In a previous article, I briefly explained (as one could reasonably spend a few pages on it) what a metabot is. I invite you to follow the link to dive into this – fascinating – reading.

Having established this difference between the two types of solutions, the question of the necessary investment and its immediate corollary of the return on investment naturally arises.

Rather than an article full of theories, it seemed interesting to me to report on the various customer exchanges I had during the presentation phase of’s offer these last few weeks.

Observation #1: Short vs. long term vision

Interesting reaction from a CIO to’s presentation: “Your solution goes beyond everything, but we need to rethink our project. We are no longer on a one shot but on a more general implementation there, we did not expect this possibility. For the chatbot was a test, if it is not conclusive, we throw away… “.

The thinking around the implementation of a chatbot is often focused on bringing a specific functionality to a specific population. Whether the functionality is linked to the support of a business function, the search for information in a specific context, the provision to users of a modern interface, or any other reason, it is undeniably linked to a desire to modernize an IS. When we come to present the “metabot” option, the thinking is much broader. The stakes are also much higher. They require a more global vision of the IS, but not in the IT sense: rather a strategic vision, a direction to give to its use.

But whether we’re talking about a “nice to have” feature or a more strategic application, thinking about the implementation of a disposable solution is perfectly legitimate. Knowing what budget can be sacrificed to implement a test helps define the scope. But how will the success factors be defined/measured?

Observation #2: Integration time for a high visibility project

This same CIO then made a very pertinent remark: “The time it takes to gather all the business needs in our company is a two-year project”.

True… and in some cases totally false. It’s all about implementation. If indeed you want a global redesign project, then yes, it takes time. And money (the people mobilized internally are not necessarily expendable). Agility being advocated everywhere, the integration of a metabot solution must follow this path: on the one hand, because it will allow easy deployment of pilots that will ideally go and promote the product internally, and on the other hand, because it will allow spreading the costs over the duration of the project. And in terms of change management, addressing the organization’s population by parts has always proven to be easier than a general change.

Observation #3: direct vs. indirect ROI

One point raised during many meetings is that of ROI. More specifically, the evaluation of the indirect ROI. In my opinion, this is a critical point in the choice of a solution. The direct ROI can be very simply calculated: the solution costs you X, it allows you to save Y, simple calculation. But is this the case for a chatbot? In the case of a B2B chatbot, yes, it is possible. A chatbot addressing support functions can replace level 1 and/or 2 operators; a chatbot for HR functions can automate certain tasks and therefore the ROI in terms of time spent can be calculated; etc. But what about the time spent searching for information? How to evaluate the impact of the adoption of the IS environment? How to evaluate the impact on user satisfaction?

Indirect ROI is also influenced by the tool’s freedom of use. If users can appropriate the tool and IT no longer has to invest in projects for the implementation of micro-projects or if the tool only remains in the hands of certain teams, then the valuation of this indirect ROI can change.

What about Hubi? is a complete metabot platform and comes with all the necessary tools to minimize licensing and maintenance costs.

The metabot provides the following elements:

  • A complete administration portal including governance, usage reporting, GDPR compliance…
  • A module store that allows you to add new functional scopes to the solution with a single click, and to extend the modules to adapt them to your context
  • A no-code model for customizing and extending modules, whether they are simple knowledge bases or more complex business scenarios
  • A universal connector to interconnect with any application that has an API

And that’s where the “magic” of the metabot comes in: thanks to this universal connector, you can also interact with chatbots that you might have already deployed in your environment… as long as they have an API and you’re happy with them. Alternatively, migration to the model used by the metabot is also possible, as here too, under certain conditions, is planned.’s licensing model is also designed to ensure that you have control over the costs of the solution. Rather than a model based on activity, which is difficult to predict, metabot is based on a subscription model. No hidden costs, everything is included (Azure hosting at, continuous improvement of the store content, support). Our partners will be there to help, if needed, to set up specific scenarios., metabot platform, allows you to serenely consider the integration of a metabot solution in your IS.