Voice shopping, or "v-commerce", is the next big convenience winning the hearts of consumers. And retailers are responding. Heavy hitters like Amazon, Starbucks, and Walmart, are already offering voice shopping. For most, though, launching a voice app that really works well is challenging. Read on to see why it’s harder than it looks.
When businesses want to add technical abilities, they usually use the resources they have: their software development or IT staff. However, building for voice requires a very different set of skills than web or mobile. Existing teams often lack the expertise and experience to create a voice app. At the insistence of management, they may try. After all, the documentation makes it look so easy! But among other special skills, machine learning engineering experience is important. Without this, the outcome is often a poor user experience - and no sales.
Higher quality voice apps use AI to learn and become intuitive. Retailers must understand that their in-house team cannot develop these apps as general IT infrastructure projects or even typical software design sprints. It is a waste of time and money.
SCENARIO: A specialty pet retailer decides to have their internal IT team develop a voice app. They want their shoppers to be able to voice search and shop their full product line by the end of the year.
Their IT team built out the retailer's website from the ground up (with great success!), and a member of the team even develops apps on the side. The IT team feels pretty confident they can build the voice app.
However, six months after release, customers are not using the app despite a lengthy marketing campaign. Although customers showed initial interest, they are not making purchases or are failing to make repeat purchases. While performing technically well, the app is found to be cumbersome for users and fails to make useful suggestions to encourage higher value checkouts.
Some retailers hire agencies to build out their voice apps. The option is typically an extremely pricey one and does not guarantee you are accessing a team incorporating the right methodology or the best technological resources to create an app that will be a delightful user experience for your customers.
Working through an agency also means ongoing updates can be expensive and drawn out. When a voice app cannot quickly evolve with customers' needs–and customers find better user experiences with competitors–those customers are unlikely to continue using the app.
SCENARIO: You are a cosmetic retailer with thousands of products. You are seeking to outsource a team to build your voice app. Your initial hurdle is the development costs of building an app from scratch, which, depending on the functionalities you require, can be extremely high.
The agency ends up providing you with a reasonable solution. However, you find that it is not as robust as some of the market's voice apps, despite your high investment. You are also frustrated with the ongoing charges and delays in making updates to your app.
Another reason voice apps fail is that the in-house team or agency doesn't understand that voice search semantics are different from a web search.
Voice is different.
People don't speak the same way they browse a website. If an engineering team doesn't understand or have experience with this, they build something that is technically everything a website can do but is a poor experience for users. They end up producing an app offering little convenience and unnatural conversation for users.
Voice is about convenience.
For example, some quick voice app functions are:
These are all behavioral functions that are more convenient tasks to do over voice. The functionality must be quick and simple to present a pleasant user experience.
Voice app developers should never encompass scenarios that require a lot of back and forth because there is no convenience and customers will drop off.
SCENARIO: You're a retailer building out a product catalog that has 10,000 products. Your app needs to give customers the ability to quickly find products with a single phrase, such as Find me size 7 blue high-heeled shoes.
Your development team does not successfully deliver this functionality. Instead of enabling customers to voice search naturally, the app includes tedious prompts such as Do you want the shoe to have high heels? What color shoe do you want? Etc. Long conversations via voice require too much user patience. Your app does not increase sales.
To deliver a successful UX that your customers will use repeatedly, your development team must understand machine learning and natural language processing technologies.
Voice apps must have the ability to learn much like your sales reps. Sales reps use experience to infer what customers want and need. AI helps you leverage past voice orders to understand user intent and quickly deliver what customers want.
Natural language is involved when users use voice apps. Users need a way to have a seamless dialogue with your product catalog or store. The art of natural language processing (NLP) and natural language generation (NLG) is complicated.
If your development team builds an app structured like a website, they will fail to deliver natural dialogue. Users end up feeling like they are talking to a computer rather than a person. Customers expect interactions like they have with Alexa and Siri.
In particular, younger users almost exclusively casually converse with their virtual assistants, even sharing jokes with Alexa and Siri. These users will never follow a rigid menu structure of command prompts.
SCENARIO: You outsourced your retail voice app development and have been promoting it to your customers for a year. General pick-up and even repeat use are fairly decent, but you are not satisfied with the drop in use by new users and that the promised increase in checkout value has not gone up.
Drops by new users and customers not choosing to put more in their carts before checkout point to ineffective AI and natural language technology. When your platform cannot make intuitive suggestions (or fails to make any viable suggestions), customers will be unresponsive.
Your third solution is to use a turnkey SaaS solution that takes all the best practices for voice commerce and makes it simpler, faster, and more cost-effective for retailers to create a voice app that works.
This option is a subscription model with much lower setup costs and typical ongoing costs that retailers can afford.
Blutag offers an inexpensive turnkey SaaS solution that allows retailers to tap into machine learning and natural language expertise and technology that surpasses what in-house teams or agencies can provide.
Our strong point of difference is that we use AI to leverage individual voice order accounts that learn and grow from our overall network. This "network effect" means our platform obtains data to improve user experiences across our 500+ retailers that offer millions of products to customers each day. Additionally, our solution is the only one on the market focusing on retail. Retailers and ecommerce businesses strongly benefit from our extensive experience in retail.
Our approach considers users' conversational behaviors, and our AI technology improves upon the user experience following every transaction.
SCENARIO: Blutag's machine learning database is not something that can be duplicated in-house or through an agency. The data we capture and use throughout our underlying technology ensures your customers have the most natural and successful conversations while ordering.
Our customers chose our solution at more than a 90% renewal rate. And our pricing removes the cost risk for those trying voice apps for the first time. We make it easy for you to integrate your product lines and quickly get your customers voice shopping.
Blutag's voice app turnkey solution gives you:
Let us show you how! Contact our team to learn how quickly we can get your customers voice shopping on your own voice app!
And for more information on what is evolving in the world of retail voice apps, please check out our new guide: The Rise of Voice Apps Within The Retail Industry.