<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=275133924453494&amp;ev=PageView&amp;noscript=1">

Try Solvemate without risks! We are willing to guarantee your satisfaction, or your money back

Learn more

Go beyond the chatbot hype and boost your customer service with automation to improve your CX.

Discover our features to create meaningful conversations in an instant, scalable and cost effective way.

We integrate with the best! Solvemate is built to fit into your existing tools and workflows. Find out more.

Nov 7, 2019 11:45:00 AM

5 steps for a successful chatbot project


Many companies are interested in leveraging AI for their business, in particular for automating their customer service. It’s however not really clear, which aspects should be taken into consideration in the project discovery phase.

To this end, we’ve collected five crucial questions that will help you assess if a chatbot is in fact the perfect solution for you. Ideally, this should be done before you enter discussions or start a project with a chatbot provider. 


Any business project can go sour due to unrealistic deadlines and scope creep, but it’s particularly important to be mindful about the potential project risks when venturing to a whole new area of technology. So, to ensure your chatbot project is a success, go ahead check out our infographic and follow these prerequisites in your own chatbot project!



 Learn more about Solvemate's solution


1. Make sure your service organisation really needs one


With more than 6 out of 10 U.S. consumers saying they prefer to self-serve when they have simple inquiries (American Express 2017 Customer Service Barometer), the desire to get a chatbot to complement your service channels seems like an easy enough way to go.  


But before anything else, you still need to validate the business need for a customer service chatbot in your organisation. Chatbots are a particular technology, fitting some use cases better than others - not every customer service organisation has a justified need for one. 


The use cases generally fall into two categories: 


  • A professionally run call center with at least 15 support agents
  • A fast growing start up without existing customer service structure


For both use cases, the fundamental issue is usually large quantities of repetitive questions with relatively simple answers. In addition, an established customer service organisation will have their typical problems: high staff turnover & cost of training new staff, global customers in a variety of timezones and languages, and general demand fluctuation. 


The startup might share some of these issues, even if at a lesser scale, and setting up a chatbot as a “gatekeeper” to the customer service will help curb the growing pains. With a chatbot controlling the number of inbound contacts, the number of human agents can be scaled up as needed, without sacrificing the quality of service.



2. Define quantifiable goals


Even though artificial intelligence and machine learning are relatively new technologies, there is in fact no reason to look at your chatbot project as anything other than a normal business project. It follows all the same phases through the project timeline: initiation, defining requirements & planning, execution, performance tracking, closing.


When defining the requirements, think long and hard - what do you want to achieve? It brings little value to your organisation to just say, “We want a chatbot.” You need to clearly quantify what is expected as an outcome of the project.


Set goals such as “reduce contact ratio by 30%” or “decrease after-sales calls by 40%”. According to Gartner, 64% of people find customer experience more important than price. So, depending on the stakeholders involved, you might want to prioritise customer experience as a KPI for your project. 


In some cases, the quantifiable goals might be harder to set without hiring new staff, f.ex. when servicing new markets in new languages. In such cases, the project KPIs might depend on the general business KPIs (such as revenue) coming from those new markets. 


Whatever it is, make sure it’s clearly defined - with numbers, if possible. You need to know what makes your chatbot a success.



3. Calculate ROI


Like any business project, a chatbot needs to yield a positive ROI. All chatbots first incur costs, so it’s an important part of the project discovery to calculate their ROI before venturing towards execution and implementation.


Calculating the ROI is perhaps not as simple as it is for other projects your organisation has carried out, but it can be done. In fact, we’ve created a handy online calculator to get you started. 


Roi-Calculator-Teaser-Button_2 (1)


Please note that this calculation only applies to text-based customer service chatbots; it cannot be used for voice-based chatbots.


When performing your ROI calculation, take particular focus on the following key parameters: salaries, number of requests & number of agents, as well as the overall percentage of the requests routed to the bot. 


If you need a more detailed overview, you can use an Excel-sheet we’ve created for an in-depth calculation. For example, to get a full overview of the potential savings and ROI, you need to know about the types of requests that end up in your service queues. Gather your stakeholders in order to differentiate between the “self-service requests” (f.ex. available payment methods or resetting password) and the “need-agent requests” (f.ex. cancelling an order after the normal notice period). Use our guide for a detailed step-by-step explanation on how to use the Excel-sheet & perform the calculation.


However, remember that calculating the ROI is just the beginning; according to American Express 2017 Customer Service Barometer, consumers are willing to spend 17% more to do business with companies with excellent customer service. Implementing a chatbot and improving your service levels could have valuable long term implications, a positive impact on your brand reputation, and ultimately improve your bottom line. 



4. Integrate it in your customer journey


Do you have the power to integrate the bot into the customer journey? A key factor of a successful chatbot is exposing it to as many people as possible. It’s easy to direct all the focus and allocate resources to the discovery and training phase - and forget that a crucial part of the implementation is how the bot is integrated. 


Consider the journey from the very beginning. How and where your customers will find and interact with the bot?. If possible, involve a UI / UX / CX expert or a service designer in your project to make sure the chatbot is put center stage in your customer service strategy. The whole point of the chatbot is that your customers can see and use it - prioritise it as the first point of contact your customers can have with you and promote it to them at the crucial moments in their customer journey.


Don’t forget that you can also integrate many chatbots on your social media channels, such as Facebook. As the average wait time on social media is nine hours (Shep Hyken, The 2015 Customer Experience Outlook), using a chatbot to instantly solve a majority of the requests is a big opportunity for differentiation. 


Being mindful about how your chabot fits into your customer journey will help you see your project through to success. A chatbot cannot yield a positive ROI if it’s not being used. 



5. Start small


Given that after one bad experience, 39% of customers will never do business with the offending company again (New Voice Media 2018), your stakeholders might be impatient and want big solutions and big numbers from the get-go. Especially for larger organisations dealing with bigger projects, it can seem counterintuitive to start small. 


It’s important to understand that due to its conversational nature, a chatbot is an iterative tool. The project should be kicked off with the top 20 most repetitive requests, as only the live bot will tell you exactly how your customers interact with it. It would not be wise to expect to provide 100 different solutions to your customers at an early stage. In the end, you are not looking to replace your customer service with the bot; you are looking to provide the right help, at the right time - instantly. 


So, it is recommended to start small; first train a few solutions and a few problems and incrementally iterate your chatbot, expanding both your question & solution base. If your chatbot uses machine learning to improve itself, it will gather feedback and suggest further training predictively and autonomously, making the process quicker and easier to manage.


Hungry for more? Get our latest handbook:

Beginner's guide to customer service chatbots. 




Sara is a former Solvemate. She’s really into chatbots, and improving customer experience. When she’s not writing about customer service automation, she’s an Italo-disco singer and a devoted housekeeping nerd. Hailing originally from snowy Finland, the Berlin winters leave her cold (pardon the pun).