I used an RF shielding product which claims 80 dB attenuation at 2.4 GHz. Why am I only seeing 20 dB?

There is only one important 1 key to successful RF shielding: control leakage. Shade

Remember that the attenuation spec for a shielding material is how much radiation penetrates through the shield. Let look at an analogy:

In many ways, RF behaves much like visible light, and RF shielding materials behave much like two sided mirrors. Image that you are outdoors on a sunny day. You set a large mirror on a stand above your head. The attenuation specification for the mirror is very high, perhaps 120 dB or more, so basically no light comes through the mirror. If leakage was not an issue, you would be in total darkness. We all know this is not the case. Leakage from the sides easily illuminates the shaded area. Granted, the amount of illumination is less than standing in full sunlight, but the attenuation is no where near 120 dB. Maybe more like 20 dB. Furthermore, using a shield with even better attenuation will not yield any more benefit. Shade

Now imagine you are in a small room with only one window. Bright sunlight comes in the window and illuminates the room. When you place your mirror shield over the window, you get a dramatic attenuation of the light. But the extent to which leakage of light occurs around the perimeter determines how far from total darkness you will achieve in the room. We have all experienced this when trying to draw a curtain over a bedroom window. You must control the leakage to get it really dark. There is nothing wrong with the shield, light is leaking around the shield.

Because RF shields are reflective on both sides, radiation which does leak in will be reflected by the inner surface of the shield, effectively amplifying the amount of radiation in the room. Even the tiniest leakage at a seam can reduce attenuation by many dB. The obvious solution is to pay serious attention to leakage points. A great how-to book is available which describes materials and methods for controlling leakage in detail.(from lessEMF.com)