Search
Close this search box.

GUEST COMMENT The barriers to voice and visual search implementation: what are they and how are they overcome?

Remember when people had to communicate by letter in the post, or when you had to visit a library to do research? It seems so archaic, yet it was only a few decades ago. Remember when you had to type things out to get the information you want? It might not seem so ridiculous now but fast forward 10 years and the process of typing things out to search could be as laughable as getting the encyclopaedia out for a research project.

It’s not news that the search landscape is changing dramatically. If you were to imagine a world where instead of typing into Google ‘floral pink midi dress’ you say it out loud as a command, or, better still, you take a photo of one you’ve already seen, and a search engine finds it for you instantly, you would be imagining the future of search. In fact, research has found that 62% of millennials want visual search capabilities more than any other technology innovation and it’s predicted that by 2020 50% of all searches with be voice searches.

The consumer is hungry for innovation and retailers are starting to respond. Even traditional high street stalwarts like Argos and Marks & Spencer – who last month announced the launch of Style Finder, a mobile-only feature that allows customers to search products using photos – are starting to look for ways to future-proof their proposition. But ‘starting to’ is the key phrase here. If the big players are only just starting to invest in these new technologies now, how and when can we expect to see the smaller guys following suit?

Barriers Vs Misconceptions

As with any innovation there are of course barriers (or perceived barriers) to adoption; “We don’t have the right capabilities within the business,” “There’s no budget” or even “We aren’t in a position to take gambles on trends” are all examples of these. But are these really barriers, or are they misconceptions?

Cost can’t be argued with. If you don’t have the budget, you don’t have the budget. However, does the investment really need to be as expensive as you think it does? Let’s take visual search as an example. It is assumed that in order to be visual search optimised, you must have a native app, but this isn’t necessarily the case. Building a Progressive Web App (PWA) – which is accessible from a web browser – offers up the opportunity for integrating visual search, but at a much lower cost. This is because – unlike native apps – PWAs only need to be developed once, whereas native apps need to be developed separately for iOS and Android operating systems.

Not having the capabilities in-house for implementing voice or visual search absolutely isn’t a barrier to implementing these technologies, yet it is often perceived to be one.

There are plenty of e-commerce agencies out there that would be more than happy to support any business through its search journey. Alternatively, freelancers are also a great resource to lean on when going through any form of digital transformation. It really isn’t essential to increase headcount and make specific hires in order to stay ahead of the retail innovation game.

When it comes to the issue of trends or ‘fads’ new search technologies can’t really fall into these brackets. Voice and visual search capabilities have already been implemented by a number of big brands and success has been recorded. Forever 21 noticed a 20% increase in average order value when the brand first trialled visual search, and leading home goods brand Wayfair credits its exceptional customer service with the implementation of visual search. Voice search paints a similar picture with research showing that a quarter of Brits now have one or more smart devices in their home and this is set to grow further.

So, how can these perceived barriers be overcome and how can smaller players get a piece of the new search technology action? To answer these questions, we need to look at both visual and voice search in isolation.

Getting ready for voice search

There are three key areas ecommerce professionals can look at to ensure their sites are optimised for voice search:

  1. Optimise for questions

The majority of voice searches take the form of questions. This is one of the key differences between voice and text-based searches. Website content should therefore be structured to answer potential customers’ questions, instead of just targeting specific keywords. One of the easiest ways to do this is by creating a carefully considered FAQ page.

  1. Aim for featured snippets

Not only are featured snippets now the first organic result to appear, they are also the only result that will be read out by a virtual assistant. This is particularly significant for voice searches made through smart speakers, where the user will hear the featured snippet but not be able to scroll down to look at other results.

According to a report by seoClarity, over 20 percent of featured snippets are triggered by the same 25 keywords. Among the top 12 are “how”, “what” “where”, “when”, “why” and “who”. Targeting these words gives websites a better chance of ranking highly in voice searches andappearing as a featured snippet.

  1. Identify long-tail keywords

The language used in voice search is much more natural and conversational, closely resembling everyday speech. Consequently, the emphasis should be on long-tail keywords.

Long-tail keywords are longer, more specific keyword phrases. For this reason, they can be an effective way of ranking highly in voice searches, when users are likely to use more detailed search terms. As a result, the traffic they generate tends to be of a higher quality, with visitors more often already at the point of purchase.

Why is visual search relevant?

We are still discovering the full extent of the possibilities of visual search. However, a few trailblazers in the e-commerce sector, such as Amazon and ASOS, are already showing how effective it has the potential to be.

Visual search is a great way for customers to find what they’re looking for more quickly – particularly when a retailer has a large volume of products. It is also a way for brands to expose their customers to items that they would have been unlikely to come across through the traditional filters, for example products that were added to the site a while ago, or that have not been selling well.

While the way we search visually is likely to continue to evolve over the coming years, here are two strategies that can be easily adopted now.

  1. Assign textual elements to all images

Images should have a descriptive name that incorporates relevant keywords, in order to provide additional information to search engines and help them to recognise what is in the image. Similarly, they need to have a descriptive alt tag and caption, as these are also used by search engines to decipher and index images.

  1. Grow a Pinterest presence

Pinterest is currently leading the way on visual search following the launch of Pinterest Lens, a first-of-its-kind camera search app that recommends ideas to try based on a user’s own photo. The technology has proven extremely popular, with over 600 million visual searches carried out using Lens every month. It is therefore important for retailers to have a presence on Pinterest if they want to increase their chances of being discovered by potential purchasers this way.

The time is now

Voice and visual search technologies are here and they’re here to stay.

Any brand that is serious about dominating in their market and pulling away from their competitors must start getting to grips with how consumer habits will change through voice and visual search, today. Only once they’ve mapped out and understood these new customer journeys will they be able to start looking to the future of their business.

Author: Sam Rutley, managing director at ecommerce agency PushON

Read More

Register for Newsletter

Group 4 Copy 3Created with Sketch.

Receive 3 newsletters per week

Group 3Created with Sketch.

Gain access to all Top500 research

Group 4Created with Sketch.

Personalise your experience on IR.net