Copy
Google and WeChat take on Slack 
February 7, 2020

A conversational newsletter from Zendesk

Last month Edward Koo, a credit trader at JPMorgan, was placed on leave by the financial firm he’s worked at for more than 20 years. 

The reason for his suspension? He used a WhatsApp group chat to talk shop with colleagues, according to Bloomberg

While it’s still not clear whether Koo did anything wrong, the incident provides a cautionary tale for what could happen when our private conversations and professional lives intersect.
 

Google goes Slack #️⃣ 

Messaging is no stranger to the modern workplace, with platforms like Slack and Microsoft Teams joining email as the de facto way many companies communicate internally.

Last week, The Information reported Google is joining the fray, with plans to unify its cacophony of office-oriented apps — everything from Gmail and Drive to Hangouts Meet and Hangouts Chat — into one messaging-first platform. 

Yet as the workplace messaging space gets more crowded, WhatsApp remains the most popular messaging app in many offices.

WhatsApp and other modern messaging apps are how we keep in touch with friends and family, so it’s natural for employees to use them as their digital water cooler.

Things get complicated when collegial chit-chat bleeds into work talk, particularly in highly-regulated industries like banking.  
 

Your word is (not) my bond 🇰🇷

We’ve talked about how the mainstreaming of emojis is causing confusion in the U.S. justice system. 

Because emojis can have nuanced meanings and look substantially different across platforms, courts have had difficulty interpreting them as evidence. 

In South Korea last week, a district court ruled that messaging-based conversations might not be legally binding at all. The ruling concerned a bond agreement made between two financial securities companies via Telegram, the encrypted messaging app. 

The Seoul court decided that because encrypted chats don’t leave a digital paper trail, they can’t be considered contractually binding. 

According to the report, the message for Korean bond traders is clear:
 

‹‹
Don’t always trust private messaging apps when doing deals.

››


WeChat, you watch 👀

While the JPMorgan and South Korea examples are fascinating outliers, the fact is professionals around the world use their private messaging apps to do company business every day. Most of the time, it’s the companies who are left in the dark. 

Business messaging channels like WhatsApp Business API, Apple Business Chat, and RCS Business Messaging, allow brands to message customers at scale, and to store those conversations in their software of choice. 

But what about casual texts between financial advisors and their blue chip clients, car salesmen and prospective buyers, healthcare professionals and patients? 

These interactions are left out of their employers’ conversational record, and lost when an employee leaves the fold. 

WeChat Work, the Chinese messaging giant’s workplace messaging app, is aiming to solve this dilemma, according to a report in Technode.

It used to be common for brands on WeChat to have individual employees send direct messages to customers or invite them to group chats. This was great for building relationships and sharing discounts, but left brands with no visibility into their employees’ interactions with customers. 

A December update to WeChat Work changed this, enabling businesses to track these conversations, reassign them to a new employee in case of turnover, and — most importantly — add them to the customer record by connecting the app to their CRM.

This has downstream effects as well:
 

‹‹
They can learn when to engage with the customers at the right time with the right content, and easily track all the selling progress from a centralized KPI dashboard.

››


The report calls this phenomenon “social commerce.” We call it conversational business and, once again, China appears to be leading the way.

Google’s Chatbot Breakthrough 

There’s so much noise in the chatbot space it can be tempting to tune it all out. But this actually seems like a big deal... 

Google last week announced it had created a bot that “can chat about… anything.” 

Called Meena, the bot is technically a “conversational neural network” with 2.6 billion parameters. It was trained on 40 billion words, or 341 GB of text data, which includes conversations mined from social media. 

To show how smart Meena is, Google came up with a conversational metric called the Sensibleness and Specificity Average (SSA). Humans have an average SSA of 86%. Mitsuki, a chatbot from our friends at Pandorabots that won AI's prestigious Loebner Prize four years running, has an SSA of 56%.
 
Meena has scored as high as 79%.
 

What does it all Meena? 🤔

As AI expert Ronald Ashri explains in Techvibes, Meena is “very much still in the lab.” It hasn’t been incorporated into any product and Google is unlikely to make it available anytime soon. 

And while Meena’s conversation skills are “incredibly impressive,” Ashri wonders how useful a bot that can chat about anything — but doesn’t have a distinct purpose — really is:

‹‹

Being human-like is not the end-all and be-all of bots.

››

👋 Have a great weekend! Feel free to send me your feedback, story ideas, and digital water cooler gossip by responding to this email. 

Dan Levy
Editor-in-chief 
@danjl


Copyright © 2020 Smooch Technologies ULC.
5333 Casgrain Ave #1201, Montreal QC, Canada H2T 1X3
You're receiving this because you signed up for messaging news and insights from Zendesk (formerly Smooch).
You can always update your preferences or unsubscribe