Voice is one of those emerging channels that is a struggle for some brands and a revenue driver for others. It could go so many different ways -- all to answer: how do brands really prepare for the future of voice advertising?
A recent study from eMarketer estimates that more than one in three people will use voice search at least once a month this year. The technology has been around for a few years, and users are largely comfortable with it, but we are still in the early days of voice search.
It took mobile a few years for the technology to increase speed, for the industry to deliver a delightful user experience, and for ad technology and attribution to catch up. That's where we are with voice, and I predict that 5G will lead to the explosion of voice search as the value of the Internet of Things and the connected home is truly realized.
Here’s why -- and here are a few things you can start doing now to gain a deeper understanding.
Voice Search Attribution
Most attribution models don't support voice attribution, so there is little data to support assumptions or report the ROI of voice. However rudimentary, you do have the ability to review queries used on your devices by your family to log and analyze for yourself.
If you’re able to convince other family members to provide access (or more folks in your agency, business, or circle of life), you can grow your sample set rather quickly.
With access, I usually dump the queries into a spreadsheet and categorize by query type -- commands (things like “set a timer,” “remind me to”), questions (“what’s the temperature outside?”, “what does my day look like?”), transactional intent (searching for or ordering a product, hours of operation of a store), and use of a device (“show me the baby's room,” “turn off the lights”). I will then ask if any of the queries or categories can be supported in some way by a client’s product or service, and that's where the opportunity is.
Disparate Voice Devices
Devices are also an important consideration. It might seem ridiculous to have a voice-controlled toaster, but what are all the things that it can do in addition to burning your toast to the appropriate amount? How are other people using voice-connected devices?
Practitioners can discover uses through digital connectors like Zapier and IFTTT. Are any of these commands or connectors impacted by your client or industry? What are workflows or triggers that can be established with these devices? Are there supporting questions or content related to those workflows?
Monetization of voice at scale likely won't come on the device itself — it's more likely to come from shopping lists, subscriptions, and finding locations “near me.”
Smart brands like Sephora are building actions that drive in-store visits. McDonald’s even allows you to apply for a job through a voice assistant.
Privacy and Voice
Privacy is seemingly measured in terms of creepiness — users are okay with some amount of digital tracking, but there is a line between useful and leave me alone. Will users be receptive to answers coming from a paid source before the answer? Will they trust the result as much?
It seems that voice ads will still be less intrusive on a display than when delivered by a celebrity guest like John Legend — however nice his voice is. Today, some queries prompt to show the user a results page on a Google Home Hub or Android device.
Do you know the ad packs available for those searches? How about the organic content leveraged to deliver the answer? How many transactional queries are in your data set? Build more of what supports these aspects.
Paid Opportunities for Voice Search
There are a few ad packs that Google is serving today, which appear to be easy “lift and shift” products for Google. The ad packs we see in the travel space on the SERP (search engine results page) are dominated by booking engines and all pay-to-play. Similarly, the ad units for location and Map pack results seem to be the biggest opportunity where we will see paid voice results, especially when the user is prompted to view more answers on a device.
Ways to Optimize for Voice Search
On the organic search side, SEO experts have been preaching technical markup, Knowledge Graphs, and on-page optimization to capture Answer Box, People Also Ask, and other types of SERP features due to how they also support voice search results.
When it is done well, modern SEO understands that if you’re already creating the content, you might as well optimize for whatever search function that content can support. It’s about maximizing reach and brand interactions, which is where we see Google’s algorithms and indexation abilities going.
Semantic search helps all the voice search providers learn folks’ natural language as they describe the objects around them and ask questions. Google’s organic algorithms such as BERT, Rankbrain, and Hummingbird are in better preparation for voice search by understanding the semantics behind the question (or search) and its relationship to the user and devices available.
Google’s ability to crawl and understand podcasts and video content has improved enormously over the last few years — so much so that Google trusts the process enough to index podcasts and video content, often above the fold.
Some searches will even denote specific places in a video that are relevant to the searcher’s query. This is all powered by a good old-fashioned strategy that values the basics of transcripts, metadata, alt text, schema markups, and a sprinkle of amplification.
For practitioners to fully understand voice search and its implications for business, it will always be best to do it yourself first and foremost. Use an assistant and begin automating your life.
Review the data that it collects to understand the intent behind what it is being used for. Shape your website in ways that make the best use of all the devices your customers are using.
When you’re in it, and you’re experimenting, the juicy lightning bolt of an idea will position you for what’s next in voice search.