This article was originally published in Retail Touchpoints. 

Given the meteoric rise of smart speakers and mobile voice assistants, it shouldn’t surprise you that by next year, half of all web searches will be voice-activated (according to a study by comScore). What may surprise you is the fact that only 4% of businesses are voice search ready.

In the eCommerce space, it’s natural to wonder how this surge in voice-activated searches translates to online transactions. After all, aren’t most voice searches informational as opposed to transactional?

Google and Peerless put that theory to the test in a recent study and found that the number one use of voice-activated speakers is searching for information about sales, deals and promotions. This accounts for 52% of what speaker owners want from a brand and vastly outweighs what many theorized the primary use of this technology would be—finding “personalized tips and information to make life easier.”

This data paints a clear picture of the opportunity voice search presents for eCommerce businesses; there’s market share available for those willing to do the work.

Featured Snippets
For many search queries, Google returns what is known as a “featured snippet,” a box of information above the traditional organic search results that aims to answer the query right on the SERP (search engine results page). When a search is conducted via virtual assistant, the featured snippet will be a voice response (if one is available).

While more commonly visible for informational queries, featured snippets are showing up more and more for transactional queries. See this “computer glasses” query example:

The goal is to overtake the current snippet by providing a better answer with more quality signals. Observe all the elements being used, including text, bullet points and images. Where are these elements coming from? Does the existing image in the featured snippet have an alt attribute? If not, adding one can give your page an advantage. It’s essential to provide context in order to best answer the question.

Page Speed
A study of 10,000 Google Home search results “found that the typical loading speed of a voice search result was much faster than most webpages.” This makes sense considering virtual assistants are trying to provide an instant answer to each query.

Google’s algorithm uses page speed as an important ranking element. To rank in voice search, make sure your site loads as quickly as possible. Test your pages using tools such as Lighthouse, PageSpeed Insights, GTMetrix, and Pingdom.

Semantic Optimization
When optimizing content, it’s important to include all homonyms, slang and acronyms related to your target keywords. This helps search engines match the context of the content to the request of the voice user. Use bold or italic fonts to highlight words that require emphasis when spoken.

Upasna Gautam, a voice search expert, said it best: “The automatic speech recognition capabilities of the voice search system have become intelligent enough to understand accents, dialects, and stressed words, as well as decipher context of homonyms.”

Simplicity
Simple, easy-to-read content can help with voice search SEO. The average Google voice search result is written at a 9th grade reading level. Tools like Readability Formulas can help you assess the readability of the text. As a rule of thumb, aim for a readability score of 70 or above (although this varies depending on subject matter and target audience).

Upper Funnel Targeting
If used correctly, voice search optimization is a great way to open an eCommerce funnel by engaging with consumers that are looking for a solution to a problem.

Studies have found that the average word count on a voice search results page is 2K+ words. This means there’s a need to supply the full context for answers to queries. When creating long-form content to target upper funnel traffic, provide a comprehensive answer to the target query, along with answers to related questions.

FAQs
FAQ content is ideal for voice search if it has a logical information architecture. It tends to line up well with voice search because it mimics the question and answer interaction between a consumer and a voice-ready device. As is the case when targeting featured snippets, it is a good practice to aim for roughly 29 words when answering each question. It’s also worth noting that voice search results are 1.7x more likely to come from FAQ content than traditional desktop search results.

Structured Data
Structured data is important for eCommerce because it gives search engines important information about different elements of a product that they can match to a voice search. At a minimum, each product page should have all relevant product, offer, and aggregate rating schema. Article content and FAQ pages should include Question schema, especially in paragraphs that are targeted for a featured snippet.

Now is the time for eCommerce businesses to think ahead. Search engine algorithms evolved to provide the best possible experience to the end user; voice search is no different. The efforts you make today will pay off in the future.