a narrow interval covering zero tells you that the effect is most likely small (which may be all you need to know, if a small effect is not practically useful), while a wide interval clearly shows that the measurement was not precise enough to draw conclusions.
