How do you land unpopular changes with your eng teams?

It's a matter of trust. My teams are mostly self-organizing where they have almost complete autonomy over their work product in true Agile sense. The times that I have to step in to introduce changes, I pitch the benefits and intended outcomes. I'm also quite transparent with my teams. If something must be done for security/compliance-sake, I also share with them my disdain for it but also how the particular change increases our sustainability. They may not like the change but if they see the benefits and outcome, you should be ok.

Do you really believe in the change? Not just bought in to someone else's idea, but fully understand it, believe in it, and are able to confidently both defend it and highlight where the benefits to the team are?

Do you have another team you can point to/work with who have already gone through this, or who are going through this at the same time?

Can you find one or two people that you can practise selling the benefits to first - and not hoodwinking them but actually getting them on board warts and all?

If you have the above, then a good day or two out the office at a neutral location can work wonders - get people talking and focus on the clear outcomes you need. You'll need to take people through Why, How and What, but don't let people waylay you in the 'Why' if you can help it - make sure you keep the 'Why' to the benefits, outcomes and opportunities.

The worst thing you an do it let it turn into a moanfest/slanging match - try to plan to find ways to defuse this.

Of course, if you also don't believe in the pivot make sure you question those above you/peers - and see how you can improve the situation before you drop it on your team. If it is that bad, ask yourself why people should stay at the company, yourself included.

In my experience there are 3 types of justifications for big decisions. The first is data supported, any logical person would come to the same conclusion. The second is data sort of supports it, yes there is a leap, but people can see it and argue whether it's real. Finally there is no real data to support, in this case a founder type needs to convince people it's there but no one sees it yet (it's our secret and we're going to own it). What's worked for me is to have a short (like a day) period where you allow everyone to be heard, then you make the decision and make it clear that if you disagree once the decision is made you commit. Inevitably someone who thinks you're wrong will leave, that's ok. New people bought into the direction is good.

Misdirection, bring a live chicken into the office.