What resources does it take to launch an internal chatbot?
Many people wonder about the complexity and resources needed to launch an internal chatbot in their company (IT support, HR support, legal support…). Let’s explore this topic together here.
"In the past", it was quite complicated to start a chatbot project
And reserved for large companies.
Let me take you on a journey through contemporary history…
Chatbots burst into our lives during the 2010 decade. The GAFAs are banking on this technology and successively launching their virtual assistants (Siri for Appel, Google Now for Google, Cortana for Microsoft, then Alexa for Amazon…). On the Asian continent, it is the giant WeChat (more than 700 million users per month at the time) that is pushing the rise of chatbots by making it possible in 2013 to create bots evolving in its messaging by third-party developers.
On the corporate side, large international groups are equipping themselves with complex chatbots for the time, based on artificial intelligence technologies such as IBM’s famous Watson, with the help of consulting firms and IT services companies. Budgets are often above 200k€ for a single bot, and projects are carried out over relatively long periods of 1 to 3 years.
The chatbot hype probably reaches its peak around the years 2016-2017. Many customer services departments are equipping themselves to handle the ever-increasing volume of online solicitations.
In the early 2020s, the sector reached a certain maturity. Organizations working on operational excellence in their various divisions are increasingly integrating internal bot to assist their teams. In the first place, this type of project focuses on the processing of repetitive tasks with an obvious automation potential, but sometimes also covers complex use cases involving powerful workflows.
But today, launching an internal chatbot is pretty simple!
At the time of publishing this article, bots are not only more powerful technically, more ergonomic, but also simpler to set up. At least, in the case of publishers who offer an intuitive “no-code” creation dashboard and a monitoring interface with performance indicators.
This little robotics history lesson had only one goal, you will have understood: to make you understand that if launching a chatbot was certainly relatively complex and expensive not so long ago (in the middle of the 2010s)… the ecosystem has largely evolved.
Today, chatbots are becoming more and more common and the majority of players dedicated to companies are specialized on one or more use cases (customer relationship, IT support, legal…). This specialization allows them to master the challenges of the projects they accompany from start to finish, whether on the technical aspects, but also and above all on the deployment and usage aspects.
Where to install its internal chatbot?
The answer to this question is often obvious.
The rule is simple: install the chatbot where your employees spend most of their time. No, not at the coffee machine… I was thinking more of the instant messaging used in your business.
Do your employees use Microsoft Teams to communicate? Install your bot on Microsoft Teams. Do they use Slack? Set up your bot on Slack. They use Webex… In short, you know what I mean.
For example at Vizir, you can install your bots in a few minutes on :
- Microsoft Teams
- Skype
- Sharepoint
- Webex (Cisco)
- Google Hangouts
- Slack
Less common in theory for an internal bot, an installation is also possible on a web widget, Facebook Messenger, SMS or Voicechat.
Vizir also has a module that allows you tointegrate a chatbot directly into SharePoint. This is rare enough in the market to be noted. The plug & play integration is extremely simple, 3 minutes are enough for the person with sufficient administrative rights on your instant messaging application.
How much does a chatbot project cost?
This is a good question. Much better than the question “how much does a chatbot cost” which would not take into account the real cost of the project revolving around the chatbot product itself, potentially mobilizing internal employees with the salaries that go with it.
Anyway, we’ve answered this budget question in a very comprehensive article that I invite you to read here: Article “How much does a chatbot cost?”.
What resources should you allocate to your internal chatbot project?
As recommended by Gartner (see our article on developments and best practices in chatbot projects based on the conclusions of a Gartner report), if your company’s core business is not close to machine learning or data science topics, it is strongly recommended, in order to reduce the risk of project failure, to call upon a specialized provider. We will focus here on this case, which is the most frequent and which avoids investing 200 or 300k€ of salaries (internal or by calling a consulting firm) simply for the ad hoc creation of its chatbot.
If you are using a provider who gives you a turnkey service, the resources required should be moderate.
At Vizir, as an example, we solicit the following people in your teams:
- A “Bot manager” who has the knowledge of the business that we are trying to relieve with the bot. It is the person who will be able to quickly identify the most valuable content to be prioritized in the bot. It is requested
about 1 hour per week for 2 months, and then only 30 minutes per week on average
.
- A marketing manager, who will be in charge of giving the bot its personality and communicating internally:
a few hours are enough to start the project
; afterwards, it is not solicited anymore.
- An API manager, who is competent in the tools that must be integrated with the bot to set up an automation of the processes (if the bot project provides for this):
half a day is generally necessary to start the project
.
That’s it!
For the rest, everything will be handled by our Technical Account Managers, engineers responsible for accounts, experts in NLU and chatbots.
So, in the end: complicated or not?
As we have seen, the complexity and resources required to launch a chatbot project are now much less demanding than what we experienced in the 2010s.
The simple installation on a communication channel already used daily by employees, and the support of a specialized service provider greatly facilitate the launch of chatbots without mobilizing the company’s internal resources.
