Independent Voice Assistants
Jul 15, 2020
7 MIN READ

Custom vs White Label Voice Assistants: What You Need to Know

Since the early days of voice technology, the term “voice-first” has become a catch-all phrase for a variety of strategies and approaches. Initially, telecom companies and carriers began adding voice search to their products to add value and create hands-free convenience. Soon after, platforms like Alexa and Google brought voice search into homes and cars. Ever since then, multi-content, multi-purpose platforms have expanded rapidly as brands added new skills and access to their content via these third-party devices and assistants. 

As people become more comfortable with voice interfaces and expect greater accessibility and convenience, companies have come to understand the imperative for a voice-first strategy. Simultaneously, the loss of control over the brand experience and customer data when integrating with third-party platforms—like Alexa and Google—has contributed to a growing preference for custom, wholly-owned, branded voice assistants

Brands like SnapchatMercedes-Benz, BBC, Pandora, and Bank of America have all launched custom voice assistants that let them take back control of their customer relationships, brand experience, and user data. 

What defines a white label voice assistant?

Only a small percentage of companies that offer their own voice assistants have actually built their own from the ground up. Most of them rely on partnerships with voice AI platform providers, referred to by some as “white label solutions”. Despite the growing popularity of voice user interfaces in a variety of industries and for a wide range of products and services, developing a responsive, conversational voice assistant is no easy task. No longer satisfied to subvert their brands to another, leading companies are pondering the question of whether to build a voice assistant internally or build it with a partner. 

Custom voice assistants let companies take back control of their customer relationships, brand experience, and user data. 

When most of us think of white label solutions, we envision a SaaS product that can be purchased in a variety of configurations that range from “basic” to “enterprise”. These packages are assembled to fit pre-determined needs based on the size of the organization and other parameters that are typically taken into consideration when purchasing that product.

While there is a certain attraction to purchasing a pre-packaged voice assistant that you can plug-and-play, those solutions don’t often meet the goals of brands trying to differentiate themselves in the market. In reality, those solutions don’t exist as purely as they may be advertised. 

In most cases, brands that are looking for their own voice assistant, have specific goals for that assistant that align with overall brand goals. Accepting an entirely white label solution with little opportunity to customize the experience for their unique user base is not really a viable option for these brands. The very act of purchasing and implementing a pre-packaged solution negates the goals of a custom voice assistant. As more and more voice user interfaces become available in products and services, brands that sound the same will not be easily differentiated from every other voice-enabled offering, and the opportunities for creating a unique experience and gaining competitive advantage are lost.

Purchasing and implementing a pre-packaged solution negates the goals of an independent voice assistant for most brands. 

Brands that have committed large teams and considerable resources to building their own voice assistants in-house often find that they can build parts of the voice assistant, but need help with more complex and sophisticated components, including the ASR and NLU, a custom wake word, content domains, TTS and more. Getting truly conversational interfaces may involve more advanced technology, such as Speech-to-Meaning®. Even at the component level, these partnerships involve a level of customization not available through a plug-and-play solution.

Customization, whether built in-house or with a partner, is one reason a pre-packaged, “white label” solution isn’t the right solution for most medium and large brands.

The Herculean effort of developing a custom voice assistant

The desire to have a wholly-owned voice assistant is a worthy goal for brands looking ahead to a future where hands-free convenience and instant access will be something consumers expect. While building your own voice solution may be the best option for retaining control and developing a branded voice that reflects your corporate values, it may not be as easy as it seems.

In the last five years, companies have realized the difficulties of creating a robust, deep language model and the folly of over-investing in technology that has already solved the problems they’re tackling. Once they overcome the misconceptions of what it takes to develop a wholly-owned voice assistant, leading brands turn to partnerships with independent voice AI platform providers. These partnerships have allowed top brands to develop customized voice user interfaces that solve for a variety of challenges and goals, including:

  • Scalability for expanding use cases
  • Truly conversational interfaces
  • Customer data ownership
  • Branded voice experiences
  • Deep library of content domains
  • Custom wake words
  • Unique sounding voices and accents
  • Custom commands

An independent voice AI platform provider does not sell a white label solution that delivers a specific set of functionality. Instead, advanced voice AI platforms—like Houndify—help brands build voice assistants that are very specific to their needs and users. 

The difference between a pre-packaged white label solution and a partnership with a company that has end-to-end technology and proven expertise is that the latter provides the ability to vault your voice project forward while still retaining your branding, differentiated experience, and data ownership.

Best practices for creating a custom voice assistant

If not a white label solution or building it all in-house, what is the best practice for developing a wholly-owned, custom voice assistant? The simple answer is “through finding the right technology partner with deep expertise.” If you don’t find someone you can work with, who is willing to be your champion, and who will bring your vision to life, then the project is over before it even begins. 

Finding the right voice AI partner that can help you develop a voice assistant that meets your goals is an important first step.

Find a partner with a proven track record and one that not only has the technology, but enough experience to bring their learnings from other partners to enhance your solutions. Find a platform that gives you the flexibility to create custom commands to match the way your users express themselves, and custom domains to deliver the information and functionality your unique product or service requires. Advanced ASR, NLU, and TTS technologies are foundational to creating a conversational assistant that can understand and respond with both speed and accuracy. 

Lastly, look for a platform that doesn’t require you to do all the heavy lifting. Beyond the underlying technology, make sure your voice assistant can have a dedicated wake word of your choosing and that the provider has content domains available—eliminating the need for you to create all your own. 

Content domains allow your voice assistant to respond to queries on a particular topic. For example, a weather domain, such as AccuWeather, allows your voice assistant to understand questions such as, “What’s the temperature in San Francisco?” You’ll want to choose domains that fit your unique use cases. For example, if you’re creating a travel application, you may decide your application should understand queries about hotels, flights, weather, translation, currency conversion, geography, and navigation, but you may not want your application to set a timer or play games. Having access to a library of content domains will give you the range of choices you need. 

One size—or one voice—has never fit all

The reasons to adopt a voice-first strategy are well-known and published in various forms all over the internet. The missing link is “how”. How do you adopt a voice-first strategy? The answer to that question depends a lot on who you are, what market you serve, and your own brand goals. Regardless of the specifics of the voice-enabled product or service, one thing has become clear: Brands need to retain their customer relationships, own their data, and create voice experiences tailored to meet brand goals and deliver unique user experiences.

Creating  a custom voice assistant is more than just adding a custom wake word to your voice user interface, although that’s a good start. Keeping your brand name top of mind for your customers and prospects is a compelling reason to consider a branded wake word. Protecting your brand and data are good reasons to start looking at every aspect of your custom voice assistant—from wake word to natural interactions. Your voice assistant should be a reflection of who you are and provide the kind of experiences your customers have already come to expect. In the near future, the sound of your voice assistant will be as important as your logo and colors for identifying your brand. 

Our formula for success: Get your voice heard with a custom, branded, wholly-owned voice assistant powered by an advanced voice AI platform that can deliver speed, accuracy and customer delight.

Developers interested in exploring Houndify’s independent voice AI platform can visit Houndify.com to register for a free account. Want to learn more? Talk to us about how we can help you bring your voice strategy to life.

Karen Scates is a storyteller with a passion for helping others through content. Argentine tango, good books and great wine round out Karen’s interests.

Interested in Learning More?

Subscribe today to stay informed and get regular updates from SoundHound Inc.