters. For human customer services agents, its mostly faux empathy anyway. An AI can just appear empathetic.
Yes. I think one of the reasons why I also brought that up was it may get tough to (ahem😁😁) really expalin issues to an AI bot to get refunds/returns from lets say Amazon or Walmart once these are implemented. Its a just a theory.
Another masterful edification!! We as readers should be truly appreciative, for your talent is rare and is most luminous in the present moment.
Thanks Greg. Appreciate the kind words!
Thank you Greg, appreciate your kind words. Glad you enjoyed the post.
Very interesting read. Customer support everywhere is broken and needs a big fix. Hope AI can provide that. Although, Im not full sure about how capable AI is for prime time. I watched the video demo and I can already see some minor errors and the conversation getting extended for no reason. Still, Im optimistic ChatGPT will solve this really quickly.
Thank you Sully, glad you enjoyed the post. I fully agree that customer support is fundamentally broken and perhaps a big part of that is how it is treated as a "cost center" as the customer has already been acquired. I agree that we are still a couple if not few years away from AI taking prime time in customer support, as there still needs to be a lot of "hand holding" by human beings in the interim period, especially as the effectiveness of these AI customer agents depend a lot on the quality of data. But, I am optimistic long-term when it comes to genAI elevating customer satisfaction as a whole.
Thanks Sully. Im curious about some of the mistakes that stood out for you in the ChatGPT video.
Very interesting and enlightening perspective on the matter.
I was pondering about a couple of other dimensions...no real answers at this time...will have to see how things play out over the medium to long term.
1. How will AI in customer service change the growing services export industry in countries like India, Mexico and Brasil? I would think that hiring in these industries in these countries will drop as AI driven automation rises.
2. Will AI result in more nearshoring and onshoring into the US which at this time is driving almost all the AI innovation and will likely be the owners of these AI platforms?
Such an interesting time to be investing and tracking these business developments!
Indeed it is an interesting time. There is just so much to keep track of!
1. My view: AI is disrupting any kind of IT and BPO off-shoring project. With the Philipinnes, the impact was immediately felt and lawmakers last year itself have tried to take measures. This was from last year: https://www.pna.gov.ph/articles/1200928.
But I suspect its not going to be something that will sustain because those that have figured out the right mix of AI capabilities in CS & CRM departments while also assessing the customer satisfaction and ROI from the project will migrate CS workflows onto the AI platform. The others are still possibly working on PoCs or may be still assessing how to deploy AI but may move to reduce CS costs further, in that case, they actually will end up offshoring more CS workloads to Manilla or Mumbai etc. So in the interim period they wont be hit as much. Im was also trying to track any possible data such as bankruptcies etc of BPO Outsourcing companies in Manilla but cant find much.
2. AI could possibly result in near/onshoring not just in the US since US has currently outgrown any other hub around the world. Im not fully certain of this yet, because there exist many other AI hubs as well around the world like Paris, Israel, India etc... For example, I ve read MSFT and GOOG are offshoring many jobs to India and Brazil, even some AI jobs.
Excellent article—you are right on. I’d like to add one comment that colors the entire CS industry. Companies have two, or more, accounting lines in their business—COGS (cost of goods sold—including DevOps for software companies—I’m lumping CAC and COR here together) and operations/overhead expenses OpEx (internet, phones, computers, etc). They have to decide where customer service/support goes. Most decide it is OpEx and therefore as you describe. A mostly fixed cost regardless of the number of units sold and doesn’t drive revenue. This accounting report methodology alone makes CS vulnerable to cost cutting as cost is the only metric that matters in OpEx. If customers purchase because there is high quality customer service, there is no method to measure the revenue increase—so no credit to the excellent customer service team for their revenue growth impact. They are just an overhead expense. Unless they have an executive who knows this and is a tenacious blankety-blank and coordinates with sales to promote the customer support team’s excellence. The trend is those executives are fading away and customers are losing out.
The best strategy is CS as part of GPM—gross profit margin—and therefore as revenue after support to service customers. CS then has to become part of development and QA/QC—higher quality products will lead to higher revenue—post CAC customer acquisition costs. This will not come from just an AI agent—a company needs well trained support staff to lead the development work integration. AI will be a powerful tool but not a complete substitute for a wholistic support process which generates revenue through high customer experience on many fronts.
The customer support software companies salesforce, five9s, etc. have pushed for integration and automation for the past decade and this part is moving quickly towards AI which puts them at some risk—as you point out.
Thanks very much Dean! I actually did not know that organizations could also report the CS cost center under COGS. But your reasoning does make sense if they sell CS in the premium attach rate to the core product purchase. So, for example, Apple sells Applecare as a premium service. In your experience does that get recorded under COGS line item?
You're correct in asserting that customer support was ripe for AI intervention. It will be interesting to see if other companies follow Klarna's example and, if so, how well that plays out over the long-term. For me, there's nothing like interacting with a human -- even if that "human" is a highly-sophisticated chatbot. Utlimately, however, I don't care who (or what) I talk to so long as I get my problem taken care of quickly (which includes another KPI - First Call Resolution).
Thanks Paul, I agree with you on the matter that I don't necessarily care whether I talk to a human or a bot as long as my issues are resolved. And given that genAI models are "goal seeking" in nature, I believe they might actually be more effective over time to drive better customer support outcomes. Having said that, it's issues that fall outside of defined parameters for which the AI agent may not have sufficient context to produce a response that could be a hurdle, but my thinking is short-term, there will be a lot of "hand holding" with human supervision, while over time we see roles such as chatbot developers, conversation designers emerge (something that the Intercom report also hinted at).
Having worked in this space as a marketer, I do hope there is room for skilled human agents. I've seen firsthand some of the crap those folks have to put up with and I consider them heros. I've also seen the empathy that an agent expressed when dealing with a customer that no AI could even begin to emulate. So much so, that sitting there, I started to cry. I gave that lady a hub when it was over and said, "You are amazing." She was.
CS is truly the department full of unsung heroes. It is a pity that the work is often considered "tactical" and not attributed directly to revenue growth, when delightful customer care experiences actually result in repeat purchases and deepening adoption.
Right on, Paul! I will be curious to see how I can make the case for a return/refund to Amazon for a situation that AI may not be trained for. Lets say my Fitbit strap broke on the 31st day after purchase and I was trying to make a return/refund but AI may not understand that because it falls outside the 30-day grace period. Sometimes, I can convince human CS agents to take care of that for me.
Also you are right about FCR.. Just curious Paul - do you have any experience in working with CS/CRM folks in your work? Not many outsiders know about FCR as a CS metric😉
I worked as a digital marketing manager for an outsourced BPO company for a couple of years, and have some other experience writing about topics in the space.
Hey Amrita and Uttam, a great post, as always. I'm wondering whether sooner or later there'll be a company that'll make it a mission to help other companies streamline customer support? Just brainstorming :)
There are tons of those companies already who help streamlining customer support operations. Salesforce, Five9 and others were from 10 years ago and promised a certain level of streamlining of operations. Those streamlining benefits failed as we moved through the pandemic which is why there still is elevated levels of perception of decreased value - longer wait times, less empathetic human CS agents, more confusing ways to speak/chat with agents etc..
There are startups like Dialpad and Intercom which have started taking a lot of market share from legacy providers and they provide almost end-end solutions to in manage CS operations....
...The question now is that enterprises and basically, anyone, looking to add customer support for their customers aren't looking for just streamlining anymore... they're looking for productivty and the subsequent cost benefits.. Like in Klaarna's case - they were supposedly able to reduce their CS costs by $40M and yet increase the volume of interactions with their customers.
Thanks Uttam. Interesting comments.
Software is eating the world - more and more of it. Great read!
Thank you Michael, it will be interesting to watch how traditional SaaS business models adapt in the coming quarters and years, as companies do more with less.
If software eats world, does AI eat software? 😉😉
Well. It depends on your perspective. I view AI as software :-)
Customer support is usually one of the sore points of all companies: they assist you and are around you before you become a customer, once inside their network you are just a number to manage (and a burden). AI could set new and better standards, good point.
Fully agree with you Mirko. I think AI agents can actually get better at resolving customer issues faster and more effectively. Interestingly in the Intercom's report, support teams were asked what new roles they anticipate being created by AI and the answers were Chatbot developers, Chatbot analytics and conversation designers.
Yes. Very true. I've worked with Customer Support teams in the past so Im very aware about the business and the metrics they follow. While researching for this post, I do see that there is a new range metrics, standards and KPI's that are being used such as number of tickets managed by bot, bot handled time etc... After all we had to endure through the pandemic with 1 hour long wait times.. im excited to see the things that AI could do.
Thank you for explaining the customer support operation in details. I certainly feel the threat for the customer support software companies as their business model is upended. I hope they will come up with ways to reassure the customer's value. AI is eating software! But our life can be better with it than without!
Thanks Marianne. I think it is most software companies whose business models will be under pressure especially with their "seat-based " pricing model, leading to lower ACV and thinner margins. Having said that, I think large, well funded and fully-integrated software companies are probably better positioned than point solutions in this decade given increasing vendor consolidation.
As for businesses, many of them are taking Klarna's route to build AI proof-of-concepts across use cases, such as customer support and deploying it over the course of multiple business cycles while assessing the ROI on these projects. The irony with customer support has been that it is treated as "cost-center", as the customer has already been acquired and especially after the pandemic, the quality of CS has degraded across organizations as they looked to optimize their cost structure. So, I am quite optimistic that given the goal-seeking nature of genAI models, we can actually arrive at a point where we have "cost-effective" agentic customer support that actually improves the overall customer experience.
Thanks for these thoughts!
Insanely insightful again. Your posts are becoming a regular weekly read for me - I won't miss a single one.
Thanks v. much Daan! Glad you found it insightful!
Interesting as ever but lots of slightly irritating typos, which give the impression of it having been rushed.
Apologies for your experience. We will be more attentive to details moving forward. Thanks for letting us know.
No worries Amrita & Uttam. I thought you would like to know. As I said, still a fascinating post 👍
Thank you for alerting us, Allan. Somehow the grammar check tool completely skipped the 2-minute summary. We've made the changes for you. Apologize that you had to go through the experience.
Amazing read and insights! Regarding BPOs and the potential threat from chatbots, they have a window of time to adapt. By embracing AI and automation as enablers rather than threats, they can thrive, and some BPOs already began this transition years ago. And as AI continues to advance, they will shift towards offering high-value-added services while improving margins.