hero-image
How Forward-Thinking Brands Leverage Machine Translation for Better Results

Marketing and advertising content needs to resonate with its audience to be effective. That’s a challenge when campaigns span various languages and cultures, globally and locally, and budgets and resources are tight.

In recent years, artificial intelligence (AI) and machine learning (ML) technologies have improved machine translation to provide time and cost efficiencies for brands entering new markets. While that looks great on paper, what’s often missing is a strategically blended approach of people, process, and technology to deliver content that rings clear across regions and borders—something that machine translation alone cannot deliver.

How machine translation fits within a localization strategy

Machine translation involves translating content from one language to another without human input. The most advanced type is neural machine translation, which continually improves language knowledge by gleaning insight from large amounts of textual data, much like the human brain’s neural networks.

However, machine translation isn’t a one-size-fits-all solution. Different types of content require different types of language services—especially in marketing. Machine translation is one tool in a wide localization toolbelt, and it can be the right solution when applied in the right scenario.

Machine translation isn’t a one-size-fits-all solution. Different types of content require different types of language services—especially in marketing.

Localization is the adaptation of a product or content to address cultural nuances that enable it to resonate with target audiences as if it was developed natively. For example, a food or drink product may not be marketed the same way across state or county lines—and possibly neighborhoods—even though the target audience may speak the same language.

To use machine translation effectively, brands need a strong localization team that can advise where certain aspects of a campaign should fall on the sliding scale of language services.

3 tiers of language services

Localization experts can identify the right ratio of people and technology for a specific workstream or asset, choosing from three service tiers:

Human transcreation of brands’ most creative assets by in-market copywriters: Transcreation involves reimagining the core idea to achieve maximum resonance and appear natural in a specific market by blending copywriting with translation, linguistics and anthropology. Think of the classic Nike slogan, “Just Do It.” This can have a wide range of meanings if translated directly.

Computer-assisted translation (CAT) with a human translator: Computer-assisted translation can provide that sweet spot between person and machine for large volumes of brand copy, resulting in significant time and cost savings. CAT stores translated content in a library called a “translation memory.” New copy is viewed against that translation memory, and the computer identifies repetitive content automatically. This allows translators to focus their attention on new copy, avoiding repetitive translation. Additionally, brands can store ultra-creative campaign headlines and taglines within a glossary to ensure consistency across transcreation and translation workstreams.

Machine translation with or without human “post edit” or proofreading: Lower-risk assets, such as internal materials, don’t require the same level of human touch. These assets strike a balance of cost and timeline relative to quality. When paired with the right post editor, machine translation can achieve strong results.

For best results, start early

It’s important to prioritize localization in the origination stage. By integrating localization teams early in campaign development, brands can create primary assets in ways that require less rework when translating and transcreating. That includes everything from consultation on the copy to vetting the design, including images, fonts, and spacing to allow breathability for text expansion and contraction across languages.

Ultimately, better primary assets increase the opportunity for automation, resulting in faster localization. And these days, localization needs to move nearly at the pace of origination.

Leverage technology throughout the process

Machine translation works best within a larger tech stack that unifies and optimizes the creative production process. Tag’s Di Translate—a portal within the end-to-end marketing execution platform, Digital Interact (Di)—acts as a brand’s gateway to more than 118 languages and a range of machine and human solutions, including a network of more than 4,000 in-market linguists.

For example, prior to partnering with Tag for language services in 2019, SmileDirectClub (SDC) had been using only machine translation. Then, Tag applied the three-tiered model, building a strong glossary for the brand; transcreating ultra-creative assets, like taglines and television commercials; automating lower-risk assets, like print, social and email; and providing cultural consultation for new creative to improve the potential for future localization.

This resulted in improved translation quality, along with up to 58% language spend reduction and 35% faster approval times. And according to SDC’s vp of creative, the technology’s language memory “kind of feels like magic.”

Embedding localization within a tech-enabled creative production workflow allows efficiencies that improve quality, mitigate duplication of effort, and reduce costs through smart automation. And in today’s fast-paced market, those factors can make or break brand campaigns.