Most problems have many possible solutions, so how do you know which one to try first?
Try them all!
Pick every simple, easy-to-try possibility and show how you’ll test each one. To do that, you need to ensure that everyone agrees on how you measure success. Resist the temptation to explain the concept of Lifetime Value at this moment. Instead, simply validate the metric that they are complaining about.
If they say the problem is that leads are flat, then any solution that causes the number of identified leads to go up is a success. Yes, I know that is a flawed metric, but this is not the time to fight that battle. Now is the time to show how experimentation works. (Remember to do adaptive change wrong quickly—you can perfect the metric later.)
If fewer people are opening your e-mails, then whatever gets customers to open more is good. Yes, it could be a delivery problem or it could be a problem with the subject lines or with the offers themselves, but that’s what you’ll test. If you can get more opened, that is a success. Later, after folks see the value of testing, you can ask whether any of these customers actually bought anything from those e-mails. But not now. And don’t argue about whether they are measuring the open rate properly. Just assume that whatever errors are in the old measurements will be in the new ones, too, and move on.
Similarly, accept the fact that dropping online sales really is the problem. Don’t question whether people are buying offline or whether profits are unaffected. That’s later. For now, be happy to get approval to try a number of ideas in rapid-fire sequence.