SEO Strategy for Voice Search in Shopping – Supplemental for Article 3 of The Mobile-First Indexing Deep Dive

Voiced by Amazon Polly

By: Cindy Krum

Google’s transition to Mobile-First Indexing is likely to shift a lot of SEO attention to voice search, especially as more and more devices are available with Google Assistant voice search capability built in. Amazon Alexa is also being added to a number of web-enabled connected TV’s, and they have big plans to monetize it with ads as well, making Amazon and Alexa more of a threat to Google and Google Assistant than they already are. In fact, many connected devices will actually now offer both Amazon Alexa and Google Assistant, so that users can choose, change or use both, depending on their needs and what they are trying to achieve.

Price is almost always the primary consideration for a person searching for something to buy online, so it is helpful to break down a searcher’s most likely behavior, in relationship to the price and risk associated with the purchase. If you assume a marketplace where the variation on price is minimal, then the purchase behavior will generally follow a predictable pattern: the cheaper and more consumable the product is, the lower the risk it will be to the consumer, and the more likely they are to purchase it with little information or comparison. The more expensive an item is, the more potential risk the shopper feels, so they will need more information and potentially, additional time for evaluation and comparison. This is important for voice-search and Mobile-First Indexing because it will impact the primary focus and SEO tasks necessary to optimize the products for search.

It also tends to be true that, the cheaper something is, the more likely a person is to be loyal to the store that sells the product, rather than the product itself. As the price goes up, the person is more likely to be brand loyal and impressed by features, but less store-loyal. At the top of the price spectrum, the more likely the consumer is to be interested in features, evaluating and comparing a small group of brands, but not particularly store-loyal. These rules are not set in stone though, and all three behaviors are possible for all three levels of price. This is important because it will impact how initial product discovery in voice search might happen.

The voice searcher will begin a search with their primary loyalty and then move out from there, depending on the risk that they feel is associated with a bad decision. You can actually use this model to determine, not only what is the main focus of loyalty for the searcher, but also to understand their perception of risk associated with a purchase. With that information, a seller could even tailor responses or information available to the voice-search utility, to respond with the best information, that will both answer the question, but also ease the searcher’s perception of risk with the purchase. An outline of these different use-cases is included below, with potential SEO opportunities for each kind of voice-only search-transaction.

  • Low Risk/Low-Cost Products: – Voice Search Most Likely by Store, then Features and Brand:
    • Searchers are more likely to search for what a store has in their inventory so that they can add the purchase to an existing order other similar purchases from the same store.
    • A great example here is paper towels:
    • These are things that are purchased often, and so once the user finds an item they like, they will probably stick with that one as long as the store carries it.
    • If the store stops carrying the preferred item, the user is more likely to test and find new items, rather than switching stores.
    • Users may be trying to add something on to an existing order to save on delivery or get loyalty points.
    • If you just order any paper towels that a store has in stock, you might not love what you get, but they will probably serve the intended purpose.
  • For SEO:
      • Being easy to find within a limited store context is important. Optimization could focus on surfacing within different store-specific search environments.
      • Brand-loyalty is weak, being well associated with competitive brands, and surfacing that information, deals or coupons easily is important.
      • Users are generally not looking for new products or innovation here, so when new products are launched, they must be associated with default products, as a new alternative. Deals and coupons will be successful here too.
      • Store brands have a huge opportunity here, because they know the pricing of all of their inventory, and can leverage existing customer loyalty. Offering consistently low prices on store-brands, offering more loyalty points for the purchase of store brands or discounted shipping will be incredibly persuasive.
      • Features may be set as default filters, so making sure that products are categorized correctly is essential. EX: Searchers may have a default setting to only see ‘gluten-free’ or ‘vegan’ food options. Fruit, vegetables, nuts and some packaged goods may not be marketing specifically to these shoppers, but their items could definitely be a good fit for these shoppers. They need to be classified correctly so that they surface in the most possible results.
  • Medium Risk/Medium Cost  – Voice Search Most Likely by Brand, then Features and Store:
    • As products get higher in price, they get higher in risk, and thus, people are more likely to be brand-loyal. These are things that people may have purchased before and had mixed experiences with, so they know which brands they had good experiences with, and which ones they want to avoid. They may have stores in mind because they know that the store sells the product, but they are mostly interested in finding the brand that they know that they like, for the best price.
    • A good example here is a high-quality sleeping bag:
    • This is an investment that is meant to last, and keep a person camping warm and comfortable.
    • This may also be something that plays into a person’s self-concept or conspicuous consumption, so brands are associated with external acceptance and consensus.
    • People are likely to shop at any store that offers the brand and product that they want, especially if it has desirable features.
    • If they make a bad decision, they might be uncomfortable and have to buy a replacement product sooner than expected.
  • For SEO:
      • Being grouped correctly by brand, and being associated with competitive brands with similar reputations is important.
      • Sorting and filtering by features is a secondary concern but still important.
      • Being easily found within a limited store environment is of least importance. The only variation here is if there is a store-brand that has a strong reputation (like REI’s house brand of luggage – it is the best!).

 

  • High Risk/High-Cost Products – Voice Search Most Likely by Features, then Brand and Store:
    • As product prices get even higher, the likelihood of someone having lots of experience with different brands and options goes way down, and so people focus mostly on features. They may have preferences for the brand and the store, but the main considerations are price and features.
    • A good example here is a car:
    • People generally only know about their own personal criteria of features. They may have brands in mind, that they like, but are easily persuaded to try new brands, since they don’t have as much of a history of experience to inform their evaluations of different brands.
    • Searchers may have brands or stores in mind, especially if there are pricing deals to be had, but the main decision criterion after the price is features.
    • Replacement purchase are not easily made, and the person may have to keep what they got, even if they don’t like it.
    • If they make a bad decision and don’t like what they get. they may have lost a lot of money, and may even risk personal safety.
  • For SEO:
      • Search will be focused on features and feature groupings. Think first about building awareness of features to get into the comparison-set.
      • Think in terms of feature grids, because this is how searchers will filter all of the possible results to generate a consideration set. If features are optional, the safest thing is to assume all optional features are added, for filtering, except in the case of price – price categorization should be based on the least optional features, and clarification. Clarification should be added to the process as soon as possible.
      • Comparison utilities and lists may be a good way to get into a consideration set. Searches like ‘Safest’ ‘Highest rated for safety’ ‘top safety rating’ are things that could easily be entered into a comparison grid, once a consideration set is established.
      • Optimizing for features, and having a good explanation of the importance of the feature will help searchers feel comfortable with their decisions.
      • If you are a store or a brand, you will have to focus on ways to decrease the risk by promoting guarantees and good return policies.