A pull force figure on a magnet datasheet is measured under ideal laboratory conditions: the magnet sits in direct contact with a thick, flat, clean steel plate and is pulled straight away from it. Very few real assemblies look like that, which is why a magnet that "should" hold a given load sometimes lets go in the field.
The air gap costs more than people expect
Holding force falls away much faster than intuition suggests as the gap opens. A coat of paint, a plastic housing, a label or a rough casting surface is already a gap, and at a few millimetres a small magnet may keep only a small fraction of its contact value. If anything sits between the magnet and the steel, the design should be calculated with that gap, not with the catalogue number.
The steel matters as much as the magnet
The datasheet assumes thick, low-carbon steel. A thin sheet saturates: the magnet's field cannot pass fully through it, so the usable force drops. Austenitic stainless steels are largely non-magnetic, and cast iron or high-carbon steels behave differently from mild steel. A rough, rusty or curved surface also reduces real contact area.
Pulling straight off is not the same as sliding
Most loads do not pull a magnet straight away from the surface; they try to slide it, as with anything hanging on a vertical wall. Sliding (shear) resistance depends on surface friction and is only a fraction of the straight pull value. For hanging loads, design on the shear value and apply a generous safety factor, higher still wherever a falling part could injure someone.
What to tell us when you ask for a magnet
Send us the real conditions: steel type and thickness, the gap, whether the load pulls or slides, and the operating temperature. With those four details we can recommend a size and grade that holds in your actual assembly, and where the margin is tight we will suggest testing samples before you commit to a volume order.