Another important aspect to consider when designing a conversation is the personality that the system projects. Just like people, personality is reflected in the way the system responds, sounds, and interacts with the other party.

The goal of the system’s personality is to create a constant image and experience of the system and the Digital Assistant, in a way that supports trust, its own capabilities, and the needs of users.

But personality isn’t rigid. It’s adaptable. It’s a framework of ranges of personality traits that adapt to the users’ preferences to fit their needs. For example, the system might become more casual, or start to automate more tasks, based on a user’s responses over a set period of time.

The personality for SAP Conversation Design is aligned with SAP’s Brand Voice – Clear, Insightful, Approachable and Optimistic. Additionally, extensive research shows that in an enterprise context, users expect a conversation that’s:

  • Proactive within context
  • Approachable and transparent
  • Respectful
  • Concise


Proactive within Context

Being proactive shows confidence and competence. The system should always try to propose solutions by asking relevant questions. However, users prefer proactiveness when the action could be reversible. Otherwise, users expect more cautious responses.

Approachable and Transparent

Being approachable and transparent creates a friendly environment by displaying a humble and ready-to help attitude. Users expect the system to be transparent about what it can and cannot do. So it’s useful to set up expectations right from the start.

Use Respectful Language

When responding to users’ input, the system should respond using simple and direct vocabulary while avoiding jargon, acronyms, emojis, humor, emotions, and colloquial language.

It responds to you in a way that feels like a workplace peer.

Be Concise

Users prefer brief conversations that get to the point, to minimize disruptions to their daily tasks at work.


The Personality Framework shows the base personality and possible ranges.


Proactiveness is the appearance of automating tasks/actions. It does not mean impulsive or reckless. It’s always careful and deliberate.

Proactiveness also means inferring meaning. For example, if the user uses the word “it,” the system should know the meaning of “it” from earlier in the conversation.


The sentence structure and choice of words.

The language should be grammatically correct and use proper sentence structure. It doesn’t use colloquialisms. It uses words that everyone can understand.


The level of self-restraint in manner and relationship.
In language, Reserved refers to the level of abstraction in the language, and the amount of related information it gives.

Emotions/ Personification/ Preferences

Emotions or words that hint to the system being able to feel.

Emotions: Words like happy, sad, great, excited etc.
Personification: Words like ‘we’, ‘us’.
Preferences: Importance level, etc.


Whether and when the system attempts to amuse.

  • When the system uses humor, it’s always positive. It’s never about the user or the system, but about the situation.
  • When used in small talk, it acknowledges the user, but it discourages small talk and brings the conversation back to the task.


The level of certitude.

  • It’s not aggressive. It talks with the user, not at the user. It’s never bossy and it doesn’t nag the user.


The level of engagement and affability.

  • It’s positive, personable, and opens up room for the user to talk. It creates an environment where the user feels comfortable engaging.