Not every question needs the designated broker. The hard part is knowing which ones do.
Two messages land in a designated broker's inbox on the same Tuesday afternoon.
The first: "How many days is the standard inspection period on the RRPC?"
The second: "My sellers want to stay in the house three weeks after close. Buyer's fine with it. Can we just put it on an addendum?"
They arrive the same way, same Slack channel, same text thread, same "hey, quick question." But they are not the same kind of question. The first has one correct answer that lives in the contract. The second is a risk decision involving possession, insurance, liability, and money moving outside the standard contract terms. It deserves the broker's full attention. The first deserves about four seconds of it.
Most brokerages supervise both the same way. That's the problem.
The two ways flat supervision fails
When every question routes to the DB, one of two things happens.
The DB becomes the bottleneck. Every lookup question consumes the same interrupt as every judgment question. The DB spends the day answering things that are printed on page two of the contract, and the questions that actually need broker judgment wait in line behind them. Response times stretch. Deals stall on answers that should have taken seconds.
Or agents stop asking. This is the worse outcome, and it's quieter. An agent who waited four hours for an answer to a simple question last week doesn't ask the next one. They guess. Most of the time the guess is fine. The time it isn't is the time that shows up in an ADRE complaint, an earnest money dispute, or a demand letter.
Neither outcome is anyone's fault. It's a structural problem: one channel, one queue, one person, every kind of question.
Three tiers
The fix is to stop treating supervision as one thing. In practice, agent questions fall into three distinct tiers, and each tier needs a different handling path.
Tier 1 - Lookup. The answer exists and doesn't depend on the facts of the deal. How long is the inspection period by default. When does the seller have to deliver the SPDS. What's the cure period on a Cure Period Notice. What does the Entire Agreement clause mean. These questions have one right answer, it lives in the form or in the brokerage's established guidance, and the DB's job is to make sure agents can get it instantly, without the DB.
Tier 2 - Judgment. The answer depends on the specific facts, and getting it wrong has consequences that are hard to undo after signatures. Post-close possession. An escalation clause on a financed offer where the appraisal gap matters. Making earnest money non-refundable before the loan contingency clears. A seller credit that interacts with the lender's limits. These are the questions the DB is actually paid to answer, and they deserve a real review, not a hallway "yeah, should be fine."
Tier 3 - Referral. The question isn't a brokerage question at all. It's legal, tax, or title. Whether a seller qualifies for a FIRPTA exemption. Who's entitled to disputed earnest money. How to structure a transaction around a trust or an estate. The DB's job here isn't to answer — it's to recognize the tier, stop the agent from improvising, and document the handoff to an attorney, CPA, or escrow officer.
How agents can tell which tier they're in
The tiers only work if agents can sort their own questions reliably. A few signals catch almost every tier 2 question:
Money moves outside the standard contract terms: credits, holdbacks, rent-backs, non-refundable deposits.
A default timeline changes: shortened inspection periods, extended close dates, waived contingencies.
A party gives up a right: waiving the appraisal contingency, accepting the premises as-is, releasing a claim.
It will be hard to undo after signing: anything that survives close or creates an ongoing obligation.
If any of those are true, it's not a lookup. Send it up.
Setting up each tier
Make tier 1 self-serve. The DB writes the answer once, in their own voice, and agents can retrieve it any time. This is the part most brokerages never get to, and it's the part that frees up the most broker time. Every lookup question the DB answers a second time is time taken from a tier 2 question.
Make tier 2 a real queue, not a text thread. A judgment question should go to the DB with the context attached, the question, the agent's proposed approach, the relevant form language, and it should have a visible status. Pending. Reviewing. Decided. The "reviewing" state matters more than it looks: it tells the agent the DB has it and is working on it, so the agent doesn't move forward on a guess while waiting. And the decision, once made, becomes part of the record.
Make tier 3 a documented handoff. The DB doesn't need to solve the legal question. They need to show that the question was recognized, the agent was told to stop, and the client was referred. One line in the file, timestamped.
Why the record matters
A.R.S. § 32-2151.01 asks designated brokers for written policies and reasonable supervision of licensed activity. "Reasonable" is doing a lot of work in that sentence, and in practice it gets evaluated after something has gone wrong.
A flat, text-thread supervision model leaves almost no record. When ADRE or a plaintiff's attorney asks how the brokerage supervised a specific decision, the honest answer is often "I think we talked about it."
A tiered model produces a record as a side effect. Tier 1 answers are authored and consistent. Tier 2 decisions have a question, a review, and a documented outcome. Tier 3 referrals are logged. That's what reasonable supervision looks like when someone asks to see it.
The DB's time is the scarcest resource in the brokerage. Spending it on lookups isn't diligence, it's a routing problem. Route the lookups to self-serve, route the judgment calls to a real review, route the legal questions out the door, and the DB gets back the hours to do the part of the job only they can do.
