Is Your Company Ready for Deployment of Virtual Assistants? (Chatbots)?

KR_blog_widescreenSynopsis

Enterprise chatbots are commonly known as virtual assistants, which are essentially automated virtual employees. They make use of conversational Artificial Intelligence (AI), such as Natural Language Processing (NLP), which has advanced tremendously in last two years. This article details the most common use cases and roadblocks in deploying enterprise chatbots.

Business Case

Enterprise chatbots are deployed in many organizations for 3 main reasons. Firstly, it is used to engage customers through selective automation of common tasks, FAQs and general inquiries. Next, companies are using chatbots for sales conversion and customer service by interacting with prospects through profiling questions, funneled suggestions and upselling options. For existing customers, the chatbots can help in fulfilling selected transactions like checking of products, orders, updating of personal particulars or payment method. Finally, chatbots are used for internal collaboration to strengthen knowledge management and transfer, sharing business intelligence and improving staff engagement. 

Enterprise Solution versus a DIY Software-As-A-Service (SaaS)

The key difference between an enterprise chatbot and a DIY self-service SaaS chatbot is the ability for the former to learn from data generated by new chatbot users and existing data sets from the enterprise. This process makes use of supervised machine learning, which allows the machine to learn from labelled information. The chatbots are learning from data, as humans learn from experience. Natural Language Processing (NLP) is about training the robots to understand, converse, contextualize and internalize different languages just like any human would. 

Many DIY SaaS products are easy to start, but it is quick for users to hit ceiling in terms of the capabilities that they need. Common ones are:

  • Lack of enterprise level security measures and dedicated hosting requirements
  • Robust Natural Language Processing and Understanding requirements
  • Connectivity with existing systems

Virtual Assistants Readiness Assessment

With these obvious competitive advantages offered by deploying enterprise chatbots, why aren't many companies not pursuing this strategy? The answers to the questions from the following eight categories may give some insights on their strategic readiness level.  



1. Senior Management Commitment

Is your C-suite executive interested in leveraging Virtual Assistant (VA) as a new customer service channel to increase productivity and customer experience? 

Are they keen on leveraging VA as new marketing and sales channels to increase lead generation? 

Is there a designated executive who will own the Key Performance Indicators (KPI) of the project and be willing to serve as a project sponsor? 

Finally, will the C-suite executives allocate resources, staff, and budget for developing and implementing a VA project which can last from three to nine months? For any business initiative to succeed, strong leadership is needed to provide vision and to set the tone of success for the entire team. 

 

2. AI Knowledge

How familiar are you with the knowledge of different Artificial Intelligence technologies like Natural Language Processing? Do you know if AI and NLP are the right solutions to your challenges? 
Are you already utilizing data analytics, voice, and image recognition to stay ahead of the competition? Do your key appointment holders and staff keep abreast of current AI trends by attending forums, conferences, and roundtables? 
Knowing the right problems to solve and matching it with the right technology and change management process are critical to a successful project implementation.

 

3. Customer Engagement Behaviour

Many companies are focusing on the demographics of their customers, how about the psychographics and behaviours they are exhibiting? 

Firstly, does your company offer many social media and chat channels like Facebook Messenger, Instagram, twitter, SMS, WhatsApp or live chat?

Secondly, is your target market actively reaching out to your company through these channels? 

Thirdly, are measuring and tracking their the volume and top reasons for contact for each channel? 

Lastly, do you keep logs of conversation with customers over the phone, email and online requests? 

 

4. Competitive Insight

Keep your customers close, keep competitors closer by monitoring them. It is not about what they are doing, but why they are doing it. If your competitor has deployed a VA, evaluate their system by acting as a potential customer. Go through the customer journey and experience how your competitor engages and converts every interaction into an opportunity. When was the last time you audited your competitors' capability in marketing, sales or customer service?

 

5. Staff and Resources

The team assembled to manage the Virtual Assistant planning and implementation must be a task force of staff from different departments. The customer service and various subject matter experts provide the data input from the customer’s point of view and the correct content. The marketing and communications ensure that the VA represents of the organization’s brand image. The technology team provides technical support for security, deployment and ongoing maintenance. 
The risk and compliance department ensures privacy and protection of customer's and organization’s information. There is the need to ensure roles and responsibilities are well defined both during the project and routine business as usual period. 
Besides manpower resources, it is important for budget allocation to be done not just for project implementation, but also for ongoing maintenance, enhancements, training and to drive the usage and activation campaigns.

 

6. Implementation Process

The roadmap, milestones, and the timeline are essential parts of any project management. All internal stakeholders must be aware of the strategic phases of innovation at least over the next 18-24 months. The Vendor Selection Matrix must be developed to evaluate different solution providers' recommendations objectively. There may also be a Proof of Concept (POC) done to manage expectations on the capabilities and limitations of the VA project at least in the near future. Risk assessment and management must be in place to mitigate any unlikely situations. 

 

7. Process Documentation

While the development and implementation processes are exciting, it is important that the process of documentation is not just seen as an afterthought. Proper documentation of the business case, roles and responsibilities of team members, concept flow, customer conversation mapping, standard operating procedures are important for the future development of the project. It allows new members added to the team to gain a comprehensive overview of the initial initiative. Regular progress meetings are not just another project update – they are also a great chance to fine tune the processes, align KPIs and review new opportunities.

 

8. Governance and Measurement
This final category is about Corporate Alignment, Key Performance Index (KPI) and Return on Investment (ROI).  Are the virtual conversations aligned with your corporate value in delivering service excellence? Does the chatbot sound like one of your customer service staff? What are we measuring? How do you define success? If it is about achieving a higher Customer Satisfaction Score (CSAT), do you know what is your current rating and how much you want to improve? If the ROI is about data collection, are you communicating the insights to Senior Management for their decision making regarding future strategic directions?

Conclusion

Virtual Assistant Deployment is undoubtedly an integral part of a long term sustainable business automation strategy for many forward looking enterprises. While some companies may know what they need to automate, but they may not be innovating fast enough to stay relevant in the marketplace. For companies who are willing to take the next step, and move  fast enough for innovation, the final question is always "Are We Ready Today?"