Thankful

What do you need help with?

Chat GPT

How much control will I have over ChatGPT?

Brands will be able to toggle ChatGPT functionality on/off easily in FlowsNext. 

Will there be any disruption to our service?

No, your old system will still be up and running so we can revert back immediately if we encounter any issues. You will not see any drop in quality.

Is there an extra cost to implement Generative AI with Thankful?

How can I learn more about Thankful’s Generative AI CX automation solution?

What is FlowsNext?

FlowsNext is Thankful’s no-code, experience builder that gives brands more control over the CS experience, empowering support teams to create adaptive responses for customer service workflows. 

What is the bare minimum needed to implement Thankful’s generative AI solution?

You only need two things: a helpdesk and FlowsNext.

What makes Thankful’s generative AI solution different from other customer support tools?

Unlike many software companies that use generative AI that only generates a response recommendation to a human agent, Thankful takes it a step further by actually delivering a response directly to the customer. It can perform actions such as canceling an order, order tracking and reshipping, offering discounts, upselling overstocked items, and more. This removes repetitive tickets from an agent's workflow and empowers them to focus on more critical customer queries.

How does Thankful’s integration with ChatGPT impact customer support interactions?

We use generative AI when we deliver a reply to a customer to ensure we are optimizing empathy and personalization.

Generative AI is not making policy decisions– it’s communicating what we intend to say, except rephrased in a more human-centered way. 

What is Thankful doing to ensure that Generative AI is “business ready” and safe to use in customer support automation?

We have taken several precautions to minimize risk and maximize safety through prompt engineering that allows you to to be involved in a conversation on a message-by-message basis. This puts your brand in full control of how generative AI is used in customer support automation.

How will my data be used?

OpenAI is our subprocessor. 

Based on how you want to implement our solution, we can restrict what information we share with OpenAI and ensure that PII is completely restricted. You are in control of your data every step of the way. 
 
For more information on our commitment to security, please review our privacy policy, subprocessor overview, and SOC 2 certification. 

How do I get access to the early adopter program?

How does Thankful make sure Generative AI is not sending an inappropriate response or wrong response?

Just like humans, AI is not perfect and can make mistakes. 

However, Thankful minimizes the risk of inappropriate or wrong responses to almost 0 through our integration with your brand’s policies and procedures. These brand guardrails help us respond to customers and we leverage ChatGPT to help rephrase them based on nuance, empathy, and understanding. 

Thankful’s best-in-class AI models were designed for e-commerce and retail brands. This drives our ability to respond correctly and when coupled with generative AI, this empowers us to solve customer issues and respond with 99.9% accuracy. 

How does Thankful leverage OpenAI’s ChatGPT technology?

Thankful has integrated Open AI’s GPT-3.5 API into our product and tech stack.  We are using Generative AI  to rephrase replies, questions, and internal notes with more use cases coming out weekly.

What is ChatGPT?

ChatGPT is a language generation model developed by OpenAI, trained on a large collection of text data to generate human-like responses to a given prompt.

Thankful FAQ

I already have a Helpdesk/Ticketing System in place - how is this different?

Thankful isn’t a helpdesk. Thankful acts as a virtual SuperAgent resolving over 50 types of customer queries within your current helpdesk.

How does this work with my Helpdesk/Ticketing System?

Thankful integrates with your existing helpdesk and communicates with your E-commerce engine and business processing tools in order to be able to receive, understand and take action on tickets.

How is Thankful different from a chatbot?

Unlike a chatbot, Thankful uses best-in-class natural language processing (NLP) to better understand what your customers want and deliver a human-like experience to fully resolve their requests through all text-based channels. Never forgetting a customer, product, or process, Thankful provides tailored responses and consistent service.

But this is what my human agents do?

Yes, Thankful is able to fully resolve all the transactional problems customers submit, enabling your human agents to focus on more particular customer queries and provide a more meaningful experience.

This sounds too good to be true, where’s the catch?

No catch, we do what we say we do and have the case studies to prove it. And there are no hidden costs either.

Even in the fine print?

Nope.

What channels does Thankful work with?

Thankful works with all text-based channels including email, SMS, and chat.

What kinds of companies does Thankful work with?

Thankful works with leading E-commerce brands in all kinds of sectors: beauty, apparel, shoes, and subscription-based companies to name a few. We also work with SaaS companies.
If your company receives at least 5,000 customer service tickets per month, then we’re a good fit for you!

What about companies that receive more than 5,000 customer service tickets per month?

Even better! Resolving large volumes of customer service tickets is one of the things Thankful does best. Like all good machine learning, Thankful’s performance strengthens as it collects more data.

Does Thankful integrate with my company’s technologies?

Thankful seamlessly integrates with your existing helpdesk, ecommerce engine, and business processing tools in order to receive, understand and take action on tickets. Integrations are viable with (just about) everything from off-the-shelf solutions like Zendesk, Kustomer, Shopify and Magento to completely custom-built tech stacks. To learn how Thankful can integrate with your company’s technologies, set up a consultation with us here.

Is there anything you cannot integrate with?

Let us know if your company uses something that we do not already integrate with. We add new integrations all the time, so we’d be happy to build one out for you!

Which customer issues can Thankful resolve?

Thankful fully resolves over over 50 types of customer requests such as track order, exchange, return, cancel order, and cancel subscription.

How does Thankful differ from Self-Serve Customer Service Tools?

Thankful is not a self-serve customer service tool. We resolve customer problems, not deflect them.

AI Questions

How does Thankful’s AI work?

Thankful combines Deep Learning from millions of tickets to understand your customers just as well as your best human agents. Our vast dataset helps us know the difference between highly similar actions, and we natively understand unclear language from end-customers, including slang and misspelling. We go beyond surface-level intents like "Cancel Subscription" to understand reasons, motivations and mood.

Once Thankful understands what a customer wants, we take action. Thankful uses expert systems designed by the world's best ecommerce teams to deliver amazing customer journeys every time. Thankful integrates into your backend systems to process returns and exchanges, add notes to helpdesk tickets, and more. Thankful consistently enforces your policies while perfectly reflecting your brand identity and flagging potential fraud. We work directly in your helpdesk just like a human agent. The process for your agents remains the same.

Does Thankful use keyword detection?

Thankful does not use keyword detection. Our machine learning systems have studied millions of tickets to learn how people speak naturally and what they truly mean. Keyword detection results in extremely poor user experiences for a large percentage of real customer requests, since detecting keywords loses all nuance. For instance, Thankful is smart enough to know that "I don't want to return this, but I do need a smaller size" is an exchange, not a return, whereas keyword detection gets this wrong every time.

Our machine learning systems also learn automatically over time. We're constantly learning from your tickets and improving to uniquely understand your customers, products, and processes.

What makes Thankful’s NLP unique?

We use NLP models built from the ground up, specifically for ecommerce, to fully resolve the top 50 actions for any brand. Our shared training model enables us to understand your customers with over 99% accuracy on day one. As new training is added, Thankful continues to improve every day. Our expert training team understands the nuance in your business and builds models tailored for you.

Implementation

How long will it take?

Great question. It depends on how complex your processes are, but most integrations take about 15 minutes.

What resources are required to implement Thankful?

We understand that your resources are limited, so we do the heavy lifting. Most integrations require no resources. To see our full process, click here.

How long does deployment take?

Deployment can be completed as fast as two weeks.

How much does deployment cost?

Nothing - we take care of that on our end.