Filing in order is what makes fullness expensive
A shelf you append to has no shifting problem at all. You put the new thing at the end and stop. The moment the shelf is in an order — alphabetical, by subject, by classification, by date — a new arrival has to go into its place, and everything after it in the sequence moves along by one.
How far that movement propagates depends on where the empty space is, and that is the entire content of this page. If all the free space sits at the far end of the run, the movement runs all the way there. An insertion at a random point in the sequence therefore moves, on average, half the collection. If the free space is spread across every shelf, the movement stops at the end of the shelf it started on, and it moves half a shelf.
The ratio between those two is exactly the number of shelves in the run, and it is worth stating as an identity because it is unusually clean: half the collection divided by half a shelf is the shelf count. Forty shelves means spreading the slack does forty times less work per insertion. That is not a tuning improvement, it is a different order of magnitude, and it costs nothing but the discipline of not filling shelves from one end.
The arithmetic, and what it assumes
The average of half comes from assuming an insertion is equally likely anywhere in the sequence. Real collections are not uniform — some sections grow much faster than others — and that assumption is generous to the end-slack policy rather than harsh on it, because in practice growth clusters where the run is already tight.
The second assumption is that shifting means sliding volumes along rather than reshelving them properly, which is why the rate field is in volumes per minute rather than per hour. Time yourself on one shelf before trusting the total. And the model counts volumes handled, not distance, so a shift that crosses from one shelf to the next is counted at the same rate as one that stays put, which understates it slightly.
Growth clumps, so the average never fills first
With forty shelves, perfectly even growth would put two and a half percent of the intake on each. That never happens. Sections grow at different rates because subjects do, and any run filed by letter has an unfair distribution baked into the alphabet before a single book arrives. The busiest shelf takes several times its even share.
The consequence is that the first shelf overflows long before the run as a whole is full, and it is why a run comfortably under its fill target on average can still need shifting every few months. The busiest-shelf field is there to make that visible: if the share you enter is four times the even share, that shelf uses its slack four times as fast as the average predicts.
What the trade actually is
Working room is shelving you built and are not using. Shifting is time you spend later. Neither is free and there is no correct answer, only a local one that depends on whether wall space or hours are the scarce thing where you are.
What can be said without hedging is that the distribution of the slack is nearly free and matters more than the amount. Twelve percent free space spread evenly does dramatically less work than twelve percent free space parked at the end of the run, and the second arrangement is what you get by default if nobody decides otherwise, because shelving a collection from the beginning fills every shelf to the brim until the books run out.
To size the shelving in the first place, the book shelving calculator converts volumes into linear feet, bays and uprights at a fill target you set. To divide a case into bays of a workable clear height, the shelf spacing calculator handles the vertical geometry and the pin-hole rounding. If the answer is more shelving in the same room, the mobile shelving calculator compares a moving-aisle layout against a static one on the same floor.
Questions people ask
How full should shelves be?
The calculator deliberately does not name a figure, because the trade is between shelving you build and shifting you do, and only you know which is scarcer. What it can show is the shape: the free-space column tells you how many volumes each shelf can still take before it has to shift, and the headroom column tells you how many years that buys at your growth rate. Somewhere between eighty and ninety percent is where most people with growing collections end up, and a static collection can run much fuller.
Does it matter where the empty space is?
More than how much of it there is. Slack collected at the end of a run does almost nothing, because an insertion anywhere earlier still has to push everything between it and the gap. Slack distributed across every shelf stops each movement at the end of its own shelf. The ratio between the two is exactly the number of shelves in the run, which for a forty-shelf run means forty times less handling per insertion for the same total free space.
Why does the busiest shelf matter more than the average?
Because shifting is triggered by a shelf running out, not by the run running out. Growth clusters: some subjects and some letters grow far faster than others, so the intake never lands evenly. A run that is well under its target on average can have shelves that are completely full and shelves that are half empty, and the full ones are the ones generating work. Entering a realistic share for the busiest shelf shows how much sooner that first overflow arrives.
What counts as a withdrawal?
Anything that permanently leaves the shelves — sold, given away, discarded, moved to a different location. It is worth entering honestly rather than aspirationally, because for most private collections it is a much smaller number than people assume, and the net growth figure is what drives every headroom result on the page. If you have never actually removed anything, entering zero gives you the true picture rather than a comfortable one.
Does this work for boxes or files rather than books?
Yes, with the units renamed. Volumes per foot becomes cartons per foot or filing inches per foot, and the arithmetic is unchanged: an ordered sequence, an insertion, and a movement that runs until it reaches free space. Records series that are filed in order behave exactly like a shelved collection. The one thing that changes is the shifting rate, since a carton is a great deal slower to move than a book, so time it before using the hours figure.