The Conversational AI Industry is growing and its numbers are promising. The hype around conversational bots is enormous and this causes companies to experience “ Chatbot FOMO ”(i.e. fear of missing out) and are rapidly adopting the technology. There´s no denying that the benefits of this technology are real, but often companies fail to identify and prove their value.
This article will provide an overview of the process for defining and measuring the value a Conversational AI solution has to a business.
Let´s start by defining Value in this specific context. In this article we´ll consider Value as “the importance, worth or usefulness that a Conversational AI Project adds to a company”. By using this definition, we also highlight that value does comes in the form of short and long term impact.
So how to identify it? To add value, a Conversational AI solution should be always tied to a business driver, that is, your company’s** rationale for developing a Conversational AI project. So ask yourself: Why are you implementing a conversational AI solution in your company? This is THE question to ask and it’s also fundamental for defining the business case and scope of your project.
For example, do you want to provide better customer service? Or do you want to increase sales?
Your intention defines your strategy.
Now that you´ve identified your business driver, it is quite easy to see how and when value will be delivered by your virtual assistant. The most common value drivers** **for Conversational AI projects are:
- Improve User Experience (Through increased availability, reduced waiting times, etc)
- Automate Tasks
- Reduce Costs
To tie the concept of value to business driver let’s use an example of a company which its main business driver is to increase sales and has developed a bot that collects user e-mails and generate leads. This company can claim that value will be delivered when the user has submitted their e-mail.
Now that you have a high-level idea of when value will be delivered, it´s time to turn to your solution.
You´ll need to map the business driver to your dialogues and clearly identify at what point in the conversation value is being delivered: Is it when a form has been filled? Is it when the bot has given the same answer that an agent would? Is it when a conversation has been completed with no handover to an agent? Is it when helping to finalize a purchase process on the website?
There are no wrong answers here but only strict requirement: whatever point of your conversation you decide must be directly tied a business driver. Failing to do this will end up on results that don´t add any value.
Try to think about this and clearly locate it in your conversations. But can this be easily achieved? Teneo has the capability to accomplish this using Metadata. By the use of this functionality, Developers can tag any flow node that they have identified as “value adding”.
So going back to the lead generation example above, the flow will be tagged when the user gives answers with their e-mail. The great news about Metadata in Teneo is that it then can be queried and analyzed using Teneo Query Language. This brings us one step closer to measuring value.
Once your solution is tagged with moments in which value is delivered, it´s time to turn it into a KPI. If you have followed the steps so far, this process should be quite simple and straightforward.
We´ll go back again to our example of the Lead generation bot: a good KPI for measuring value here would be “Amounts of E-mails collected”. But how do you get from the flow to the KPI? As we mentioned before, metadata can be analyzed, so by querying the number of times that users provide their e-mails, the company can now have a clear KPI to include in a dashboard to assess the impact the bot is having.
When defining KPIs, remember to keep them SMART and also avoid confusing them with Performance oriented KPIs (like Bounce rate or Session Duration). A bot with a 0% bounce rate does not necessarily add value if it´s not reaching the points in the conversation that you want them to.
At this point in the process, you should have a list of clearly defined KPI´s. However, there is an additional step you can take. Try to consider the added impact that these KPI´s could have on other areas of your business if they were to be achieved.
Once more, we can use Lead Generation Example to highlight this point. Knowing that the bot has “collected X amount of e-mails last week” is clearly valuable information, but how is that helping to increase sales? To assess this, the company could track how many of the e-mails generated by the bot end up becoming customers by crossing e-mails with a CRM database.
By doing this, the company´s Business Driver for increasing sales gets tied directly to the Conversational Bot, and this company can clearly measure and quantify the value their conversational project adds.
This exercise becomes quite straightforward when it comes to assessing monetary value, but as mentioned in the beginning of this article, keep in mind that non-monetary value can also be addressed with this process. If your business driver is improving customer experience and a relevant factor for your company is time, you could for example measure the time a user takes to complete a request using a virtual assistant versus the time it would take to go to an office or make a phone call.
Finding a precise ROI for your Conversational AI solution is critical for its long-term success.
Every company, conversational AI project and business goals are unique. Rather than being about KPIs and measurements, the key for clearly assessing the value a conversational AI solution has relies on clearly aligning your chatbot to your business strategy, identifying your key value drivers, map them to the corresponding points in your solutions in Teneo and measure them.
What are your ways for assessing the value of your solution? Let us know on the comments below
#teneotuesday #conversationalai #ai