Success fee on a repositioning engagement, or just eat the retainer?

Attach the fee to something the client can't dispute and you can measure without an appraiser. Realized rent premium on the first 20 turns is closer to that than appraised value, but it still isn't clean, because the client controls the renovation quality, the leasing agent, and the concession policy. If they turn units cheap and lease at 250 over instead of 380 over, you're arguing about whether that's your recommendation or their execution.

The deeper problem is the one you named. A contingent fee on the reposition path means you get zero if the right answer is sell. That's a real conflict and a sophisticated owner will spot it later even if they proposed the structure. Some consultants handle it by pricing the decision phase flat and only putting a contingent piece on the implementation phase, so the recommendation is bought and paid for before anything is at risk. 8k flat for the decision work then a separate engagement priced on execution is a cleaner shape than one blended fee.

On your 1.5% of appraised lift, run the number before you argue about principle. If the reposition adds 900k of value on their numbers, that's 13.5k, so your total is 21.5k against a 22k quote for the same work, and you've taken all the cap rate risk for nothing.

Watch the securities question too. Depending on how the compensation is structured and what you're being paid to do, performance-linked advisory fees can implicate state licensing or investment adviser rules, which vary by state, so get that reviewed by a lawyer before you sign anything contingent.

Also, a 71% occupied asset with an unreliable comp set usually has a collections or resident-quality problem underneath. Find out before you scope, because if the vacancy is management, the renovation model is answering the wrong question.