What Happens When Brands Grab the Reins?
This is the second installment in our Owned Virtual Assistant series. Read Part I.
While conversation between a brand and a consumer is powerful, conversations won’t replace destinations; mobile apps, websites, brick and mortar, and physical services will continue to be integral to a brand experience. And when integrated with conversational and assistive functionality, these are the fundamentals of an OVA.
In thinking through why OVAs may be valuable to a brand or business, start with some fundamental questions that span business, creative, and technical considerations. What is the strategic objective for the conversational experience? How does the brand come to life through conversation and sound? And what technological support is needed to build and scale the “brain”?
With these considerations in mind, we’ve established the drivers that will lead brands toward building OVAs in the years to come.
Driver 1: “Own vs. Rent”
There are multiple factors to evaluate in considering owning your voice tech vs. “renting” it. While it’s free to publish a skill or action, and hosting is a minimal cost, the “rent” comes in the form of being but one application on the proverbial store shelf, “paying” the platform by virtue of access to your voice app’s data. Renting is, having little to no visibility as to how your app will be recommended and/or namelessly invoked, and being beholden to platform updates that may require costly updates to your code.
Two critical factors within the “own vs rent” dynamic are control over data & insights and domain expertise. Let’s look at each.
Data & Insights
The promise of voice-related data is vast: a granular view of what consumers say, and how and when they say it, is endlessly useful to a brand. And while voice-related data can help improve voice experiences themselves dramatically, the insights gleaned from voice tend to have far broader applicability than conversational AI efforts. These insights can fuel product innovation, inform new ways of messaging to consumers, and provide the fodder for deep personalization. There are a number of opportunities for data capture across the voice landscape, including MVAs, but OVAs represent a unique opportunity.
MVAs present some major limitations on what data brands can use from their voice apps. For obvious reasons, big tech is being protective of what it shares with third parties for fear of appearing fast and loose with privacy. While Google does provide transcripts of what users say to 3rd-party actions, Alexa skills only provide an abstraction of these utterances in the form of intent-based variables that match expected user input relative to the app’s conversational design. Neither Alexa nor Google provides raw audio, and neither allows much flexibility to understand a specific, identifiable individual’s behavior.
Naturally, the privacy concerns that are imposed on big tech are far greater, in most cases, than those placed on brands. Therefore, the platforms need to solve for their own risk profile, and brands have less reward to reap. That’s not to say that brands are immune to privacy concerns, but rather that they can assert their own control over how data is collected and used.
But by owning the natural language processing (NLP) voice layer within the OVA, brands are positioned to capture not only utterances but raw audio, and to use these to glean sentiment at all points of an interaction. Over time, this information can be used to anticipate a consumer’s disposition at any point in their journey – from searching to buying, from using to advocating – and adapt all digital channels to meet these needs.
In verticals like retail, food service, manufacturing, and more, an OVA would result in a bevy of data collected through numerous interactions within even a single workflow.
Multiply by hundreds or thousands of sessions, and businesses can make consequential gains in understanding their customers, supporting their workforces, and tailoring customer experiences.
What’s more, there’s the confidence from knowing that these data stores are owned – not shared – with companies who may compete with your brand now, or down the road.
The more sensitive the information that flows between stakeholders in any given industry, the more likely it is that an OVA posture will lead, and prove out a valuable return.
Having an expertise in a field like this – think financial service firms, or a healthcare provider, or a multi-dimensional logistics company – means that you’re less likely to involve an outside computing partner and that there are likely already government stipulations that make the medium of voice in the current landscape highly restrictive to use, or simply inoperable.
We see this today in healthcare, where voice-enabled experiences making use of Personally Identifiable Information (PII) are scant. This is a function of trying to have deeply personalized experiences on general-purpose devices and platforms, which are operated by companies who have a high propensity to collect and act on large volumes of data.
But because of the complexity and regulations of assessing financial portfolios for clients or reviewing patient information necessary for filling prescriptions, an OVA can shine to replace clicks and ‘taps’ in a closed digital ecosystem, rich with custom domain knowledge not easily replicated by others.
Domain expertise can also represent valuable IP, where brands want to keep their own knowledge graph proprietary, and not at potential risk of being replicated or lifted.
Driver 2: Channel Agnosticism
Building an OVA doesn’t necessarily mean ignoring MVAs; they can be a key end-point through which your OVA can reach consumers.
So if you care about reach, but also want control, it’s best you start with a channel-agnostic point of view. For OVAs, the idea is to build the conversational model and intelligence without a preferred voice assistant partner, and deploy outward to the channels where your users are. Many of these channels might be ones you own and operate, but in many cases you may want to also leverage big tech’s scale and their brand loyalty.
Amazon, Google, and Apple report the best brand loyalty scores in any sector, and as previously mentioned, have the greatest reach. If your consumers stand with a certain voice assistant provider but you don’t have any presence there, there’s always risk in mismatched expectations and a soured relationship with your consumer. There’s reason to believe that any given individual may have a blended diet of voice assistants in their lives, tapping into at least two of the MVAs through mobile or smart speakers or cars. Brands work to be digitally relevant in the smartphone app market, to win in search, or to sell with Amazon, and consumer expectations will be to talk to their brands, in some form, in most if not all voice-enabled interfaces.
But the experience a brand offers on owned channels may be far more customizable or robust than that which they offer on an MVA, depending on the brand’s decisions and calculus relative to controlling their experience and their data. With the new deeplinking functionalities Google, Alexa, and Siri are providing into mobile apps, the boundaries are blurring, and one might think of an MVA presence as a launchpad into an owned channel, rather than the destination itself.
Every company needs a voice channel strategy, and the friendliest, most flexible approach to managing both control and reach is to build an assistant that interfaces well with both brand-owned and MVA touch points.
Driver 3: Discovery
You’ll hear both enthusiasts and detractors of voice share the same major concern over voice activations: “How will our customers know about it?”
Our answer is always that without a system-approach to promoting and integrating the voice experience, they won’t. With OVAs the native roll out of your voice offering focuses on where you already do business. In customer service? Insert a voice-enabled mobile app, mobile website, and mobile chatbot that also happens to be an Alexa skill. In activating this way, a user is presented opportunities to engage in a conversation in-line with situations the brand controls, instead of relying on the user to seek out the 3rd-party application when the behavior is so new. Ad dollars shouldn’t have to be poured into marketing Alexa (or, your brand on Alexa), but rather marketing your brand’s conversational presence, wherever it may be.
The central brain of the OVA exists to introduce voice where it can be tailored to channels as they are relevant to the user. The more you control the touch points where your assistant can be surfaced, the less you’ll have to spend touting the assistant – it will simply be found by your existing customers as they interact with your brand.
Driver 4: Brand Identity
Regardless of whether brands are considering owned channels or mainstream tech channels for their voice efforts, brands simply must evolve their identities to be relevant in an audio-centric context. How do you remain differentiated and salient when you’re largely invisible?
Compounding problems, when it comes to general voice assistants as we know them, a 3rd-party voice app is always subservient to the mainstream voice platform brand as a wake word (“Alexa, open Brand X” or “Hey Google, talk to Brand Y”). This secondary placement inside an invocation utterance is by default an awkward position of relegation. What’s more, big tech places major restrictions on how your brand manifests on their platform. Want to use a custom text-to-speech voice you’ve engineered to be the perfect voice for your company? Good luck integrating it into that skill or action.
As mediums and behaviors change, so too do brand requirements. Although not a new concept, a cohesive sonic brand identity is of newfound importance in a voice context. The “jingle” exists to this day as a holdover from the golden era of radio, but it’s far from the complete manifestation of a brand through sound. Decades of brand focus on more visual media emphasized the importance of dynamic logos and derivative assets, robust guidelines, and all manner of treatments in marketing and advertising. Brands that evolved to be relevant and memorable visually must now do so aurally.
While your control of sonic expressions may be diminished in a mainstream 3rd-party voice app context (where reliance on sound effects and musical elements substitutes for the absence of a unique, identifiable brand voice”), any foray into voice comes with the mandate to manifest your brand sonically.
By design, an OVA starts at the core of your brand’s essence, so instead of an assigned voice from an external body, you’re forced to confront the identity of your voice persona. How do you speak? How do you listen? How do you sound when non-verbal cues are needed? Establishing a foundation with these auditory considerations is key to owning a sonic brand identity, necessary for any brand activating in voice, and an invaluable prerequisite for successful OVAs.
The Bottom Line
As these drivers demonstrate, what OVAs afford brands most prominently is greater levels of control. Control over data. Control over channels and touchpoints, and how to scale across them elegantly. Control over how users find and return to use your assistant. And control over brand expression, with sonic expressions taking on elevated importance.
Read Part III: The Rising Wave of OVAs in the Market.
What’s a Rich Text element?
The rich text element allows you to create and format headings, paragraphs, blockquotes, images, and video all in one place instead of having to add and format them individually. Just double-click and easily create content.
Static and dynamic content editing
A rich text element can be used with static or dynamic content. For static content, just drop it into any page and begin editing. For dynamic content, add a rich text field to any collection and then connect a rich text element to that field in the settings panel. Voila!
How to customize formatting for each rich text
Headings, paragraphs, blockquotes, figures, images, and figure captions can all be styled after a class is added to the rich text element using the "When inside of"
nested selector system.