The Chatbot Hierarchy of Needs

apple-email-iphone-17663 copy.jpg

In the 1940’s, Maslow defined the hierarchy of needs: for humans. 70 years on, automated agents, or bots, are commonplace. We interact with programmatic responses every day and the implementation of these automated agents is a multi-billion dollar industry. So with bots now a critical part of the global workforce, isn’t it time we imagined a digital hierarchy of needs?


Deficiency vs. Growth needs

Maslow determined the first four needs as deficiency based: i.e. should you not have them, they become a motivation when they are not met. Let’s look at how this theory works in the world of automated agents:

Physiological needs

In humans, our need for water, shelter, food etc is clear. For bots, the basic requirements to function are still discernible and without one element, the

  • Platform. Somewhere for the bot to be present: this could be messenger, LINE, Google Home or any other environment where a user can interact
  • Content. Being connected to information: be it static (an airlines product), context based (a person’s PNR or FFP status), dynamic (pricing or operational info) is critical to a bots value.
  • Language. Natural Language Understanding. Mostly freely available and often taken for granted, like water, without NLU a bot cannot respond
  • Context. Knowing the difference between the context of language: baggage, carry-on vs check-in
  • Trained. Like rest, bots need training to remain fresh, relevant and up to date

Safety needs

Defined for us humans as security, or protection from the elements. When we hire people, we ensure their safety. The same goes for our digital colleagues, too: a digital layer of protection and security has to be added.

  • Compliance. GDPR and other regulatory frameworks are there to protect users, good bots build trust by ensuring that they comply and remain compliant as regulations change.
  • Structural. Depending on the purpose of the bot, the experience may require it to touch sensitive information. Being set up in such a way that anything sensitive or private is safeguarded is a minimum. Examples of this from a UX perspective is the use of web-views or ‘conversation extensions’ that shield and protect data.
  • Reliability. 99.9% uptime. As a standard that means 24-7 monitoring and ensuring that the infrastructure is build to scale way into the future.

Belongingness needs

The first of two ‘psychological’ needs. Bots, just like humans need to feel part of something and to know that they have a clear responsibility.

  • Purpose. Bots need a reason for being. The best bots work because they have a specific set of purpose, or scope, which guides their UX, content and flow.
  • Committedness. We’ve seen companies work on automated agents, give them the basic physiological, esteem and safety needs but then fail to commit to them as a channel. What happens when a great bot is built and then becomes part of the core brand proposition? It becomes meaningful as a communication tool and then delivers on the next set of needs..

Esteem needs

Maslow refers to ‘feelings of accomplishment’. As humans we take great satisfaction from a job well done. Bots going from being proficient to being capable and then to feeling ‘prestige’ is about measurement. Setting out to measure the effectiveness of your bots, with KPI’s that are meaningful and relevant to the utility and purpose you’ve set out is the only way to understand the bots performance.


By celebrating the results of a bot, through the improvement in end-to-end automation, or the time and value of available hours for the bots human colleagues or (insert other example), the greater the recognition of the robot colleague and by turn, the greater the investment.

Growth needs

Having a bot that has satisfied its deficiency its needs is an achievement in itself. There are many examples of companies that have implemented bot technology, ensured the bots basic needs and then stopped. Just when things begin to look interesting. Moving beyond the basics of implementation, useful utility, assimilation within strategy and measurement is crucial to getting the most from bots.

In the same way that we humans need to realize our potential to feel fulfilled, so do bots.


Often, our partners have come to us with one or two specific use-cases. It might be in customer service automation or post-sale engagement, notifications or retail. What typically happens is that as the basic needs are met, from the basic physiological requirements through to the esteem of the bots delivery on task, is that people begin to ask “what more can we do?”

Increasing the scope of a particular use-case is mostly self-determining: users ask a lost-baggage bot often enough about baggage allowance and good UX developers will begin to add the capability to answer. This actualisation by increment is an easy way to build long term, purposeful bots.

Where bots start in silos, such as marketing vs. customer service, broadening a bots scope can often be problematic when faced with competing demands. Where we see the best results is through collaboration, with long term enterprise goals being mapped out into phases that represent the best scaling from the existing deficiency goals.

A solid foundation for the future

With robot colleagues being ever more prevalent in ever more industries, understanding what makes them successful is crucial for understanding what further use they can be.

As the technology behind bots improves, we will see the basic deficiency needs of physiology and safety become more hygiene as the infrastructure to support them becomes easier to operate and more scalable. Where companies want to utilize bots properly, the less straightforward, psychological-equivalent needs become more important. These are the less scalable requirements that need guidance and experience.

Actualisation is something that is often alluded to as part of digital transformation strategies. In reality, the process of working from a solid foundation of understanding and applying it to more and more adventurous and meaningful use-cases.

At Caravelo we build chatbors for industry leading travel brands around the world. If you'd like to learn more about how we could help your company with conversational commerce feel free to get in touch here .