/ Google Chatbot: a look at Dialogflow
Alexander Wijninga
|
4 min
Chatbots are mainstream in customer service. Alongside customer service chatbots there are also marketing and sales chatbots. They catch first-line customer contact, or support the sales or marketing department. The Google chatbot recently changed its name: from API.ai to Dialog Flow. In this article you can read how Dialog Flow works and whether it suits your organisation. Read on and discover both the pros and cons of Google's Dialog Flow.
The Google chatbot builder, known as Dialog Flow, makes it possible to build FAQ bots easily. Via Dialog Flow you build chatbots, including an integration with voice devices via Google Home, or Amazon's Alexa. But Cortana, Twitter, Slack or Skype can also be integrated with Google's software. Google's chatbot offers extensive possibilities provided you have the right budget (and the right amount of time). Google works with NLP (Natural Language Processing), a form of artificial intelligence.
Because of the low learning curve the Google chatbot offers, you easily start building your own FAQ chatbot. This is easy to do by uploading your existing FAQs. You fill in the frequently asked questions (including question variations), as well as the matching answers the bot offers. This is quite a job, though; without AI this manual task is a considerable time investment. An advantage of this service is the easy integration of Google Home or Amazon's Alexa. Because of this easy form of voice activation you will soon want to get going with voice instruments yourself.
With the Google chatbot you have access to a wide range of choices. These extensive possibilities are not easy to realise, though, when you do not have the right (programming) knowledge. When you want to build a dynamic app you need what are known as Fulfillment hooks. For this you need quite a set of technical knowledge. This hook is a connection to an external system, with which Google retrieves information. When you want a bot that retrieves order information or browses through your webshop, for example, you need this. For the info that is then retrieved, entities (comparable to categories) also have to be created, otherwise the chatbot does not understand it. A big disadvantage is that each of these has to be entered manually. So you spend a long time entering only part of your bot.
With the help of integrations from third parties such as Slack, Google Assistant and Facebook Messenger, you build a chatbot without programming. When you do not use these integrations, though, you do have to write code. You also have to interact directly with Dialog Flow's API. Do this with every conversation; that way your digital assistant understands the right matches and nuances. Writing code is a complex matter for the average entrepreneur or organisation. Dialog Flow does offer you some assistance in their starter guide. Besides this starter guide the platform does not offer you many (free) lifelines. It is therefore important that your knowledge is in the right place, or that you take account of bringing in a developer. When you get going yourself the possibilities are great, but so are the time and costs you put into the build.
Dialog Flow offers extensive solutions for enterprises. With a development team in house you build your ultimate chatbot entirely manually. For smaller companies this step is harder to accomplish, though. You soon need programming knowledge to realise what you have in mind.
You get support in building your own Dialog Flow chatbot in the form of content. When you have a question, or want to speak to a member of staff, you can only turn to the external integration partner. Do you work with a Facebook Messenger integration? Then that is also your point of support. Large organisations will experience less of a problem with this. With in-house developers and a team working on the bot, problems are quickly cleared away. For the small and medium-sized entrepreneur this is often different. An alternative to a chatbot with broader support is a chatbot platform, such as Watermelon. Here Customer Success Managers stand ready from the start of the process to support you as true chatbot experts.
The costs of Dialog Flow are calculated on the basis of different categories. Google charges $0.002 per interaction request via text. When you make a voice interaction request, this costs $0.065.
Starting with a Google Chatbot goes as follows.
Open the website. Click on ‘go to console’ and then log in with your Google account, or create a new account.
In the menu on the left-hand side you then see ‘create an agent’. By clicking on this you bring your first bot to life.
Give it a fitting name. Do not forget to enter the right time zone and language either.
Go to ‘create’. Congratulations! The beginning of your first chatbot is ready.
Now you are going to start entering the bot's content, code and intents. This process depends on the type of bot you choose. You can read how this process works here.
Dialog Flow offers very extensive solutions, provided you have the right knowledge. The program works intuitively and by choosing Google you ride along on the company's success. When you want to go one small step further than just an FAQ bot, though, then building it is not for the layperson. Do you have the right programming knowledge and do you enjoy letting your creativity loose within the build possibilities? Then Dialog Flow is certainly something for you. When you do not have programming knowledge, though, you benefit more from a chatbot platform with a code-free chatbot builder. In this you only place the content, and the AI arranges the rest. Are you curious about building your own chatbot, without programming? Then take a look at Watermelon's chatbot builder here.
Would you like to know more about successful chatbots? Download the free whitepaper and read about the success stories of Stichting BKR and Hunter Douglas (Luxaflex), among others.
Share this article
