Data can be used to expose an affinity to a brand, gender or category. It can also tell us about a device and its location. But what can data really tell about intent?

User intents are typically extracted by mapping linear sequences of events whose realisation may be a purchase, and whose previous step would have been an add to basket. Intent is implicitly explained by these sequences; the higher the repetition of a sequence, the higher the intent.

A problem arises, however, when human intent is explained in the same terms as the intent of lifeless objects. Humans and users are not the same thing.

To better understand this problem, I encourage you to ponder the way objects exhibit their intentional states, for example, a thermostat that operates unmysteriously.

In the ordinary physical world a thermostat installed in the street will respond to the particles around it, as a direct correlation with simple causation, in absolute terms – meaning that what caused the thermostat to rise did so in isolation of other possible causes.

Nonspuriousness is an essential quality behind this simple form of causation, as it needs to be established that no other variables participate in the cause-effect relationship.

A user’s state of mind is left behind when describing intent.

Since mental states (and the way humans conceptualise their world) are beyond the data’s scope, the digital industry decides to ignore them and instead conceive information systems with no consideration to the emotive markers behind intentionality and human behaviour.

Everyone has difficulties in understanding feelings. But this does not justify that we reduce them to a linear interpretation of intentionality, one that makes the following assumptions:

  • No completion (i.e. not checking-out) means that no intent was present.
  • Not following the previous steps (i.e. not clicking or not adding to cart) is an anti-intent declaration.
  • An intent is one thing alone.

Such simplicity in interpreting user intents is similar to the simplicity that is applied to infer the intentional states of objects.

User intent must be recognised as a kind of intent that exists through human behaviour, hence different from the intentional state of an object. If this distinction is not made, professionals of the digital world are put in a data hole, where they can’t escape the limitations contained within that view which unaided data can see.

Most digital professionals have given a resolute cold shoulder to emotions in the interpretation of digital value, neglecting feelings because of data’s inability to understand them and deciding to downgrade humans to the qualities of lifeless objects.

We have codified human intentionality to bring it down to data’s capacity, as if making users comprehensible were to give us an entitlement to forget our own unfathomable nature.

Nevertheless, the ground has already begun to shift. For example, UIs have started to be recognized as causal to the elicitation of positive perceptions in key fields of digital, such as Information Retrieval.

Naturally, the incorporation of feelings into the interpretation of digital value represents a discomfort to many, as it requires us to break into the realms of neuroscience, neurobiology, psychology and even philosophy and cultural history: Disciplines that on the other hand are an inevitable part of any effort to understand “users” — who are, after all, “humans.”