What We Should Learn From Machines: How Metadata Drives Effective Feedback
Hyperconnectivity rules our lives. Approximately 330 billion emails are sent per day. 19 billion texts (not counting in-app messages). 2.4 billion phone calls. Be it business or personal contexts, we are expected to be available on a near-constant basis.
The systems we’ve built are designed and maintained to enable communication on a scale and immediacy never before seen in history. Entire new languages, their accompanying standards, and the logic of their grounding exist purely to enable the transfer of information that is never intended for direct consumption by humans.
But what can be learned from their approach?
Metadata
Simply put, metadata is data about data. From summaries and categories to a data object’s information structure or its administrative restrictions. An array of self-references governs how data in the modern world is proliferated, stored, and retrieved. It’s not an exaggeration to say that without good metadata, and standards to govern it, our information systems could not function.
Take an mp3.
Included in the file’s data will be the track title. The artist. Genre. Perhaps when the track was originally released, or the creation date of the file, or its listing number within an album.
None of this information is intrinsic to the track itself — the data which encodes sound waves in digital format — and yet without it, searching for the track or usefully interacting with it would be extremely difficult.
But what about personal communication?
The Importance of Context
Subconsciously, we use contextual metadata in our day to day speech. “Adding context” enables our audience to better intuit which specific objects in reality our words refer to. It enables symbolic representation to become precise and the efficiency of language to increase in use. We seldom have to offer highly specific descriptions of people or things, and when we do, it usually represents a knowledge gap between those communicating.
Let’s take a quick look at that in practice:
“I was speaking to Jamal from Accounts yesterday, and…”
What were they doing?
Having a conversation.
With whom?
Jamal.
Which Jamal?
Jamal from Accounts.
When?
Yesterday.
None of this information is intrinsic to the message (the conversation that took place), but might be necessary for interpreting it.
Given the assumption that the two participants in the conversation work in the same organisation and their circles overlap, the information given in the statement will be easily understood. In the event that they don’t, further details (metadata) about Jamal must be provided.
This all seems very simple. Intuitive to the point that we process these decisions subconsciously when offering information, and seldom notice the constructions when receiving it, unless the order or context of the information poses a challenge to interpret. Failure of communication becomes easier to spot than good practice.
But are we using our information constructions as efficiently as we could be? On a structural level, are we good at communicating?
Feedback
Feedback is extremely important in enabling workplace growth. No level of the organisation should be exempt from either criticism or praise, and the opportunities for learning and development should be available to all.
Feedback requires organisational safety and confidence to enable. Without willing participants, and the ability to speak freely, achieving feedback is a fantasy.
But once this is achieved, what underpins the structure of feedback? What is required for effective communication?
Shared Understanding
Similar to the example sentence above, all communication is predicated on a shared understanding of the symbols and references used. Without a translator, two people who don’t speak the same language will struggle to communicate. With sufficient differences in background, identity, or knowledge specialism, two people can miss a majority of implied information.
In the context of feedback, how do we ensure a baseline of shared understanding?
Feedback Literacy
Much like any specific communications medium, literacy in organisational feedback is essential for both parties.
This latter aspect is often overlooked at an institutional level, where power dynamics prejudice the idea that the paternalistic responsibility of ensuring that feedback is business-effective lies solely in the hands of the person giving feedback. Endless articles will proclaim the best ways of giving feedback, or offer top tips on how to present your ideas so that others will listen.
This misses an entire half of the conversation.
Feedback is not a monologue. For feedback to be effective, it must be understood and internalisable by both parties, and a dialogue must be possible. In order to gauge that information given and information received has been equally understood, language and context must be shared.
To ensure that this is the case, both parties must be literate in their organisation’s form of feedback, both giving and receiving. If you throw someone into a 360° feedback session without ever explaining to them how that differs from old-fashioned 1-to-1 feedback, they can’t be expected to achieve good outcomes.
This is the most extrinsic form of feedback metadata, the handshake that ensures both parties are capable of parsing the information that will be shared.
The Medium Is The Message
The next extrinsic check to ensure good outcomes draws from Marshall McLuhan’s 1964 theory. The medium of feedback itself forms an important context for how the message it contains will be received.
The basics of this concept are usually clear and internalised in most business contexts. Regardless of level within a company structure, employees have some sense of whether a context is better handled in person, or can be easily communicated via email or messaging app.
It’s well understood that the context in which conversations happen can be as important as the conversations themselves, from the presence of outside observers onwards.
Whilst this consideration is obvious when considering the roles of giving and receiving feedback, the role of feedback requests is often overlooked. Modern feedback frameworks are predicated on the ideas of drawing feedback from multiple sources, but how that feedback is gathered can impact internal and external business relationships.
Is a Google survey form better than an emailed request? Do sliding metrics focus response, or prevent articulation of true intent? What is needed from feedback?
Functional Feedback
Whilst the most basic “function” of feedback is enabling iterative improvement, this isn’t specific enough to be useful. Similar to the layout of a user story in product development, or negotiation preparations, each party or role within the feedback process has individual aims, needs, and standpoints.
When requesting feedback, the requester is demonstrating proactive willingness to learn or improve. If the request is not made out of routine obligation, but genuinely desired, is asking for feedback generally the most efficient method?
Research suggests no. A Harvard Business School study concluded that soliciting “advice” rather than “feedback” helped increase the future orientation of the information relayed, ensuring that, similar to the requester’s desire of future improvement, the advisor would match their aims and deliver future-mediated advice.
This is a good example of shared goals being achieved through the contextual (metadata) presentation of a request. The framing of the request itself enables both parties a shared understanding of the functional result.
This outcome could equally be achieved through a context request on behalf of the feedback provider. If someone asks you for feedback, understanding how, and what contexts they will use it can let you better tailor feedback to their purpose, and enable more productive conversations.
A 2012 study by Pelgrim et Al. supports this interpretation, noting that a learner requires the answers to three questions for received feedback to be effective:
- Where am I going?
- How am I going?
- Where do I go next?
This formulation shares the principle of future-oriented behaviour. If the recipient of feedback can’t be sure of their position, direction, and method of progress, no amount of additional information will help them.
This information has to be accurate. It has to be interpretable. It has to share goals.
So how can we structure the most basic context of the information itself to ensure that it will be consistently interpreted and useful?
Feedback at Scale
So far, we’ve identified three role-identities within structured feedback: the requester, the provider, and the recipient. However, up until this point, the assumption has been made that the requester and recipient are likely to be the same individual. In a medium to large modern business, this is frequently not the case.
Feedback can be requested by or for others. It can be requested on behalf of a team or of an entire organisation. It can be given or received through networks of interlocutors or passed through existing hierarchies.
This gives rise to a fourth, generalised, category: the communicator.
The communicator may also encompass any of the previous three roles, or they may form an intermediate point in the communications chain, passing information from source to intended target. It is in this final role that the specific structure of communications network metadata can play the greatest role.
For communications to exist at scale; source, route, and destination must be known. Permissions mediate these relationships and their impact on the enveloped data must be recorded to facilitate transport.
Whilst truly objective feedback is often impossible or of limited use (as few real world behaviours can be neatly reduced to fully scientific metrics, nor do people think in those terms), all efforts should be made to ensure neutrality and detachment in the provision of information, and the separating of more-objective and more-subjective elements to aid the recipient in understanding them.
But how can this be achieved?
Radical Clarity
I’d like to suggest the following 6-point framing for passing on feedback, be it critique or praise:
- Who originally provided the feedback? Note its source, or if it was provided anonymously. Be precise and unadorned, don’t hypothesise.
- How did the source appear to feel when providing the feedback? Give context of tone, behaviour, or anything else that might impact how the message should be received. Note that this metadata is subjective, so be clear on who made the observations.
- Hand over the feedback, in as close to its original form as possible. Give time for the feedback to be understood, and answer questions the recipient may have.
- What purpose do you, as the communicator, feel the feedback serves? Again let the recipient know that your view on this is subjective, unless an explicit purpose was provided alongside the feedback.
- How did the feedback make you, as the communicator, feel? Link this to your relationship with the recipient, and any information you may have about what they were looking for from the feedback. Was the feedback expected? Unexpected? Open a dialogue.
- How does the recipient feel about the feedback? Be open to both emotional and action-based responses. Check if participants’ goals have aligned. Check if the recipient has an idea of how to use the feedback, and what to do next.
This system aids in bringing together the needs of all involved parties, ensures clear signposting of the multi-factor context of the information, and makes the feedback as a whole easier to parse and transform into positive learning outcomes.
Thank you for reading. Hopefully this has provided a few new perspectives to incorporate into your feedback processes.
What have your experiences with business feedback been like? What do you wish you’d known earlier?
Let’s open a conversation in the comments.
