The future of voice and visual search
20 June 2018
The future of search is a topic that’s widely covered and, of course, many people have their own opinions on how search is evolving and will continue to do so in the future.
Here at Bing, our own internal research is focused around how best we can adapt to and, indeed, predict the changing search landscape to create tools that better facilitate positive experiences across platforms and devices.
At SMX in London last month, we invited a panel of experts, led by Impression’s Laura Hampton, to discuss their thoughts on the matter. Below, we have brought together all their answers to provide you with one of the most thought-provoking takes on the matter of search evolution and how, in particular, voice and visual search will influence the way you do search marketing in the future.
- Arianne Donoghue - Epiphany Search
- Ben Wood - Hallam Internet
- James Murray - Bing
Tell us about voice and visual search and how it is currently being utilised in your agency and more widely.
Voice and visual search usage is clearly growing but we're still having to educate many of our clients on why investing in voice and visual search could benefit them now and in the future.
Our clients are operating in a much more competitive SEO landscape with less clicks available than in previous years due to the evolution of voice search and the rise of instant answers in the search results. Our agency is currently combating this by investing in featured snippets through targeting informational queries, optimising structured data and cleaning up local SEO information for many of our clients.
A couple of our fashion clients have already started testing the use of 'actions' and 'skills' as we suspect these will be key to brands being able to capitalise on the evolving voice search market in years to come.
Some brands are already leveraging the capabilities of voice search to boost brand engagement and sales using Alexa Skills and Google Actions. For example, Tide's Alexa app, Stain Remover, doles out advice on various types of stains. Campbell's Kitchen reads out recipes and then helps guide the cooking process. Nestlé has also rolled out a "GoodNes" skill, which delivers voice cooking instructions on Alexa devices.
When it comes to visual search, I'd say it's an easier sell to clients right now, especially those operating in fashion and home decor markets. A few of our fashion clients are dabbling in Pinterest ads and getting strong results and this is definitely something we're planning to continue to push as an agency.
I'd look no further than ASOS (lens technology) and B&Q (visual search ads) for best in class examples when it comes to visual search.
With my focus being on the paid side, there's isn't a huge amount that can be done at present since neither Google or Bing support ads via voice at this point.
Visual is a little different. We've started to see examples of Shopping ads being pulled through into Google Image Search, Bing has released the ability to include up to 10 images within your product listings, in addition to their "Intelligent Image Search" feature which allows you to find items within an image that you like. This doesn't yet appear to be using ads, but it's only a matter of time.
Pinterest is another platform where visual search is being utilised well - they've been running advertising for a while now, offering a number of different targeting options and we've seen great results when we've run campaigns for clients, in some cases driving traffic to site for a lower cost than we can through generic paid search. Brands who do well here are ones which are naturally very visual to start with - companies like Made.com and B&Q within the interiors category, for example. Within clothing, brands like Burberry, JD Williams and Dorothy Perkins do a good job of serving up ads which blend in well with the rest of the content on the page.
It’s still very early days for voice and visual search and we are seeing some advertisers experimenting with these emerging technologies as part of their overall strategy.
However, we don’t want to limit the scope for advertisers to just our product offerings. Microsoft’s vision is to democratise artificial intelligence, which means empowering every person and organisation to create their own AI solution, tailored to their specific needs. What’s exciting is we are seeing brands already doing this to solve problems in their own organisations. For example, McDonalds used Microsoft’s speech recognition to improve their DriveThru service, transcribing customer orders in a quicker and more efficient manner. Uber have used our facial recognition technology to create a driver verification system that improves customer trust and safety.
This is not voice and visual search as we typically think about it, but it’s the same core technologies being used in creative new ways to solve customer problems. I can’t wait to see what advertisers can come up with to drive the next wave of innovation in our industry.
What factors currently prevent you from investing more into voice and visual search, and what would need to change for you to increase that investment?
The biggest thing that stops me doing more with visual and voice right now is that there's not a marketplace for us to buy ads for yet.
I'm sure this will change in time, but the search engines in particular need to consider very carefully how they integrate ads into their offerings.
There's a number of interesting considerations to think about as voice in particular rolls out - if the aim is to return the one "best" answer for you, how do you represent your brand online so that the assistants can find out everything they need?
Arguably brand recognition may become less important and products which are less well-known but a better fit may get returned more often. Alternatively of course, if your assistant knows you usually prefer X brand, then it may be inclined to only ever show you that instead.
With visual, we're definitely nearly there. Google are now showing Shopping ads in image search, although this appears to be an occasional, rather than a regular thing. I also don't think it would be a huge leap for Bing to start serving paid product results to users who were using their Intelligent Image Search product, although you'd be relying on their AI to make the product matches for you.
For many clients it is a case of priorities. We are all busy people and finding the time to dedicate to new disciplines like voice and visual search, where the opportunities are still nascent, is difficult to justify. I think voice search will be the next digital obsession, in the same way that mobile has been for most of the last decade. Much as we saw with mobile, those advertisers that are willing to experiment whilst the technology is still in its infancy, will be ahead of the curve once it becomes a marketing “must have” rather than a “nice to have”.
Attribution and reporting are big roadblocks when it comes to showcasing the value of voice and visual search to our clients right now.
Many of the leading reporting platforms out there don't enable us to report on visual search impressions for example, and voice search reporting and attribution is currently non-existent.
I'm hoping that the major search engines make this data available to webmasters in the near future so marketers and agencies can use it to accurately judge and report on the value of voice and visual search over the coming years.
What are your top tips/takeaways for businesses looking to invest in voice and visual search now and in the future?
With both voice and visual search we’re at a stage where it’s too early to go “all in” but it’s too late to be doing nothing.
For voice, the best tip is to think about the variance of natural language that people use to describe the same product or service. A person searching for a family holiday might just as easily search for “holiday for two adults and two kids” as “holiday for a family of four”. Both searches are looking for the same thing but understanding the nuances of how your customers talk when they are describing your products will be the key to unlocking voice search.
For visual search the simplest thing to do is ensure image extensions are added to all of your campaigns. Image extensions help differentiate your ads from the competition, have high engagement rates, and will be increasingly important for future product developments across search. Taking the time to implement image extensions to existing campaigns (as well as making them best practice for new campaigns) is a worthwhile investment.
For voice search, right now I'd simply encourage businesses to continue to focus on answer box SEO and gaining relevant featured snippets optimised at voice search queries. Remember to use natural language where possible as digital assistants are getting better at natural language processing.
In addition to this, focusing on the latest structured data capabilities such as 'speakable' schema and how-to schema may be safe bets, as well as investing some effort into learning how to develop apps for Alexa skills and Google actions.
In terms of visual search, I'd encourage start optimising images for search, compressing images and creating a library of high quality assets that can be used on search and shopping ads. Exploring other platforms such as Pinterest will increase the potential reach for many brands so this is also worth investment.
Let me break down my top tips by each area.
Don't be afraid to use Broad match - as the syntax of how we search begins to change as we move from a search box to voice, it'll be impossible for us to capture every possible variation of a query. Clever use of Modified Broad match will save you trying to come up with every permutation, which just won't be possible and will take a crazy amount of time.
Use tools like Answer The Public to research how people do actually search - when using voice, we tend to use more question words and talk more conversationally - rather than the techniques we've "developed" for typing in queries to a search box. ATP is a great way to dig into a specific keyword or topic and come up with keywords that more closely mirror how actual people speak, rather than search.
If you're a brand that can run Shopping campaigns, make sure you do. It's also worth looking at the range of images included for a product. Do you show the product from multiple angles so that you're prepared for your products to be included within relevant image searches?
Give platforms like Pinterest a try. You can target by interest, keyword and audience - in the first instance I'd suggest keyword targeting as you'll be able to draw insights from paid search and the ads which match keywords blend into the results much better.
In summary, our panel of experts have provided us with some great insight into where they believe the future of search is heading, specifically voice and visual search. Even though voice and visual search may not have fully evolved just yet, it is only a matter of time before they become integrated into the consumer’s decision journey. So what steps are you going to take today to optimise your search campaigns? Start of by advertising on Bing Ads.