Slang Vs Google : Best Voice Search experience for e-commerce apps?

Slang Voice Search vs Google Voice Search - Which provides the best Voice Search experience for e-commerce apps?

Voice Search is growing and is being rapidly adopted by many e-commerce apps, especially as they attempt to grow into the tier-2 and 3 markets in India.

Popular e-commerce apps in India with Voice Search enabled

But until now, brands either had to rely on adding a simplistic Google Voice Search option.

BookMyShow with Google Voice Search option enabled

Or had to spend a lot of time building out their own full-on Voice Assistant if they needed anything smarter.

With Slang’s e-commerce specific multi-lingual Voice Assistants, brands now have a better alternative. Powered by Slang CONVA,  Slang’s multi-lingual Voice Assistants provide advanced capabilities that brands need but with the same simplicity as adding Google Voice Search into their apps.

Slang’s Voice Assistants have the following benefits over Google’s Voice Search option for e-commerce apps — 

  • Higher accuracy
  • Smarter
  • Customizable for your app’s needs
  • Automatic translation
  • More interactive
  • More responsive
  • More platforms supported
  • Better support

What is Google Voice Search?

Google Voice Search is basically an option to the default Android Text Search widget or can be invoked explicitly as a simple Voice to text service. In its default mode, it will open up a dialog that will listen to what the user is saying (in the language selected by the developer) and will pass the raw text of what the user spoke back to the app. The app will then fire its regular search operation using the text that it got from the Voice Search engine.

While it's very trivial to integrate, but it comes with the following limitations — 

  1. Not tailored to e-commerce needs: The voice recognition offered by this is a generic voice-to-text engine, which will not understand the context of your application, and thus is subject to have inaccuracies in its recognition.

2. No translation: The Google Voice Search engine provides a literal Voice-to-text conversion and so when users speak in other languages (assume you use that option), the result you will get back would be the literal text that was spoken. You need to do the additional job of translating that back to English, as that's how most e-commerce companies maintain their backend SKU details in

3. Not smart: The speech-to-text engine from Google does just that. Gives you the literal text version of what your users spoke. So if they say things like “do you have onions” or even a more natural “2 kg onions”, your app will receive the exact text, and you will end up searching with the exact literal text and end up with poor search results.

4. No synonyms: If the user speaks, your user could speak in their own words the items they are looking for. For Eg when searching for potato, one could just search it using the more colloquial name “aalo” (which itself could be spelled as “aloo” or “aaloo”), the Google Voice Search engine would not provide any synonym support for your app. Your backend search engine has to be smart enough to handle the synonyms.

5. Not interactive: If your users are having trouble speaking to your app, the Google voice search option does not guide the user. It does not speak back to them when they say something wrong or guide them to speak the right thing.

6. Not patient: If you have used the Google Voice Search offering before, one of the most frustrating issues with it is that it times out very quickly and in very unpredictable ways. Many users, esp first-time Voice users, take some time to collect their thought before they speak. But if the voice experience times out before they can start speaking, or worse while they are trying to speak, it ends up as a very poor experience and could lead to users giving up on using voice.

7. No training: The Google Voice Service does not provide an option to help the end-user get trained on what to speak.

8. No navigation or other actions: Because the Voice Search engine is not smart and does not detect any “intent” of the user, all apps end up using it only for search. Even if the user says something like “where is my order” or “checkout”, the app will end up using them as search terms.

9. No browser or iOS support: The Google Voice capability is only available locally for Android apps and developers need to find other native approaches for adding similar experiences inside web and iOS apps.

10. No tech support: Google is great when it comes to building great software, but as a developer, if you want to have any additional capabilities or have issues, you are limited to community support only.

Slang’s Voice Search

In contrast, Slang’s Voice Assistant for e-commerce apps provides the following benefits — 

  1. More accurate: The Slang Voice Assistant is optimized for e-commerce apps and understands the SKUs more accurately. It comes pre-trained with over 50K SKU items for grocery and Pharmacy and other sub-domains and can easily be extended to support more custom SKUs. In fact Slang Voice Assistant is 46% more accurate than Google Voice Search!
  2. Smart Voice Search: The Slang Voice Assistant provides a lot more than just a simple voice-to-text engine and its integrated NLP system provides a smart offering to its users. Users can now speak commands naturally, and the Assistant will understand the intent of the user and extract the relevant search items from it.
  3. Interactive Voice Search: Unlike the Google Voice Search engine, Slang Voice Assistants provide two-way interactions. The Assistants can also speak back to the user and give them feedback when they are doing something wrong.
  4. Smart Multilingual: The Slang Voice Assistant understands the language of the user but translates the user’s request to English. It can also handle synonyms of things users speak and pass only the canonical value to the app for doing its actual search.
  5. Smart Timeouts: The Slang Voice Assistant dynamically extends for how long it should wait for users speech.
  6. Smart Training: The Slang Voice Assistant can be trained to understand your specific needs more easily and have high levels of accuracy.
  7. Cross-Platform: The Slang Voice Assistant is available on multiple platforms and multiple languages like Android, Web, iOS, Flutter, React Native, etc.
  8. Best-in-class support: The Slang team has an highly responsive support system allowing customers to directly interact with the development team when required.

Read more about the features and their functionality here.

Voice Search is where it all begins with!

Voice Search is growing in importance and your users are going to expect that to be a standard part of your offering. If you want to provide a “simplistic” experience then Google Voice Search is ideal. But if you want to make sure that your app actually uses Voice and has the great experience that the rest of your app provides, you need a Voice Assistant that understands e-commerce apps and gives you something much better than just a plain old voice-to-text service.

And it's actually free to get started. And super easy to add it. Your mobile and web developers can hook it up in less than 30 minutes. Signup for free here and give it a go.