Assistant app quality does matter (apologies for the click-bait title), but not as much as we’d all like to think outside of a few outlier categories like entertainment and utilities where quality is paramount. In the early voice assistant space, discovery is more important than quality when it comes to driving usage. This will change over time, but for now that's largely the case.
When voice assistant app ecosystems first launched, the overwhelming focus was creating a large quantity of apps. In particular, Amazon put in massive effort to hit the 100K skills mark, using that as a way to cite a vibrant emerging app ecosystem.
Over the past 18 months, however, there has been a material shift to focusing on app quality over quantity. The rationale is that by increasing app quality, we can increase app usage and ultimately find voice assistant’s “iPhone moment” where voice assistants become the go-to platform with a thriving developer ecosystem.
Above all, voice is a command-driven interface because of its superb data input speed and poor output speed. For voice assistant apps, the correlation between app quality and app usage is there, but not nearly as much as the correlation between usage and the app’s ability to be discovered. Despite this, we’ve seen the major Assistant platforms continue to focus on producing tools to develop better experiences. While quality is important, the real problem we should be solving is app discovery.
The problem with voice assistant apps is their often poor business ROI when compared to other digital channels, largely because of poor discoverability. Big brands do circumvent this problem by being able to leverage an external brand to drive awareness. For indie developers, discovery is a much larger hurdle.
Every digital channel that a business maintains must either cut costs or increase revenue in some tangible or intangible way, full stop. Today's major voice assistants don't have the technical capabilities to be a cost-cutter like a replacement for a call centre, so they have to focus on being a revenue generator.
The problem with voice commerce and brand experiences today is not a lack of quality or conversion; it's a lack of volume. Any CMO evaluating voice assistants as a channel would be thrilled to see a 1% conversion rate on this new untested channel with 1M users /mo, even if the channel is highly unprofitable from an investment standpoint. Why?
Because with that volume, the CMO knows that with investment in great Conversation Design and UX, they can turn voice assistants into a large profitable channel. Today, however, the inverse is true — voice has a good conversion rate with low volume.
There are incredible agencies like RAIN and Vixen Labs that have the expertise to design and build top-tier apps, but for these apps to drive serious business impact, there needs to be volume. Does this mean the CMO shouldn't build a voice experience? Of course not. For said CMO, voice assistants will be a significant channel, and having their foot in the door building up expertise now is incredibly important given the relatively low cost vs. other channels. Plus, if the brand is large enough discovery is a much smaller hurdle. The point is however, for most brands it is hard to merit significant investment today akin to other digital channels like mobile. Every brand should invest, but not overly so until discovery is solved.
What made the internet a place where every business needs to be? It was search engines that matched consumer intention with business supply through the medium of a website.
We have the same opportunity with voice assistants.
The more we view voice assistants as conversational browsers to connect consumers with the right 3rd party service by matching intention, the closer we’ll be to a thriving voice assistant ecosystem. This isn’t a new opinion. This has been the industry’s (and my) sentiment for years.
However, it’s become clear that the assistants are facing a “chicken-and-the-egg” problem where they can’t fully open discovery because there aren’t enough good apps to match these intentions. Furthermore, developers aren’t getting enough traffic to incentivize building said apps.
To get around this, albeit slowly, the voice assistant teams are manually connecting 3rd party apps with direct intentions, but it’s a slow unguided process.
We need the assistant platforms to give developers anonymized aggregated intention search data in the same way google provides keyword and search data to web developers.
We can solve the chicken-and-egg problem through supply and demand dynamics. Developers want to make great apps that get usage, and the platforms want to be able to serve users great apps for every intention. We can plug the hole in the middle by giving developers the insights they need to spend their time building apps that will meet real user intentions vs. blindly building another sleep sounds, quiz, or weather app.
Through this supply/demand dashboard of user intentions, developers could apply for access to match their apps with a given intent which would go through a certification process by the platform. This is already happening on a small scale but could be ramped up dramatically and eventually automated.
By empowering developers with the tools they need to increase discovery, not quality, the quality will inadvertently increase as a result of increased business ROI for the assistant channel.
Comments? Please send me an email: email@example.com. You can also subscribe to our blog to stay up-to-date with the latest Voiceflow news.
Taking a look at some of the more nuanced challenges that conversation design teams are addressing when building conversational products.
From visuals to smarter prototyping, an open-source SDK, and even a desktop app – here's the breakdown of what happened at Voiceflow V2.