Against Digital Decay: A Case for Letting Data Die

If you spend any time in the circles of digital preservation, you quickly learn the catechism. Data wants to be free. Save everything. More archives are always better. The dominant tone is one of urgency and proactive salvation, a digital Noah’s Ark racing against a flood of planned obsolescence and neglect. The prevailing fear is loss, the irreversible decay of our collective digital memory. But what if this noble impulse, this drive to hoard every byte, is itself a form of decay? What if the real risk isn’t losing data, but being buried by it?

I want to make a counterintuitive, perhaps heretical, argument: that a healthy digital ecosystem requires not just preservation, but also planned, thoughtful deletion. The common advice to archive everything ignores a fundamental truth—context is the most fragile element of any record, and an archive without context is just a landfill. We are so focused on keeping the bits alive that we’ve forgotten that meaning is what matters. A terabyte of saved data from a forgotten corporate intranet, its internal links broken and its purpose opaque, is not a preserved record. It’s a ghost, haunting a server farm, demanding energy and attention while offering nothing but its own cold, incomprehensible presence.

The Preservation Paradox

This is the preservation paradox: the more we save indiscriminately, the more we dilute the value of what we have. Archivists of physical collections have long understood the principle of selection. They cannot, and should not, keep every single draft, every memo, every duplicate. Appraisal—the act of deciding what has enduring value—is the very soul of their work. It is a difficult, subjective, and deeply human process. In the digital realm, however, we’ve outsourced this curation to silicon. We tell ourselves that since storage is cheap, we can defer the hard questions of value to some future expert or an AI that will magically sift the wheat from the chaff.

This is a dangerous fantasy. It assumes that future technologies will not only be able to read our inert data formats but will also possess the wisdom to understand them. More likely, we are creating a problem of monumental scale for our descendants. We are not building a library; we are constructing a midden heap of digital detritus, leaving future generations to scratch their heads at our obsession with saving every transient comment on a defunct social media platform or every version of a now-irrelevant marketing brochure.

The argument for letting some data die is not an argument for carelessness. It is an argument for intentionality. It suggests that the highest form of preservation might be a conscious, documented decision to allow certain information to complete its lifecycle and expire. It forces us to ask difficult questions in the present: Why are we saving this? For whom? What story does this data tell? If we cannot answer these questions satisfactorily, perhaps the most respectful thing we can do for the historical record is to let that data return to the void, clearing the way for the information that truly deserves our limited attention and resources.

Our fixation on preventing all digital decay is, in a way, a denial of a natural cycle. Not everything needs to be permanent. The ephemeral nature of a conversation, the natural obsolescence of a draft, the quiet death of a trivial file—these are not tragedies. They are the necessary processes that give shape and significance to the things we choose, with purpose and clarity, to keep. Perhaps the most radical act of preservation is to occasionally let go.

Notes & further reading

A few pages I came back to while writing this: