Intercom on Product: Our manifesto for next-generation customer service

A lot has changed since we started Intercom over a decade ago.

The iPhone, first released a couple of years before, was making rounds across the industry, introducing a new standard for smartphone design that has since been replicated by countless competitors. Facebook Messenger had just been released, and social media platforms like Facebook and Twitter, as well as messaging apps like WhatsApp, were rapidly gaining popularity. In the years that followed, as we now know, messaging exploded.

Intercom was born out of a desire to make internet business personal. We wanted to move away from spammy, asynchronous emails or cookie-cutter replies and towards personalized, genuine connections. Throughout the years, we’ve seen a lot of what we believed in become real, and we’ve certainly worked hard at pushing the needle that way. But we’re not much for resting on our laurels. It’s time to start thinking about new ideas – not from the past decade, but for the next one. And that’s where our new manifesto comes in.

In today’s episode of Intercom on Product, Paul Adams, our Chief Product Officer, and I sat down to talk about our new product manifesto for the future of customer service and what it means for us.

Here are some of our favorite takeaways from the conversation:

  • The future is omnichannel: although messaging has become the dominant mode of communication, channels like phone and email are still key in many settings.
  • Pairing humans with AI will lead to better and cost-effective experiences. Bots excel at dealing with simple, repetitive queries, freeing up reps to do more rewarding, complex work.
  • The world is filled with proactive support examples. We just have to imagine what they look like in software – from onboarding checklists to timely tooltips and banners.
  • Over the past decade or so, especially in this subscription economy, traditional boundaries between sales, marketing, and support teams are increasingly blurred.
  • The best approach for ensuring a smooth orchestration of all of these components is to connect all data, content, and workflows within a single platform.

Make sure you don’t miss any highlights by following Intercom on Product on Apple Podcasts, Spotify, YouTube, or grabbing the RSS feed in your player of choice. What follows is a lightly edited transcript of the episode.

Why manifestos matter

Des Traynor: Welcome to Intercom on Product. I’m once again joined by our Chief Product Officer, Paul.

Paul Adams: Hi.

Des: And today we’re going to talk about manifestos and what they mean in product. The core idea here is that Intercom’s been around for 11, 12 years.

Paul: Over a decade, yeah.

Des: And things we used to proclaim about the future have, over the course of those years, turned true. And as a result, we need new shit to say. That’s what we’re talking about today. Let’s start off with the idea of the manifesto itself. What is a product manifesto to you?

“The manifesto, as opposed to a vision, is a set of things you believe about the future and things you think will come to pass”

Paul: It’s an interesting question because people talk about visions and missions and manifestos and goals.

Des: There are a lot of words, yeah.

Paul: For me, a manifesto is really powerful. And the manifesto, as opposed to a vision, is a set of things you believe about the future and things you think will come to pass. And you may have a role in helping them come to pass, of course. But there are things you believe in. You can see trends.

Des: They’re possibly independent things, as well. They can happen independently of each other; one of them doesn’t cause the other.

Paul: Exactly. And I think that’s why they’re different from a vision. A vision for me is something that’s really concrete – consistency and coherency matter in a vision. That’s how our product would look in five, ten years. Whereas in a manifesto, it’s “We believe X and Y, we’re going to enact on those things, but they don’t necessarily come together.” It’s kind of TBD.

Des: You’ve been at Intercom for about 10 years. When you joined from Facebook, one of the things you did early on in our old office was put up these posters that declared certain aspects of what you saw happening. One was “All businesses will become internet businesses.” Another one was that the future of customer service will look more like the past than, say, the present – it’ll look more relationship-based and less transactional-based. You got that idea, I assume, from Facebook. Is that right?

“It just seemed that this Google Analytics era was a temporary moment and that software would catch up with humanity”

Paul: Yeah. When I was at Facebook in 2010 or 2011, it was very different than today. It was much smaller, and there was a really clear sense of purpose. Zuck and the leadership team were fully promoting the idea that the world would be better if it was connected. Connecting people means you connect ideas, and when you connect different ideas together, you get this kind of one and one equals three dynamic. And obviously, things play out in all sorts of different ways, but the ideas at the time were core. Everyone believed in them. It was exciting. And Facebook did a fantastic job, at the time, of creating beautiful artifacts for the walls, quoting people like Marshall McLuhan, really inspiring stuff that brought them to life.

Des: And then when you came to us, we created a set of what we believed would be the things that would happen over the next few years. What were some of them?

Paul: Yeah, back then, as you said, one was this belief that the future’s going to look much more relationship-based. If you look at the kind of commerce going back decades, if not hundreds of years, all commerce is relationship-based. And at that time, things like Google Analytics were in vogue, and that’s a great tool, but not relationship-based. It’s a very transactional and analytical tool. And it just seemed, to me and other people at the time, that the future would look much more like the past, i.e. relationship-based. That this Google Analytics era was a temporary moment and that software would catch up with humanity. Now, it’s not controversial to say that you think business will be relationship-based.

Once in a generation change

Des: What about communication? At the time, we were seeing the rise of messaging, right?

Paul: Yeah. I think the same year that Intercom was founded, WhatsApp was founded, iMessage came out…

Des: Facebook Messenger. Snapchat maybe.

Paul: Yeah. And there’s definitely a connection with the iPhone. Suddenly, people had the thing in their pockets and could message you. But messaging was the thing. And again, I was probably influenced pretty heavily by Facebook. Facebook was betting on messaging. And so, when I came to Intercom, it made sense to think that business messaging was the future, too. And again, that’s no longer controversial. Look at Apple Business Chat. Everyone’s getting WhatsApp for business. Everyone’s past the point of believing that that’s the future – hence us needing new opinions.

Des: That’s what I was going to ask. Obviously, you can have all these opinions, but opinions do kind of naturally expire. Sometimes, they expire because they’re wrong – no one’s talking about gamification or whatever. And other times, they expire because they just become woven into the fabric of the internet. And a third reason is that technology moves on, in a sense. Let’s say you had an expense-tracking app before the iPhone was launched, and then the iPhone comes out with an app store and a camera, and all of a sudden, you can expense something by taking a photo of the receipt. If you’re in the expense-tracking category, your entire world has changed. Your entire product roadmap should get thrown out because the capabilities are groundbreakingly different from what they were a couple of weeks ago. What happened to our original ideas?

“Does the world happen to you, or do you happen to the world?”

Paul: I think all of those, actually. With things like messaging, technology and society evolved. Messaging is the future. At the same time, though, another opinion we had at the time that I think we were wrong about was that we underplayed the importance of other channels. We thought messaging would replace email or phone. Now, new types of channels came in, like WhatsApp for Business and so on, and we’ve come to realize that the future looks omnichannel, whereas, in the past, you might have said it was very messaging-heavy. I still think, for what it’s worth, that messaging will be the dominant channel in the future, but omnichannel will be the thing. And so, some of them did expire. Society has changed, and society keeps changing.

Des: Expectations of what people want out of a business are rising.

Paul: Absolutely. Ten years ago, all the apps on your phone were not brilliant, beautifully designed apps, right? They were the first versions of everything. Whereas nowadays, anyone can open their phone and there’s Google Maps, Uber, and all these apps are beautifully designed, really well-built, architected, fast, seamless. That’s definitely evolved pretty significantly since we started.

Des: And maybe this is too recent of a thing for us to quantify how big it’ll be, but the advancement of AI and automation has obviously presented a lot more opportunities for rip waves in the industry.

Paul: Yeah, absolutely. Is AI over-hyped right now? Maybe. I tend to believe it’s going to have a foundational impact on so much of our general life. It might be a little bit over-hyped at the minute. The big thing for me is: does the world happen to you, or do you happen to the world? And I think that lots and lots of businesses, in the last three to six weeks, or maybe back to when OpenAI launched ChatGPT, suddenly realized, “this technology is: a. pretty transformational; and b. accessible.” Suddenly, if you’re not on top of this, the world’s happening to you. And if the world’s happening to you, you need to get a manifesto, get on the front foot and start to make your product so that you’re happening to the world.

“The idea of emailing a ticket and there’s this asynchronous, slow-moving back and forth – customers are way beyond that now”

Des: The other thing that changed for us is that late last year, we made the decision to go all-in on this area of customer service. Even if the other tectonic shifts hadn’t happened, that would’ve forced us to have a kind of reset: What are we doing? What do the next 10 years look like? And where to from here? What’s your belief about the future of customer service?

Paul: Again, this is going to sound big. I think customer service is undergoing a once-in-a-generation change. And AI is certainly going to play a big role in the future. But even prior to the AI breakthrough we’ve seen in the last few months, customer expectations have changed, technology has changed, channels have changed, and customers expect a lot more from businesses. The idea of emailing a ticket and there’s this asynchronous, slow-moving back and forth – customers are way beyond that now. They expect a now economy because they have it elsewhere. If Amazon can ship you something overnight or in an hour, why can’t the business get back to me in that same timeframe? Whether it’s via automation, bots, people. Expectations have changed. And I think it’s actually going to be better for businesses and customers when we get through it.

Des: Whenever people talk about generational changes, or even, honestly, when ChatGPT launched, my go-to question, no matter the technology or the societal breakthrough, is: does this make it cheaper, faster, easier, and more accessible to use this technology? Can you use it more? Can you use it more places? Will you use it more often? Whenever you’re trying to bring it back to software, I always come back to those questions because, ultimately, the jobs customers are trying to do rarely change. It’s just that the way in which they do them changes an awful lot. Most businesses want to support their customers by giving them fast, high-quality replies that address their issue and ultimately render the customer at least neutral, if not happier, as a result of contacting them. What are the changes you’re seeing if you were to try and break this generational shift down? What does it look like?

“A lot of customers aren’t really hoping they get to talk to a person that works for the company – they just want the answer”

Paul: I think it looks like a few things. Something which will be better for businesses and customers looks like speed – it’ll be faster for customers and faster for businesses, and that will be through automation. We’ve had automation for a while in customer service. You’ve got rules, and you’ve got efficiency improvements like macros and things like that, and some things can trigger other things. It’s a whole new generation of technology that will make it extremely fast. You’ll have immediate answers from bots. Sometimes, they’ll actually deliver the answer right there and then. And it’s better for the customer, in most cases. A lot of customers aren’t really hoping they get to talk to a person that works for the company – they just want the answer. And for the business, it’s just way more efficient. If you talk to a customer service rep, they’re not saying, “I love answering the same question over and over and over again every day.” It’s a win-win.

The quality of the experience will improve because of the speed, and it’ll just be cheaper. It’ll be a lot more operationally efficient for the business, and it will certainly reduce costs. That means they can serve more customers with fewer people. It means that the people who are working in the customer service team get to do different things, not just answer the same thing over and over. They can start designing bot experiences, working on harder problems, and building relationships with important customers.

Human touch meets automation

Des: Let’s get into it. What do we believe about the next 10 years?

Paul: We have a new manifesto and we’re pretty excited. There are four parts to it. The first part of the manifesto is on automation and humans. We believe that the future is both automation, which is a combination of bots and AI, and human support; people talking to people. The “and” there is really important because we believe one of the biggest things here is the bridge between the two. When the automation side doesn’t have the answer, which will happen, they will seamlessly hand it off to people. That’s an end-to-end experience that you can design and orchestrate. That’s a big part of it for us.

Number two is proactive and reactive. The history of customer service is reactive – businesses trying to manage workload, incoming volume, just trying to manage it as efficiently as possible, put people in a queue, deflection, and all that stuff. We think the future looks proactive as well as reactive. Again, these things will work in combination with each other. And proactive means you predict what customers are going to ask based on what’s going on and all the data you have, and then you give them the answers before they’ve asked it.

Three is conversational and omnichannel. We talked a lot about this already. Conversational has to do with messaging as the primary channel of customer service, and omnichannel means it’s not the only channel. Sometimes phone is the best, sometimes email is the best, and sometimes there are back-office ticketing requirements where a question has to go through an org and a team, and so on. We think it’ll be conversational first, but all the other channels will work together.

“That’s what I see in this world of support. I think you’ll have humans ultimately controlling the intelligence that the AI works from”

And the last thing is a single, seamless platform. We believe that today, you have a lot of customer service tools that are quite siloed, and there are pretty complicated integrations to get them to work together. We believe the future solution will be a single, seamless channel that can do the top three things we mentioned above. It just will create a much better and faster customer experience.

Des: Automation alongside humans. What is our thinking there?

Paul: Our thinking, at the highest level, is that the vast majority of customer questions in the future will be answered without needing a human. That’s number one. The second part is that sometimes, you won’t be able to do that, whether it’s because it’s technically impossible or it’s preferred to talk to a human. So automation and humans work together.

Des: There’s this idea in AI of an augmented machine, as in human plus machine is better than a machine, which, in turn, is better than a human. That’s what I see in this world of support. I think you’ll have humans ultimately controlling the intelligence that the AI works from. I see a future where the customer support team is answering every question for the first time and the last time in parallel. The first meaning the first time the AI didn’t spot it, so we couldn’t configure it. And the last meaning that once the support representative is finished replying, the AI might pop up and be like, “We hadn’t heard that question before. Am I right in saying that we don’t charge for that?” And you say, “That’s correct.” And it says, “OK.” And that goes into its own intelligence such that you’ll never see that question again. I like the idea of the support team being the team of first and last. That way, the support org takes a step up in the value chain to actually being in control of the entire set of business intelligence.

“Implicit in the ‘treat our VIPs with human service’ thing is an assumption that human service is better. And it’s not guaranteed to be better”

The other side is that there is a world where the customer’s and the business’ desires might be slightly at odds. And it’ll be interesting to see how we have to design for that. In practice, we’ll probably end up having to follow whatever the business prefers. But you can imagine a world where I don’t want the artisanal hand-typed reply, I just want the answer, whereas the business might be like, “Oh, this is our one chance to talk to Des. We should take this opportunity to see if we can upsell him.”

Implicit in the “treat our VIPs with human service” thing is an assumption that human service is better. And it’s not guaranteed to be better – it’s definitely better in some queries and worse in others. If it was a race of who can explain how to reset your password quicker, the bot’s going to win. If it’s a challenge to deal with an emotionally-loaded situation regarding a late flight and missed luggage, chances are the human’s going to win by a substantial margin. It’s for that reason we want them both together.

Paul: Yeah. And it’ll differ from business to business. One thing we hear consistently when we talk to customers is most customers don’t care – they just want the answer. They’re not looking for the bespoke, they’re just like, “How to reset my password? What happened to X? How do I do Y?” And in-product helps us do that. They’re delighted and hence better satisfaction and engagement.

But sometimes, and increasingly in this automated world, businesses want to differentiate on human service. In other words, “Hey, with us, you get the human service.” But sometimes, there’ll be a mismatch. Customers don’t value that. And I think we’re going to have to work this out. People will learn as they try and start hearing from customers, “Hey, thanks for talking to me, but I really would have preferred…”

Des: The bot’s quicker. And honestly, one of the biggest advancements that ChatGPT might have caused is a societal acceptance of bots being useful. Historically, the chatbot was seen as, at least, a second or third-class citizen, and now I see it taking a step up the chain.

Paul: A big thing here is businesses want to build relationships with customers. I guess the known best way to do that is to talk to people and build a bespoke relationship as best as you can, people-to-people. And a lot of people see bots, especially AI, as transactional and not relationship-building. And I think that’ll change. You’ll see these augmented experiences, and when the bot talks in the brand and tone of the company, they’ll actually feel like a part of the brand.

Proactively engaging customers

Des: The second piece is this idea of proactive and reactive. The reason this resonates with me is that my first job, believe it or not, was in a gas station. One day, some customer was complaining about something like the carwash not working, and the first thing the manager said was, “Thank you.” He was so effusive in his thanks for the complaint. And I was like, “What’s going on there? The dude’s shouting a lot of shit, and you’re here saying thanks.” He said, “Des, most people don’t complain. For every guy who comes in and says that, 800 people just drive off. We’re going to fix the carwash and sell a lot more stuff. Whenever you see those things go wrong, always get ahead of them. If you see someone struggling, get in there because you’re going to cost me a business opportunity otherwise.”

I didn’t realize that 25 years later, I’d be sitting here talking about proactive support, but that idea sticks. If you have a reactive-only support strategy, you’re just waiting for people to notice that something goes wrong, care that something goes wrong, and contact you that something’s gone wrong. If you have a proactive support strategy, instead of waiting for things to go wrong, you ensure everything goes right. We used to say it’s the difference between preventative medicine versus the emergency room, right? You get out ahead of the problem.

Paul: Well, just as you were talking, it made me think that proactive and reactive service is a bit like an iceberg, where the support team under pressure only sees the top of the iceberg, the visible part. That’s all the stuff you get. What they don’t see is the invisible part of the iceberg – all the people leaving without asking, all the things you just said. And they’re actually the opportunity. We think that by building tools, people can be proactive, recognize all these things, and start dealing with them ahead of time.

“When we talk about proactive support in customer support, it feels quite new and hot, but the world is a museum of proactive support examples”

The other thing is that a lot of times, in technology, there are these before and after moments. ChatGPT was one of them. Before, you didn’t think about it much, but then you use it, and suddenly, your world has changed. We’re old enough to think about using Wikipedia for the first time. It was just mind-blowing. Oh my God. Or, you know, use Google Search for the first time. It used to be the case that you typed in Google Search, and nothing would happen. You hit the search button, and it comes back – it’s reactive. And suddenly, one day, I started typing in Google, and stuff starts going down below.

Des: The auto-complete, yeah.

Paul: I was like, “Whoa.” OK, now we’re in this whole new world. It’s still reacting, but it’s much more proactive.

Des: Yeah. It’s guiding you to a better outcome, basically.

“One thing we’ve seen happen over the last decade is teams are blurring boundaries”

Paul: Totally. And now, we’re in a world where you open Google Search, and there’s already stuff there – the weather in your location, the things you search for, news articles. I get loads of random football things because they think I’ve come here – and oftentimes, I have – to ask these questions.

Des: When we talk about proactive support in customer support, it feels quite new and hot, but the world is a museum of proactive support examples. If you walk into the courtyard of a car dealership, through the front door of a hotel, or into a restaurant and stand around long enough, you’ll see proactive support. People are like, “How can I help you? Clearly, you need some help.” If you stand scratching your head in front of a television in a store, people will be like, “Oh, this dude probably wants to buy a television. I’m going to come in and help.”

The world is just literally full of it. You have to imagine the equivalent of that inside a product, which looks like, for example, someone hovering their mouse over a button that says merge. They’re like, “What does this merge?” And the reality is proactive support looks like things like a tooltip that pops up and says, “‘Merging will combine the two products,” or projects or whatever the appropriate thing is in this hypothetical product. Other versions of it are, let’s say, a banner telling you your credit card is about to expire, a popup to tell you about a new feature you haven’t used, a checklist to tell you about the stuff you haven’t done yet. They’re all examples of what we would consider proactive support to help you ultimately get to the best outcome. It’s a new stance for software, but I think it’s inevitable given how the rest of the world works.

Paul: I think so too. The hotel example is brilliant. Because some people might say, “Well, Des, a lot of that is sales, not service.” And again, these two worlds are very, very interconnected. But the hotel one’s great because you’re already spending the night there, you know? That really is true proactive support.

“If you’re supporting somebody on day 14 of a 15-day trial, you’re also selling them. If you do bad support, they don’t buy. If you do great support, they buy. How is that not a huge part of the sales role?”

I think it’s also going to be really big in the future – bigger than we realize today. One thing holding it back, when we talk to people in customer support teams, is that they’ll say, “We don’t have permission to send messages.” Who puts the banner live? Typically, marketing. I think that all of that world is blurring together. Back to the manifesto, one thing we’ve seen happen over the last decade is teams are blurring boundaries. It used to be the case 10 or 20 years ago that sales, marketing, and support were quite siloed, transactional, marketing hands leads to sales, et cetera. These days, they’re blended. You have teams like success, growth, and product in there too. And so, I think it’s going to become the norm that a lot more customer service departments will either have success teams within them who send out messages, or they’ll just start doing it themselves in all sorts of different ways.

Des: Yeah. And I think the idea of all support being sales is very true, especially in this subscription economy that we’re in, where if you’re supporting somebody on day 14 of a 15-day trial, you’re also selling them. If you do bad support, they don’t buy. If you do great support, they buy. How is that not a huge part of the sales role? I think we will see that blurring of the lines. Even if somebody’s an advanced user in year two and you launch this new fancy feature and they’re struggling to use it, you’re either selling them to use that feature, if you can be proactively supporting them, or you’re ignoring them and letting them not use that feature and thus have less engagement than they could have had or as the optimal level with the product you’ve built.

To me, of all the areas in the manifesto, this is the one that is going to be most emergent. AI is obviously going to run, whereas I think this is one that will be a combination of us making it happen and showcasing examples of where it really does transform a business.

Messaging first

Des: The third area is this idea of conversational combined with omnichannel. Could you say more about that?

Paul: Yeah. Let’s go back to the start again. People talk and communicate with each other in all sorts of different ways. Messaging, phone, email. And the big change we’ve seen in the last 10 to 15 years, especially with smartphones – iPhone and Android are now obviously commonplace in everyone’s pocket – is the emergence of messaging. Messaging has taken over as the dominant way our species communicate. We forget about it, but it’s actually nuts that, in the last seven or eight years, messaging has become the dominant way humanity communicates. And it’s already happening in business. It starts with live chat, which is basically messaging, and now these messages can do much more than chat. It’s taking over.

“Humans communicate in multimodal ways, and a support tool, to allow the richness of human communication, needs to be multimodal too”

That said, sometimes other channels are important too, and sometimes they’re better. Sometimes, you need to get on the phone. Sometimes, if someone isn’t available for a message or something’s happening, email can be better. Or if something has to go through a back-end system that’s email-heavy, email can be the best channel. These will persist, but it will very much be conversational and messaging first.

Des: I think there’s probably some Marshall McLuhan-type theory. New messaging channels don’t kill off other ones, they just kind of sit on top of them. You can have an SMS conversation with somebody, but if it’s going to get more formal, even with friends, like when they invite you to their wedding, it’s an email, not a WhatsApp. When the wedding invite gets real, and they’re definitely going to get married, it turns into written form. It’s almost like you have this spectrum or hierarchy of formality. Then, when you also have this other thing about urgency and emotion, that’s when people reach for the phone, “This thing is blowing up, we need to talk and use more than words” – intonation, sympathetic tones. Humans communicate in multimodal ways, and a support tool, to allow the richness of human communication, needs to be multimodal too.

“Messaging can be lower bandwidth, but messaging is way more flexible. People can message from anywhere”

Paul: People who talk about different communication channels and how people use one or the other, or prefer one or the other, talk about bandwidth all the time. Face-to-face is the highest bandwidth. You’re getting everything – what I’m saying, what I’m doing, my gestures, expressions, what I’m not saying.

Des: Eye rolls, the whole lot.

Paul: Yeah, absolutely. The phone is the next closest to that. Because you’re getting full bandwidth really fast, because people can speak faster than they can type, and you’re getting the emotion. Messaging can be lower bandwidth, but messaging is way more flexible. People can message from anywhere. They can be in a quiet or loud place – it works everywhere. It’s omnipresent.

The power of orchestration

Des: And then, lastly, it’s a more technical idea, this belief that all these things – automation plus human, proactive and reactive, conversational and omnichannel – need to be connected by a single platform. There is a reason why these things are better together and ultimately kind of broken when they’re apart.

Paul: The thing in the manifesto that I love is the ‘and’ between each of the first three. It’s automation and human. Proactive and reactive. Conversation and omnichannel. And I think, to do that well, you need orchestration.

“When you want to put things on the same platform, it’s because they either share data or content or common workflows. And I think all things are true here”

We mentioned earlier that customer service teams’ roles will change. Orchestration, I think, is going to be a huge new emergent job in customer service teams. You’ll have things like conversation designers – people orchestrating when bots fire, when they don’t, how to hand it off to a human, how to teach it to get better, and all that stuff. Orchestration’s going to be really big.

And to do orchestration really well, you need the things in one tool. If you have one support tool that does automation only that then hands off via APIs, webhooks, and all sorts of data integrations to some other human-only tool, it’s going to be really hard to orchestrate that experience. Then, you buy a third tool that does proactive messaging. Suddenly, with the proactive thing, which is hard to do, you’re now trying to figure out how to not over-message people, or you’re going to be messaging a person who’s just had a really angry phone conversation… It’s just hard to orchestrate it.

Des: Generally speaking, when you want to put things on the same platform, it’s because they either share data or content or common workflows. And I think all things are true here. Ultimately, you want to have as few customer records as possible. Every soft platform out there is like, “We’re the single record of truth,” but in general, you don’t want to have multiple conflicting records, which would be the case if you end up with the outbound tool, the support tool, the phone tool, the email tool. It’s going to be a mess. They share a lot of content, too – when you’re explaining how a feature works, that explanation should also be in the tool clip beside it, in the knowledge base, and in the bot. You need to move towards this convergence and, ultimately, to fewer copies of the same data.

“Even if it’s all really well set up, they’re still clunky. There’s still latency and delays between things”

And then, with the actual workflows, a bot handing off to a human is going to be very hard to do if you’re going to rely on the high jinks of API calls back and forth. To be clear, it’s technically possible but practically infeasible. That’s my experience with a lot of these tools where your AI bot hands over to human support, who looks up the knowledge base somewhere else, who then jumps on a phone call paired with something else. Yes, it can work in the sense that everything can, in theory, be connected, but in practice, the people selling that tend to have a lot of botched installs.

Paul: Yeah, I agree. Honestly, though, even if it’s all really well set up, they’re still clunky. There’s still latency and delays between things. And at the end of the day, going back to one of the earliest things we said, customer expectations of service are changing and customers now expect service to be immediate, high quality, just as good as Google and all the other cool stuff on their phone. And the messaging apps they have – people are just going to get upset, annoyed, frustrated, and have lower satisfaction and engagement. The clunkiness will lead to that.

Des: And then, lastly, the other benefit is that, as products have many stakeholders in a business, it’s easier if you don’t have to multiply everything by seven to give everyone seven different logins. They can be like, “Hang on, I’ve had to jump on a call with this customer. What’s the last thing we said?” It can be quite messy if you spread that across a proliferation of tools. And that’s why we think it’s better with a single seamless platform. That’s it. That’s our customer service manifesto. And hopefully, we’ll be able to use that for a few years to come. But who knows, it’s technology. Things change fast.

Paul: That’s right.

Des: Cool. Thanks for joining, Paul.

Paul: Yep, it was fun.