4 ms·

Hm, I could supply a list of utm parameters observed in the wild. Then you could substitute a randomly picked but valid set of parameters...

• virtzzz 8y ago
  The only problem with that is that they'd have no idea the data would be bad. Even if they did, there's nothing stopping the marketers from taking the bad data as proof one or the other campaign was actually working.
  • wumpus 8y ago
    I'm OK with both of those "problems".