Last year when I first got into this whole chatbot topic, I was a little skeptical what value this could actually add to people’s lives. All of a sudden, they were everywhere, everyone was talking about it, you could find them on a lot of websites. But, to be honest, I haven’t really been using them much in the past. It was just hard for me to think a machine could answer a question with the same quality than a human.
So, in order to evaluate this new hype, I started building chatbots on my own via tutorials that showed me how to use the technology. While helpful to get to know the platform, the use case was quite simple, and it wasn’t really adaptable to my day-to-day life, that was cute but not very satisfying. After this, I was thinking, In what part of my life could a chatbot actually help me?
Chatbots in Our Day-to-Day Lives
Baaaaam, suddenly ideas kept popping into my head. If you think of the most annoying things in your life, the question is actually quite simple to answer. First thing that popped into my head was to create a chatbot that answers all the questions that my fiancée asks me repeatedly.
- When are you going on a business trip again?
- When does our flight leave again?
- What’s the flight number again?
- When are we meeting with your parents?
- When is your birthday again? 😉
(The last question he actually hasn’t asked yet, if I am being honest, but who knows, we’re not getting any younger.)
So, after all these questions raced into my mind like fireworks, I was thinking this is going take a lot of time to realize and it requires too many integrations with systems like outlook calendar, several apps on my phone, and too many synapses in my mind. I can also think of a 1000 other use cases like a chatbot for the Citizen Registration Office that helps people to go through the German jungle of documents, websites, rules and regulations, etc.
What about Our Professional Lives?
So, after more or less thorough evaluation of this use case of my private life, I went to my professional life where I got my colleague Sarah involved. Of course, there are several examples that come to mind in terms of HR, Marketing or IT. Helping an employee to set up a leave request, change the personal data or just simply answer an IT Service Ticket are great use cases.
But we wanted to go more into the challenges of my specific role. We work in presales—which is an awesome job—but it comes with some challenges. Our job is to go to the customer with our sales colleagues and present whatever solution we’re specialized in and the customer (hopefully) is interested in. This requires very close collaboration with our sales colleagues. The process usually goes like this:
- Customer shows interest in a topic because they have a problem that needs to be solved à Sales qualifies that interest and evaluates which the solution is the best fit and can solve the problem best à Sales calls the corresponding Presales colleague and asks them if they can come to the customer and present the solution.
Sounds pretty simple, right? Wrong! See, there are two process steps that we should have a deeper look into.
First let’s have a look at the second process step: the qualifying.
- That is especially important for us in presales because we should only get involved in a deal if the opportunity has been qualified thoroughly. This means, the problem was properly discussed and understood from both sides to ensure we suggest the best solution. So, both sides will not waste any time looking at products that only fit half to the problem. It also needs to be ensured that the people we are talking to are actually the right stakeholders, e.g. the ones who will implement or later on use the solution to help them with their daily work. The customer has thought about a use case or an operational scenario. This is just to name a few of the questions.
Now, let us have a look at the third process step: the calling.
While easily executed from anywhere, this is not the way how sales ideally request presales. Because it results in:
- First of all, important information about the deal can get lost by being passed on via the phone or in the hallway without any documentation and there is a high risk of leaving out important procedures like getting the Presales Manager involved or putting the presales support into the system. Or just plain and simple misunderstandings. This is actually crucial because if it’s not documented somewhere that Presales was involved in a deal, it creates big in transparency.
- Second of all, this way the involvement and customer meeting history are not documented and could get lost. We all know how it is when things get stressful, the documentation is the first thing that we forget. So, we could end up involving various colleagues to work double. No person likes that. This also has another side effect: people that are new to the team will have a hard time getting involved in deals because nobody knows them.
In the end, the non-transparency of the execution of those two process steps results into presales doing a lot of unnecessary work and unhappy customers.
Chatbots and the Qualifying Process—A Use Case
So, there we are, confronted with this amazing opportunity of optimizing a process that could actually improve our life and make day-to-day processes easier. A chatbot would be able to cover all these questions for the qualifying process even before sales talks to any presales. It can bring some transparency, standardization, and a free documentation into the process.
In order to get this project going, we started with first creating a conversational flow, which basically means, thinking about how a conversation between a human and the chatbot would go. How does the chatbot react when this request comes in, how does it react when the user of the chatbot says no, how would it react if the user answers yes and so on and so on. You really need to think about every possible way and outcome of the conversation.
Then we thought about which systems need to be connected in order to get or write back certain information during the conversation. And there’s another big potential here, because while a human needs to get into the system finding an information or filling in a text field, a chatbot can just do that automatically within seconds.
Afterwards we talked to a couple of colleagues about this idea and now we build this little team together that is working on this solution. To build just the first prototype took us one day. So, you can see, this technology takes little time to realize compared to other software projects.
After going through this experience, we are not only absolutely convinced of the great value that chatbots can add to peoples lives, but we also came up with an idea to improve my way to work.
Try It Yourself
- If you also have some ideas in mind that can help you in your life (and let’s be honest—you do) check out our chatbot platform: https://cai.tools.sap/, sign up for free and start prototyping. It is easier than you think.
- This might help you to get started: https://cai.tools.sap/blog/build-your-first-bot-with-sap-conversational-ai/