The development of organic light emitting diodes (OLEDs) and the use of emitting molecules have strongly stimulated scientific research of emitting compounds. In particular, for OLEDs it is required to harvest all singlet and triplet excitons that are generated in the emission layer. This can be achieved using the so-called triplet harvesting mechanism. However, the materials to be applied are based on high-cost rare metals and therefore, it has been proposed already more than one decade ago by our group to use the effect of thermally activated delayed fluorescence (TADF) to harvest all generated excitons in the lowest excited singlet state S1 . In this situation, the resulting emission is an S1 →S0 fluorescence, though a delayed one. Hence, this mechanism represents the singlet harvesting mechanism. Using this effect, high-cost and strong SOC-carrying rare metals are not required. This mechanism can very effectively be realized by use of CuI or AgI complexes and even by purely organic molecules. In this investigation, we focus on photoluminescence properties and on crucial requirements for designing CuI and AgI materials that exhibit short TADF decay times at high emission quantum yields. The decay times should be as short as possible to minimize non-radiative quenching and, in particular, chemical reactions that frequently occur in the excited state. Thus, a short TADF decay time can strongly increase the material's long-term stability. Here, we study crucial parameters and analyze their impact on the TADF decay time. For example, the energy separation ΔE(S1 -T1 ) between the lowest excited singlet state S1 and the triplet state T1 should be small. Accordingly, we present detailed photophysical properties of two case-study materials designed to exhibit a large ΔE(S1 -T1 ) value of 1000 cm-1 (120 meV) and, for comparison, a small one of 370 cm-1 (46 meV). From these studies-extended by investigations of many other CuI TADF compounds-we can conclude that just small ΔE(S1 -T1 ) is not a sufficient requirement for short TADF decay times. High allowedness of the transition from the emitting S1 state to the electronic ground state S0 , expressed by the radiative rate kr (S1 →S0 ) or the oscillator strength f(S1 →S0 ), is also very important. However, mostly small ΔE(S1 -T1 ) is related to small kr (S1 →S0 ). This relation results from an experimental investigation of a large number of CuI complexes and basic quantum mechanical considerations. As a consequence, a reduction of τ(TADF) to below a few μs might be problematic. However, new materials can be designed for which this disadvantage is not prevailing. A new TADF compound, Ag(dbp)(P2 -nCB) (with dbp=2,9-di-n-butyl-1,10-phenanthroline and P2 -nCB=bis-(diphenylphosphine)-nido-carborane) seems to represent such an example. Accordingly, this material shows TADF record properties, such as short TADF decay time at high emission quantum yield. These properties are based (i) on geometry optimizations of the AgI complex for a fast radiative S1 →S0 rate and (ii) on restricting the extent of geometry reorganizations after excitation for reducing non-radiative relaxation and emission quenching. Indeed, we could design a TADF material with breakthrough properties showing τ(TADF)=1.4 μs at 100 % emission quantum yield.