There’s a surge of excitement around messaging as a new platform and the idea of ‘Conversational Commerce’. There are a number of great pieces and podcasts that have been published documenting and forecasting the new ways we will see messaging used in 2016. I recommend reading them all. What I would like to add to the conversation is the why, how, and who will get us there. This is my snapshot of why messaging is so interesting and a map to help understand the stack. There is a lot to cover, so this is broken up into three sections:
- What is Messaging and how did it become such an integral part of our lives?— part one explains what makes messaging so powerful in people’s lives today based on its product principles. I also make the case that calling messaging a singular ‘market’ is inaccurate. TLDR: your messaging conversations are your relationships.
- Messaging in 2016, what’s the hype? — part two attempts to explain why there is so much recent attention on messaging as a platform. TLDR: messaging relationships could extend beyond friends and family into relationships with services and businesses.
- What’s the current landscape? — the last section discusses the existing messaging platforms, the emergence of assistant and B2C services over these platforms, and the infrastructure necessary for this future. This section also looks at the new SaaS businesses that are coming about to help messaging achieve the hype covered in section two. TLDR: there’s a lot happening across the stack. Except for messaging apps themselves, this is still an immature market with a lot to build.
What is Messaging and how did it become such an integral part of our lives?
The first thing to understand is this:
There is not a ‘market for messaging’ anymore than there has ever been a singular ‘market for websites’ or a ‘market for email’.
Messaging is a form factor, a paradigm, and most importantly a set of user expectations that can carry across many different markets, use cases, and approaches. Before diving into the latest buzz around messaging in 2016, we should lay out these user expectations and understand what got us to our messaging powered lives. For the purposes of this discussion let’s constrain messaging to mean ‘asynchronous, long-lived, mobile-centric messaging,’ e.g. SMS, Facebook Messenger, WhatsApp, iMessage, etc. Let’s also focus on consumer messaging as opposed to workplace/enterprise messaging, which has a lot of related, but different, interesting things going on.
What makes messaging so powerful? If you look across all messaging products you quickly come up with a few baseline behaviors that have become its foundation:
- Asynchronous by default, synchronous when appropriate: this is the user expectation that messages can be sent, even if the other side isn’t there to read them right now. Yet, conversations can seamlessly escalate from a slow back-and-forth, sometimes spanning many hours, to the quick synchronicity of old-school desktop chat. As a result, we’ve largely done away with the ‘brb’ and ‘u there?’. The sender knows you’ll be there when you’re there. This behavior is fundamentally enabled by mobile — the product expectation is that a sent message will sit in the recipient’s pocket until they can and desire to answer it.
- Long lived conversations: One of the most powerful aspects of messaging is the treatment of messages within the context of a long lived conversation. This is not how it always was — old SMS clients used to have an inbox approach that split up messages from the same person and instead focused on ordering messages only based on when they came in. Email is fundamentally organized around a topic-based thread, which behaves differently than a person-based conversation. This is one major reason why messaging isn’t just ‘faster email’. The conversation is the container that is defined as between two (or more) people, not based on topic. In fact, much like real-life conversations, topics change constantly. These long lived, multi-topic, conversations are what make messaging the best digital representation of your relationships. If you ask someone “pull out your phone and show me where your significant other exists on this device?” the answer is not in the address book, email threads, or their social profile, it’s in the messaging conversation.
- The conversation list: If conversations are the powerful container of your relationships, then the conversation list is the hub of all of the relationships in your life. The list is quite simple — the conversation that had most recent activity (inbound or outbound) is also the one that you’re most likely to hop back into, so it’s up at the top. This simplicity is powerful — I don’t need to signal to my messaging apps who is most important to me today. The most important conversations naturally live at the top. Conversations that are days old fall down the list gracefully, but can be resurrected just as easily. So why is this conversation list even important? Aren’t those recent conversations also generating notifications on the same device? Three important distinctions between the message and conversation paradigm vs the notification and notification center paradigm:
- The conversation list is not ephemeral — viewing or responding to a message does not get rid of its position (in fact it can keep its position high). Conversely, notifications, evaporate as soon as they are dealt with.
- It represents both inbound and outbound activity — notifications are a one way street. In fact, to deal with the barrage of inbound notifications, there need to be better ways to triage, prioritize, and snooze out the noise. There have been startups that have tried to help with notification overload.
- Conversations are mostly from people you care about. This, combined with the above two behaviors, means that the conversation list is a great approximation for ‘who do I care about talking with right now’.
Asynchronous messages, conversations, and the conversation list form the basis of mobile messaging. It is the combination of all of these behaviors and expectations in messaging that make it such a dominant part of people’s digital lives. It is what makes messaging the most comfortable communication medium ever invented.
I believe comfort, not convenience, is the most important thing in software, and text is an incredibly comfortable medium. Text-based interaction is fast, fun, funny, flexible, intimate, descriptive and even consistent in ways that voice and user interface often are not. — Jonathan Libov
The comfort of messaging has had real consequences in making it the most engaging and popular activity on our most personal devices. Even just looking at the current top messaging apps, which do not cover all messaging behavior, we can see an engagement that is so far beyond anything we’ve seen in any app category.
Now we are on the same page about what messaging is and its dominance in our connected lives. However, most of these foundations were established in 2005–2011 (SMS growth, WhatsApp, iMessage, Facebook Messenger, WeChat, Hangouts, etc.) and have grown tremendously since. So, what is the new excitement in 2016 about?
Messaging in 2016: What’s the hype?
Over the last year a number of factors tipped that caused an industry-wide awakening. I’d summarize this awakening in one question:
If messaging is so useful and so popular…why do you only message with your friends and family?
Why can’t I message my bank to get a bill paid? Or message the department store for advice? Why not the airline, or the local bakery? If these channels open up to allow for messaging businesses and services, how does it change my relationship with them? Will I now use something like a travel agent, when I’ve never used one before? Will it change how I communicate with insurance companies, doctor’s offices, and department stores? With therapists and non-profits?
Here are the factors driving the recent excitement:
Distribution and ubiquity
Messaging, like any new platform, shifts the power dynamics around distribution. It turns out that the primary platform of today, mobile apps, is really, really hard to get going, and these challenges have been well documented. First you have to develop an app for one platform, then get it approved, then try and get people to download it, and then get them to enable a bunch of permissions. If you manage to get a lot of people to do that, then you get to do it all over again for the other platform. If you feel you’ve gotten that far, and you want to grow usage of the service, you can then spend a crazy amount on a mobile app install ads and try to play games in the app store. If successful with distribution you just get people to the top of this drop-off-prone funnel.
Messaging changes this: instead of mobile platforms, app stores, installs, notification permissions and the app install ad unit, we’ll have messaging platforms, conversation start buttons, and ads to start conversations. Instead of building native apps for many different platforms, each messaging platform you support gives you an always-on, many-screen presence in a user’s life. In the long run, being found by consumers will continue to be hard, and while there’s likely a first-mover advantage to being early on a new channel, it’s likely going to erode quickly, as it did in the app stores of today.
Enabling technology: Context, identity, and AI
Each new platform brings with it a set of assumptions, largely based on the functionality of the platforms that came before it. If you ever wonder why mobile OSes aren’t more aware of your social graph and identity, it’s largely a matter of timing — they developed alongside each other. If the first iPhone came out today, social identity would be less of an afterthought. It would permeate throughout the entire operating system, from home screen to sharing to notifications. Messaging has come about in a world that can carry with it all the assumptions of both mobile platforms (always-on data, location, camera, security, payment integration) and social platforms (strong representation of identity and graph).
Messaging is essentially the child of the social and mobile platforms and can bring with it the best from both.
The other enabling technology that is relevant now is the potential to apply conversational AI. Messaging as an interaction medium lends itself very well to blended human-machine systems. The input and output to a user are both text, so you can skip the complexity of voice parsing. More importantly, the async nature of messaging allows the opportunity for handling each response in a way that is dependent on the level of machine understanding. In other words: by talking with Siri, you quickly can tell Siri is a machine. It would be foolish for Apple to try and act like there’s a chance Siri is even part human. What if you were texting with Sherri though, and some messages were sent by a human and some by an algorithm? If the combined system of software+agent ensures that a user sent message is always well understood and that the system sent messages are always sensible responses to the user, do you think of Sherri as human or a machine? If Sherri’s communication skills are at least as good as a human, but she has the tools and data access of a powerful machine then we think Sherri is a super-human.
Now that’s sufficient context on what messaging is, and why now is an interesting time for messaging to emerge as a platform.
There are a lot of new companies trying to do something ‘in messaging,’ so it’s worth trying to create a map of the different participants:
- Channels: Messaging apps and their APIs
- New B2C ‘Assistant for X’ companies: New businesses getting started as assistants.
- Existing B2C companies that should utilize messaging.
- Infrastructure companies being built to enable (2) and (3).
Messaging apps: a mature market
The battle of the last 5–6 years has clearly been over who wins as the end-user’s messaging app. This has been dubbed by some the ‘messaging app wars’ and while it’s an ever evolving landscape, this is now a very mature part of the stack. The quick summary is that players dominate by geography: FB Messenger, iMessage and SMS continue to do some battle in North America, Asia is split up between WeChat (China), LINE (Japan), KakaoTalk (Korea), Telegram and Kik have some scattered market share. WhatsApp has won pretty much everywhere else ranging from most of Europe (except the UK), India, the Middle East and Central and South America. With combined Facebook Messenger and WhatsApp, Facebook certainly has ownership of the most interesting platforms outside of Asia. All of these messaging apps continue to add new things for users and set crazy new usage records every month. 2016 will only continue this trend.
Platform APIs: new openings
This is a big year for a shift in the availability for backend services to access these platforms. Here’s the current landscape:
- SMS: It’s taken years of complex business development by players like Twilio and Nexmo who have done the hard work of going carrier to carrier and cutting deals. There are finally now decent APIs for texting almost anyone in the world. However, it remains ridiculously expensive to send messages (in the US, which has relatively cheap SMS, it is $0.0075 per message — which is 780X the cost of normal wireless data), and there’s very little innovation on the content or interaction of SMS. Even simple functionality like group texts or MMS photos are plagued with issues. However, it remains ubiquitous.
- Facebook Messenger: While it’s still slow coming out, Facebook is working very hard to make Messenger into a platform for businesses. This is distinct from the platform to integrate other apps with Messenger. Messenger for Business is still in a closed beta. We’ll see how F8 changes things.
- WhatsApp: New public statements by WhatsApp indicate that they are beginning to think about the business-to-user messaging case.
- Twitter DMs: Twitter has made some great strides in enabling customer support scenarios and promoting an initially public complaint into a private DM. There is already great API support for Twitter DMs.
- iMessage: Nothing really to report here and seems unlikely Apple will be opening up iMessage anytime soon.
- WeChat and Line already have platforms that have had some well documented success. They have some distinction between subscription style bots (mostly for marketing) and more 1–to-1 direct messaging interactions.
- Telegram has a well supported bot platform.
- In-app messaging: This is an interesting space to watch as some new entrants (such as Smooch, Layer, and Twilio IP Messaging) are coming in to help developers add messaging into their apps without having to build the whole messaging stack. From banking to commerce apps, this is an opportunity to bring long lived conversations to already successfully distributed apps.
New companies — ‘Assistant for X’
Many new companies are getting started that pitch their service as ‘message us, and we’ll assist you with X thing.’ Most of them either have a more rigid command-line interface where it’s clear you are messaging with a bot (like Digit, Assist, Peach) or they support less structured messages and it’s clear that you are messaging with a human. Companies where you are messaging a human today are claiming they will become more efficient with NLP and AI over time (like Magic, Operator, Pana, Facebook M, PS Dept, etc.). These assistants cover many different use cases: local, travel, on-demand, commerce, fashion, banking, finance, and news. Some of these new products are targeted at a specific vertical and others try and act as horizontal aggregators.
There are important questions for each of these consumer facing assistant services to answer to be viable:
- Even in this new platform, how will they ultimately achieve end-user distribution in a cost effective way?
- If they aren’t human powered, is the interaction actually better than a UI?
- If they are are human-powered, how and when will the unit economics make sense?
Most of the great experiences people have discussed so far have been human powered high-touch interactions. Most of these services promise long-term AI automation without a clear path or the team to get there. To survive, many of these new companies have shifted to high fees to cover the cost of human labor. The broader and more complex the potential requests from customers, the more automation (and profitability) looks like a dream.
Many of these new companies are today creating their own tools for customer communication. As the market for messaging platforms matures they will increasingly be able to build great assistant products on top of software solutions like Rep. A good analogy for this is what happened with e-commerce stores — early e-commerce stores had to build their own systems from scratch, whereas now most stores are built on systems like Shopify or Magento.
Existing Companies and Organizations
This is where we get excited. We’re excited because we think people long for a better experience when they interact with the companies already in their lives. This becomes most acute in the support scenarios with huge companies like Comcast or United Airlines. However, even when you’re buying online from a boutique store or trying to find out some information from your local government you often get lost for the best way to get in touch with a human. As a consumer, you’re usually stuck with only one, very painful option: pick up the phone (hopefully between 9–5pm EST) and try and get through to a human who may be able to get you to an answer.
The future looks much more like this: you’ll still do some browsing, especially for heavily taste-based products (like fashion). When you know exactly what you need, there will still be a great use for 1-click buy buttons. However, most businesses that sell anything directly to customers will find a combining of product sales, support, customer happiness, etc. into an essential, new, external facing team that deals in conversations (i.e relationships) with customers over messaging. For lack of a better term we can call this the assistant team. This isn’t actually a new idea. When you walk into a physical Apple Store, a Nordstrom, an insurance agent’s office, a car dealership, or the doctor’s office, you are often there to get assistance. You browse around the store, and then turn to an expert for help. What messaging is going to do for all of these businesses is create a channel and a user expectation that you don’t always have to show up in person for great assistance.
One very important distinction here is that assistance is not just ‘customer support’. Only a small number of interactions with an Apple Store employee or a Nordstrom employee is about support, it is very often about educating customers about products and then transacting.
As businesses moved online, the interaction has put the onus of user product education completely on the user. Today the focus is instead on large scale marketing campaigns (particularly email, social media, SEM), getting some users through a funnel to browse, select, and buy something. As tactics for funnel optimization have become well distributed, the online shopping experience has become homogenous. The end-user experience of interacting with a business online is the same whether it’s a Nordstrom, a Macy’s or a Target. However, their in-store experience is vastly different, with a primary differentiator being in how they provide assistance. Messaging is going to bring this human touch back to person-to-business dealings online and give businesses a chance to provide assistance to their customers in a way that has never been better for customers.
To make this transition work well, businesses will need three things:
- Well trained assistants. Some more customer-forward businesses already have these: there are HotelTonight Aces, Bonobos Ninjas, ModCloth Stylists. These are employees who are trained to be the digital front-line and build relationships with customers.
- Software that makes these assistants’ lives easier. Many platforms have been built to try and make customer support more manageable (Zendesk, Desk.com, Freshdesk, Front, etc.). While these are tuned for support tickets, there will need to be new platforms that shift the focus from support cost reduction to seeing customer communication as revenue generating. Such platforms may integrate product catalogs and recommendations, NLP and intent recognition, and partially or fully automated responses — all in the service of improving the quality and efficiency of the assistants’ ability to communicate with customers. This software will need to be a blend between CX, CRMs, and a messaging client. It will need to help manage teams of assistants and their workloads, and will become more powerful as it becomes more integrated into the workflow of the business. This platform is what we are building at Mensch Labs — we call it Rep.
- Investment in the presence of messaging channels. This will come through the above platform openings. Businesses that have a well distributed mobile app will see value in adding messaging into their mobile app. Those with large social media presences will see value in focusing on those same platforms for messaging. For many SMBs or local businesses that market their phone number, the main messaging channel will likely be SMS.
Infrastructure companies to be created
There are lot of new companies to be created out of this transition. The companies that power B2C email communication today give a good sense of what we’ll see coming out for messaging. Products will be needed for marketing, analytics, optimization, NLP/AI, payments, and personalization. The stakes will be even higher as customers will only respond well to companies that do a great job of using this highly personal space for building a useful relationship. As the underlying infrastructure improves, I believe there will be many more non-profits like Crisis Text Line, travel agents like Pana, commerce experiences like Operator, and vets like Treat. Anyone using messaging as a platform today is effectively a tech company, because they have to build almost everything from scratch. This will soon change.
The main takeaway is that there is a tremendous amount to build. It’s ridiculous that it’s 2016 and you still can’t message a business like you would a friend. Let’s fix that.
Other reading on related topics
Future of messaging apps by USA Today’s @ndiblasio based on an interview with Chris Messina and me
Clippy’s Revenge — Smart Messaging as Platform Shift by Greylock’s Sarah Guo
Conversation Commerce overview and Podcast from Chris Messina
Futures of Text and Server Side Humans by USV’s Jonathan Libov
The Message Is The Medium: Reasons ‘Assistants-as-App’ Work by Nir Eyal
On Facebook’s front:
Here’s to 2016 with Messenger by David Marcus
Facebook Messenger: inside Zuckerberg’s app for everything from Wired
To better understand what’s going on with WeChat in China: When One App Rules Them All: The Case of WeChat and Mobile in China by connie chan
Product Hunt Slack bots collection
On Bots, Conversational Apps and Fin by sam lessin
Index of companies and categories for consumer messaging
This list is not comprehensive. If there is an interesting company that should be on this list, please let me know, as I would love to learn more about them and add them in!
Assistant for X over messaging:
There is a very long list of consumer assistant for X companies. The best set of these that I’ve seen so far is in this Product Hunt collection. They range from travel, to shopping, to finance, and food delivery. Some of theses are full bots, some are human powered, and some are a bit of both.
Agent Tools companies with some messaging focus
Rep — what we’re building. Targeted at B2C conversations and blending human agents with AI techniques.
Grata — WeChat focused
Line@ — Line’s in-house products targeting small businesses
Front — Shared email inbox with support focus, starting to add more messaging functionality
Sonar — SMS focused agent tool
Chatbox — multichannel product focused on enterprise
Intercom — communication platform, main market is B2B internet businesses
Smooch — Messaging middle layer across in-app, many messaging platforms. Like Stripe for messaging.
Layer — Messaging infrastructure.
Twilio SMS & IP Messaging — Twilio’s well known SMS APIs. They’ve newly added IP Messaging.
ZipWhip — Enables landlines and 1–800s for SMS
Telerivet — SMS platform for emerging markets
Existing, potentially relevant, CX Companies