Joe Borgione not sure why you need a dictionary but...
z = np.asarray([str(i) for i in np.random.randint(84110, 84120, 100)])
c = np.random.randint(1750, 1780, 100)
dt = [('ZipCode', 'U8'), ('Customers', '<i4')]
keys = np.unique(data['ZipCode'])
kv = {i : data['Customers'][data['ZipCode'] == i].tolist() for i in keys}
{'84110': [1757, 1755, 1766, 1753, 1755, 1774, 1760, 1765, 1772, 1776, 1761, 1762],
'84111': [1753, 1762, 1760, 1769, 1766, 1761, 1766, 1764, 1772, 1760],
'84112': [1761, 1765, 1766, 1760, 1770, 1752, 1763, 1761, 1770, 1763, 1774],
'84113': [1762, 1766, 1773, 1751, 1778, 1762, 1756, 1751],
'84114': [1770, 1759, 1774, 1760, 1757, 1775, 1759, 1763, 1775],
'84115': [1756, 1766, 1753, 1754, 1767, 1758, 1768, 1775, 1769, 1774, 1772, 1777, 1764, 1778, 1750],
'84116': [1759, 1776, 1762, 1778, 1765, 1763, 1763, 1774, 1775, 1770],
'84117': [1762, 1750, 1769, 1756, 1759, 1763, 1779, 1766],
'84118': [1753, 1776, 1752, 1775, 1758, 1770, 1761, 1756],
'84119': [1771, 1768, 1750, 1764, 1768, 1762, 1759, 1761, 1759]}