By John Campbell, Managing Director, Rabbit & Pork
Following Amazon’s recent announcement of their latest suite of Alexa enabled products, which included Omni smart TVs and Astro – a small robot for your home, the reliance on voice-enabled products is growing.
Of course, it’s going to be a few years until we all have Astro robots zooming around our houses (which are priced at $1,499). However, it isn’t these flashy new devices which have caused a shift in users’ searching behaviour. It’s the smart speaker in your living room or kitchen, Siri on your phone or the assistant in your new car that has shifted the dial of voice search. And it is crucial brands sit up and take notice.
Research has shown that a sizable 30% of UK consumers have used Voice Search in the past month. And COVID-19 has had a direct impact on voice search too. 55% of UK smart speaker users increased their usage through the pandemic – with more users at home resulting in higher usage.
The considerations for brands, when it comes to voice technology doesn’t stop and start with usage. The most pressing for many brands is around voice search. What happens when a consumer using Voice technology asks a question about your product.
For years brands have invested their budgets and technical focus on website development, SEO and paid search, but what about voice search?
Worryingly, brands that are yet to engage in voice search optimisation run the big risk of delivering misinformation. Considerations such as can Alexa or Google recognise your product or brand name, when answering a question and where exactly is that information being pulled from?
The Importance of Voice for Brand Recognition
Can the voice assistant correctly perform ASR? Automatic Speech Recognition (or ASR) is the process of the smart speaker converting speech from the user to text. For established brands and products, brand recognition shouldn’t be too much of a concern.
However, there are cases when Alexa and Google struggle when it comes to FMCG brands. Some working examples include:
1. New brand and product names
If you are launching a new brand or product, it’s going to take time for Google and Alexa to understand it. A great example of this would be “Cadbury Plant Bars” which launched this week.
Google and Alexa will need to build up an understanding over time that “plant bars” is a product sold by Cadbury, which will take multiple and authoritative source references.
2. Complex naming conventions
An example here would be Vita Coco – the popular coconut water drink who recently launched “Vita Coco Choc-o-lot”. All the packaging has hyphens in “Choc-o-lot”. Asking Google by using “choc-o-lot” broken up in three parts we can see Google struggles to correctly perform ASR – picking up “Choco light” instead of “choc-o-lot.” Careful consideration for these product variations are becoming more and more important if brands want to be able to expand offerings and still engage with a loyal customer base.
3. Pronunciation of brand name
A classic problem when it comes to voice search for brands, is if there is any ambiguity around the pronunciation of a word. This is evident when you see the auto-complete on Google for “how do you pronounce…”
All of these problems can be rectified overtime with optimisations and building out content. However, FMCG brands now need to consider voice search when deciding on brand and product names in a similar way they would when checking for trademarks, domain names or social account handles.
Where are your brand answers coming from?
If you are lucky enough to not have any of the aforementioned ARS issues your focus needs to be on where Alexa and Google are getting the answers from.
Both assistants use a wide range of sources to answer questions. The majority of the answers for FMCG brands comes from scraping the web for answers. Google have been doing this for years to create answer boxes which sit at the top of search results on a web search.
In Rabbit & Pork’s recent FMCG voice search ranking report we tracked that in 85.7% of the results returned in the report, Google did not read out information from the brands’ official sites.
These “non-owned” results are where Google grabs the result from any website that
contains the answer.
Websites need to be of a certain authority and quality to gain these results. But one very concerning point here is that users posting on popular forums or UGC websites could be providing these answers. In this, and past reports we’ve seen examples such as:
– Incorrect ingredient information for well-known brands, like Coca Cola
– Answers coming from forums, such as Mumsnet
– Completely incorrect answers, e.g. “does x contain y”, “yes it does” – when in fact it does not or ‘You can buy this product from location x, y, z’ – when the product isn’t on sale
One of the areas that Alexa struggles with is voice search. Google has the advantage of using all its search products in answering questions. Whereas Alexa uses more third-party integrations to supply information.
One solution to plug the gap is Alexa Answers. The platform live in UK, USA and Canada allows Amazon customers to answer questions that Alexa can’t answer. Users provide an answer which then is live on Alexa in minutes. Answers are then up and down voted to find the best answer.
Taking control is key
As you can see from our 2021 report, there are currently huge gaps for brands, in terms of what they are doing with Voice Search, and where opportunities lie. Taking complete control of your brand’s narrative and customer conversation is vital to any marketers’ success; however, Voice is unsurprisingly an easy touch-point to overlook, and it needs to not be. It should be a key part of planning from the start and form how you want to interact with your customer. With the growing rise of Voice usage, and certain brands already taking the front foot, it isn’t the time to be complacent with your Voice Search planning.
You can download Rabbit & Pork’s full Voice Search report, which focuses on FMCG brands, here.