Newsletter #008 • 10 October 2018
Customer Centric
The weekly digest for the customer service obsessed
Dear customer-obsessed subscriber,

Would you trust a machine to translate everything you need to tell your customers? 

Considering that Google Translate is presumably the first thing that comes to mind upon mention of machine translation (MT), probably not—for one reason or another ... or another.

Most customer service managers are skeptical about introducing this technology into their workflows simply because the quality is not good enough. 

But does this mean you shouldn’t use machine translation in customer service? Not necessarily. 

You just need to learn how this technology works, understand the most common mistakes, and how you can adapt the current MT models to customer service. 

For Maxim Khalilov, Director of Applied AI at Unbabel and one of the leading experts on MT for customer service, it’s all about adding humans to the process, because if you don’t, these are the kind of messages your customers might receive

Thanks for reading! 

Olá (did I mistranslate something?),

🔖  This week's top picks 
🤖 The good, the bad and the ugly of machine translation for customer service
There’s one significant detail that prevents most customer service managers from automating translation: quality.

Maxim Khalilov, Director of Applied AI at Unbabel, writes about the fascinating mistakes machine translators can make—mistakes that Unbabel's thousands of bilingual editors catch and correct every day.
👌 The future of customer service needs a new model of operational excellence
According to Esteban Kolsky, founder of ThinkJar and strategist with more than 30 years of customer-facing positions, the future of customer services requires a new model of operational excellence that is built on five pillars of customer centricity. Find out what how to deliver exceptional service in the age of the connected customer by reading this great interview.
😰 Under Pressure
There’s a sentiment in hiring that Jim Mackenzie from Basecamp has run into recently. The idea goes that you want to see how someone works “under pressure” before you hire them. What does that look like? You build in a step with a task, add a tight deadline, and wait to see how the applicants cope. In the worst cases, there is an expectation that giving any help to the applicant would skew the results. And besides, that quick turnaround helps keep the hiring timeline nice and short! That’s always a win. Or is it?
😡 How emotions can influence customer support agents productivity
Productivity is the #1 trait for customer support agents to develop if they want to satisfy clients with their services and, as a result, with a brand they represent. Managers understand that, so they try hard to give a fair deal for customer support teams: comfortable workplaces, bonuses, corporate trainings, out-of-office activities… You name it! But one small yet dangerous saboteur remains there still, deliberately damaging all endeavors and killing productivity at its grassroots level: procrastination.
🎄 Take it from Zappos—holiday hiring is everything
Though it’s only October, many companies are already deep into preparation for the upcoming holiday season. Kohl’s, for example, started hiring seasonal store associates in July. At Zappos, planning for the next holiday season begins just as soon as the current season ends. Since holiday sales historically account for 20-30 percent of a retailer’s annual revenue, it’s imperative to begin early to make the most of the season.
💡 Speaking words of wisdom

“Machine Translation systems are trained to read parallel sentences, which is a bit like teaching a parrot to talk; the parrot may be able to do it, but they are never going to truly understand what they are saying.”

— Vasco Pedro
Pssst, tweet me.
Copyright © 2018 Unbabel, All rights reserved.

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list.