The Top 3 Gaming Customer Service Challenges (And How AI Can Solve Them)

游戏玩家面临的三大挑战(以及AI如何解决它们)

2020-10-14 12:00 unbabel

本文共723个字,阅读需8分钟

阅读模式 切换至中文

Gaming is an enormous and fast-growing industry, with revenue projected to top $159 billion in 2020 (a 9.3% jump over 2019). In some ways, the pandemic has created an explosion in gaming, as global lockdowns led to more time spent at home and on-screen. And, one way gaming companies can stand out against the competition is by offering high-quality player support, sometimes within the game itself.  The industry presents a unique set of challenges for customer support teams, however. First, many games cater to global, multilingual audiences that want answers fast. Second, customer support agents are expected to have the level of expertise needed to “talk the talk” alongside players. And finally, many players attempt to troubleshoot on their own, which leads them down a rabbit hole of unofficial, English-speaking forums. These three challenges are, nevertheless, big opportunities to use AI and empower support agents to serve players quickly in their native tongue (and also, the language of the game itself). Here’s how: Challenge 1: Massive global audiences The audience for games – whether they’re played on consoles, mobile devices or PC – is uniquely global. Generally speaking, there are no geographical barriers to playing the large majority of games. Because of this, most companies localize them in order to handle cultural differences of multiple audiences. It’s a complex process that not only involves your standard translation tasks but also voice acting for all the different characters. As a result, however, gamers expect that they can get support in their native language, which is not always the case.  The reality is that many companies still work with lean customer service teams, as it makes very little economic sense to hire full-time agents for languages with relatively few requests. Unfortunately, that means some lower-traffic languages may experience longer first response times (FRTs), which can add to players’ frustration.  AI can help solve this by empowering agents with multilingual machine translation for international customer support. In other words, a human is still in the loop, but a machine aids in the translation process. A human checks the translation for accuracy before it reaches the customer, and feeds any corrections back into the algorithm. This process can help gaming companies reach a global audience with their support operations more efficiently.  Challenge 2:  Knowing the language of the game A gamer can spot another gamer a mile away. When support agents don’t know the lingo of the game, they immediately lose goodwill and credibility with players. Some gaming terms are consistent across languages, while others aren’t (for example, slang or abbreviations). Add to that challenge the global audience of players, and the right profile of support agent can be nearly impossible to find.  In this case, multilingual machine learning models for customer service can be trained on the unique terms for each game. As more terms emerge (or terms differ across languages), the model can be fine-tuned to “learn” them as well. That way, companies can focus on hiring agents that are also gamers, and leave the translation challenge to AI (with the lingo intact). Player support in their native language (and the language of the game) is a surefire way to improve customer satisfaction (CSAT) scores.  Challenge 3: Multilingual self-service Gamers are among the most tech-savvy of people who seek out customer support. That means they’ll try anything to find the answers themselves before contacting an agent. If these answers aren’t available on the website, they’ll turn to gaming forums to find them. Most of these forums are in English, which leaves out a wide community of multilingual players.  In this case, AI can help translate common support requests into multilingual FAQs for both the website and in-game help centers. This is the most efficient, low-cost way to support multiple languages at scale. Not to mention, a multilingual help center can alleviate frustration among non-English speaking gamers searching on English-only forums. Key Takeaway: Gaming is a prime use-case for AI Given the global nature of games and players’ big expectations (I should know, I’m a gamer myself), this particular industry is a perfect match for multilingual technology. It gives companies the flexibility to focus on hiring the right people with gaming expertise, whilst letting gaming-tailored AI take care of the rest.
游戏是一个市场广阔、成长迅速的行业,2020年预计收入将超过1590亿美元(比2019年增长9.3%)。在某种程度上,疫情让游戏玩家爆炸性增长。全球性的隔离导致了人们大部分时间都在家玩着手机电脑。而且,游戏公司能够在竞争中脱颖而出的一个方法是提供高质量的玩家服务,这有时候就体现在游戏中。 然而,该行业为客户服务团队带来了一系列独特的挑战。首先,许多游戏面向全球多语种的观众,他们都想要得到快速的回应。其次,客户服务代理应具备与玩家一起“畅所欲言”所需的专业知识水平。最后,许多玩家试图自己解决问题,这常常导致他们在非官方的英语论坛越陷越深。 然而,这三点既是挑战也是机遇。我们可以利用人工智能和授权服务代理商用玩家的母语(以及游戏本身的语言)来快速服务玩家。如下所示: 挑战1:庞大的全球受众群体 游戏的受众——无论玩家是在游戏机,移动设备还是个人电脑上玩——都具有独特的全球性。一般来说,绝大多数游戏是没有地域障碍的。正因为如此,大多数公司会对游戏进行本地化处理来解决多种受众之间的文化差异。这是一个复杂的过程,不仅涉及到标准翻译任务,还要为所有不同的角色配音。所以游戏玩家总是期望可以获得母语服务,然而事实并不总是如此。 现实情况是,许多公司仍然与高效的客户服务团队合作。因为玩家对语言的需求相对较少一些,为此雇佣全职代理商很不划算。不幸的是,这意味着一些使用人数较少的语言可能会经历更长的首次响应时间(FRTs),这会增加玩家的挫败感。 AI可以授权代理商多语言机器翻译来提供国际客户服务,以此来帮助解决这一问题。换句话说,人工担任主要工作,但机器在翻译过程中起辅助作用。人工在翻译到达客户之前检查其准确性,并将任何更正反馈到算法中。这一过程可以帮助游戏公司用更有效率的运营服务吸引更多全球性受众。 挑战2:了解游戏语言 一个玩家可以在一英里外认出另一个玩家。当服务代理商不知道游戏中的行话时,他们很快就会在玩家中失去好感度和可信度。有些游戏术语在不同语言之间是相同的,而另一些则不相同(例如,俚语或缩写)。此外,玩家的全球受众也是一个挑战,而且几乎不可能找到合适的服务代理商。 在这种情况下,客户服务的多语言机器学习模型可以在每个游戏的独特术语上进行训练。随着更多术语的出现(或者术语在不同语言之间是不同的),模型可以被微调来“学习”它们。这样,公司就可以专门雇佣同样也是玩家的代理商,而把翻译的挑战留给人工智能(保持行话不变)。用玩家的母语(和游戏的语言)提供玩家服务是提高客户满意度(CSAT)的一个有效方法。 挑战3:多语种自助服务 游戏玩家是寻求客户服务中最懂技术的人群之一。这意味着他们在联系代理商之前会尽一切努力自己找到解决方法。如果在网站上找不到答案,他们会转到游戏论坛去找。这些论坛大多是英文的,这就忽略了一个广泛的多语言玩家社区。 在这种情况下,AI可以帮助将常见的服务请求翻译成多语言常见问题,供网站和游戏内帮助中心使用。这是大规模支持多种语言的最有效,低成本的方法。更不用说一个多语言帮助中心可以减轻非英语玩家在只使用英语的论坛上搜索问题的挫折感。 要点:游戏是人工智能的主要使用领域 鉴于游戏的全球性和玩家的巨大期望(我知道因为我自己也是一个游戏玩家),这个特定的行业是多语言技术的完美匹配。它给了公司灵活性,让他们可以专注于雇佣具有游戏专业知识的合适人才,同时让专门为游戏量身定做的人工智能来处理其他的事情。

以上中文文本为机器翻译,存在不同程度偏差和错误,请理解并参考英文原文阅读。

阅读原文