Take your CX strategy to the next level with automation & AI: Intelligent Virtual Assistants

What is the real impact of implementing an Intelligent virtual assistant for startups and how can The Nest by Webhelp help you do that?

Take your CX strategy to the next level with automation & AI: Intelligent Virtual Assistants

By The Nest by Webhelp, 10 June 2021

Artificial intelligence and automation have provided companies with the opportunity to differentiate themselves from their competitors when it comes to customer experience.

Now, what is the first thing that comes to mind when you hear the word chatbot?
Whatever your answer is, let us explain why intelligent virtual assistants can be the key to success for your business. From their creation until today, virtual assistants have become one of the strongest pillars when it comes to developing customer experience strategies. This is one of the reasons why companies are incorporating Automation and AI by adding chatbots to their customer experience approach.  However, having a competitive CX strategy involves not only cutting-edge technology, but also knowing how to manage resources in a more efficient, results-driven way.

Some data points on chatbots

34% of people say their trust in chatbots is increasing over time.
The share reaches almost 50% for people under 44 but is lowest for older people (32% for people between 45 and 54).

58% of people are happy to use chatbots for customer service
The younger they are, the easiest is the adoption: 68% among the 16-24, ~60% among the 25-44

74% of people aged 16-24 would be happy to use a chatbot for an update on an upcoming delivery.
Acceptance varies depending on the nature of the query (when it comes to sensitive information, people usually prefer to discuss with a human)

CCMA research initiative – performed on December 2020 with UK users

Finding the balance between 4 key components to effectively implement a chatbot:

Although many organisations understand the need and benefits of integrating a chatbot into their customer service, many face challenges to make it succesful. We are convinced that the success relies on the balance between 4 components:

  • Thoroughly understand the client’s needs to provide appropriate responses to them. This avoids customer frustration preventing them from abandoning the process.
  • Be present at the right time in the customer’s journey. Achieving full integration with other systems by detecting, for example, when interaction with an agent is needed.
  • Develop an appropriate and bespoke design according to the operation of each company and its needs, allowing it to be managed as smoothly as possible.
  • Use the right technology to achieve results-driven automation.

Not all platforms suit every business and a start-ups should carefully consider their brand, tone of voice and the type of customer they serve.

So, what is the real impact of implementing an Intelligent virtual assistant for startups? 

The exponential growth of companies can force people to diversify their efforts, slowing down processes. Having an expert partner in CX processes using AI and automation can get your business off the ground without any hindrance.  By automating simple questions and flows, higher volumes can be managed without needing a larger workforce. 

In addition, the implementation of virtual assistants leads to a significant reduction of calls and their diversion to messaging channels can be achieved. This allows for a decrease in resolution time while increasing automation and customer satisfaction

Now, all these things are easier said than done and that is why there are so many unsatisfied customers out there. The misuse of chatbots makes companies hesitant to implement them.  Expertise is needed to ensure they are built strategically so that they become a strength and not a liability. 

Chatbots vs conversational apps?

Efficient chatbots tend to become conversational apps which sit at the intersection of text-based (chat) and graphic-based (web, apps) interfaces. Even if a chabot often starts with a “low effort dynamic FAQs”, the ability to progressively move on to an end-to-end transactional bot is critical to deliver the right CX and to benefit from conversational automation.

The Nest by Webhelp designs and implements intelligent virtual assistants which are key to drive voice deflection & automation

To be more specific, we:

  1. SHAPE – Working jointly to analyse insights and ecosystem, to build use cases and journeys that are eligible.
  2. DESIGN – As adoption is key, our conversational experts create flows & intents, fine tune tone of voice, depending on your brand policy & the channels where the IVA will be deployed.
  3. SET UP – We implement these scenarios within one of our partner solutions
  4. TRAIN – Our AI trainers coach the IVA so it can properly answer customers, either for initial training and for support and improvement (new intents, new vocab.)
  5. HANDOVER… to our live advisors if for any reason the IVA cannot answer the question or if the customer wants to speak to a real Human

Also, we team up with partners that have been selected in regard to their ability to integrate with 3rd party solutions, to their omnichannel availability, to their language coverage and the quality of their AI/NLP engine.


Implementing a chatbot to provide technical support for a global Japanese electronics company.

Implementing a Conversational App has changed our client way of working, introducing a new paradigm in Customer Service and Technical Support.


  • Reduce the number of chats handled by human agents, as well as an overall reduction of contacts in high-cost channels like voice and
  • As part of their digital strategy out client wanted to implement a virtual assistant that would support its omnichannel strategy


  • Drifting email and phone interactions to the chatbot and/or human chat (cheaper, more efficient channel) while continuing to provide an excellent customer experience.
  • Additional instant messaging channels which helped our client to be recognized as an innovative company in their sector.


  • 24/7 support: Customers can get 24/7 support through the chatbot; other channels (phone, e-mail, and social media) are limited to 9-18h Mon-Fri service hours. While humans usually struggle with multitasking, chatbots can handle multiple conversations and hundreds of customers simultaneously, without waiting times or queues, and no errors.
  • 27k Chatbot sessions completed in 2020.
  • 38% Chatbot sessions closed without human transfer.
  • 70% Customer satisfaction rate being achieved by chatbot.
  • Savings in operational costs: Even having many processes that are not automated, the duration of chats handled by human agents has dropped an average of ~20% due to the previous checks and instructions provided by the chatbot.
  • Best practices and customer education: As we have full control of the chatbot contents, we have now the opportunity of instructing the customers on following the brand’s best practices when using their products and when troubleshooting technical issues. We also promote additional self-service resources such as video-tutorials and FAQs, which can potentially avoid contacts in the future.
  • Faster mitigation capabilities: When an issue arises, we can immediately react and take action, as the contents get updated in real time, and address any related queries in the chatbot and accurately monitor its performance and impact (e.g. Online store being temporarily closed due to issue in customs shortly after the Brexit – the day after the issue was detected, customers could already find all the relevant information and what to do in the chatbot, preventing more of the 50% of forecasted contact increase). At the same time, it is critical to keep every document up to date.
  • Additional source of customer feedback: customers have now more ways of engaging with the brand by providing product, troubleshooting and generic feedback that helps in the continuous improvement of the services.
  • Detection of other, indirect business areas with room for improvement: Both when creating and reviewing contents, and when analyzing customer feedback, we had the opportunity to report about missing or outdated resources, inaccurate or incorrect contents, accessibility of the information, customer needs not being met, etc. Prior to the implementation of the chatbot, many of these were not detected nor reported.


  • Reporting: Switching to a contact to a user approach has required to work on new ways of measuring volumes, we have renewed the entire chat reporting system by creating tailored reports and combining different analytic tools together.
  • Standardization of processes: Due to the technical nature of the chatbot’s scope, the automated processes need to address and be able to detect the root cause of the issues and provide a solution in an extensive variety of circumstances such as: the product itself, their different tech specs, the troubleshooting steps, the device and/or operating system the product may be connected to, etc. Thus, both the creation of content and its maintenance are demanding.
  • Standardization of language: as the contents are pre-determined, a chatbot cannot adapt its way of addressing our customers depending on their profile (age, technical savviness, origin, etc.) as a person would do. It’s crucial to have this in mind when creating and reviewing the contents to make them clear, simple, and easy to follow without sounding condescending. Audiovisual resources are helpful but also need to be accurately used (e.g., using screenshots of a previous OS version can be misleading).
  • The randomness of human beings: Humans are beings with emotions, and usually controlled by them. Mood is a big factor. It is also difficult to redirect customers when unsatisfied, as the chatbot has not AI capabilities yet.

If you are thinking of implementing an automated assistant for your customers, these are some recommendations to set you up for success:

  • Define project goals in the beginning and evaluate them throughout and at the end of the implementation.
  • Involve all key people from your organization, forming coherent teams that includes both your client and your supplier.
  • Find sponsorship from leaders of your organization.
  • Select the right-tech partner. The first months are critical in terms of exploration of use cases and forming a proper scope for the project.
  • Use an agile methodology and a sprint-based action plan.
  • Use visualization techniques to analyse the vast quantities of data that you will obtain from your chatbot.
  • Involve your client from early on, including the deployment and testing phases; and have at least one key user who can manage the forward after the initial phases.
  • It is recommended to use conversational designers, copywriters, and user experience professionals.
  • Analyze and understand the customer journey and create strong datasets.
  • After the implementation, a continuous improvement approach should be adopted. The chatbot needs to be constantly trained and its performance monitored.
  • To monitor customer experience, satisfaction surveys can be implemented as part of the conversation flows.


  • Being creative is needed when finding ways of preventing human error: using acknowledgement statements, adapting our language to expressions we observe in customers, and always giving the customer the opportunity to revert a decision are the most used techniques.
  • Instruction length: due to the nature of the business, tech support requires most of the time to follow a specific, fixed set of steps, and in a particular order, to get their queries resolved.
  • User engagement: A chatbot user can have a limited attention span and/or abandon the conversation any time, so sometimes it is difficult to assess the efficiency of the instructions when the customer does not confirm that or no additional feedback from the end customer is received.