Scope creep rarely arrives as a big decision. It arrives as a series of small, reasonable additions, each one easy to say yes to, that together quietly move the date no one agreed to move. Protecting the outcome means saying no, and saying it well.
The wrong way is to say no by reflex or by authority. That just makes you the obstacle, and people start routing around you. The right way is to make the cost visible and let the math do the arguing. "Yes, I can add that, and here's what it pushes and by how much." Most requests withdraw themselves once the trade-off is on the table.
The key is that the no is really a trade, not a wall. The program has a fixed amount of capacity. Every addition spends some of it, which means something else gives. My job is to make that exchange explicit so the person asking can decide whether it's worth it, with full information.
Done consistently, this changes how people approach you. They stop bringing casual additions because they know you'll show the cost, and they start bringing only the ones that are genuinely worth the trade. That's scope control working. It isn't about refusing. It's about pricing.