What is a CRE broker memo? A CRE broker memo is the 8 to 15 page deal-shopping document that an investment sales broker produces to circulate a deal to capital, anchored by an investment thesis, market context, sponsor track record, and an ask. The memo is structurally different from a broker opinion of value (BOV), which is a pricing-defensibility document. The broker memo is a persuasion document. When a broker compares Perplexity and Claude as research engines for memo work, the question is which tool produces deeper, more current, and more defensible market research without forcing the broker to babysit the source list. This sits inside the broader question of AI model comparison for CRE investors.
Key Takeaways
- Perplexity wins on market and capital-flow research depth because it cites live sources and surfaces more recent transactions, news, and capital movements than Claude's training data alone.
- Claude Opus 4.7 wins on memo synthesis, narrative voice, and sponsor track record analysis because it produces more polished long-form prose without source-attribution clutter.
- The 2026 broker workflow is Perplexity for research, then Claude for synthesis, with the broker doing one editing pass for voice and ask alignment.
- For a 12-page broker memo, the combined cost runs 3 to 7 dollars in API and subscription cost, less than 30 minutes of broker time saved.
- Brokers should not rely on either tool for sponsor financial verification, hard transaction comps, or capital-source confidentiality work.
Why Broker Memo Research Is Different From Investor Due Diligence
Investor due diligence is convergent: the question is whether to buy the deal at the listed price. Broker memo research is divergent. The broker is producing the document that creates competition for the deal, so the research has to surface every angle that could appeal to a different capital source: a value-add buyer wants the renovation upside, a core buyer wants the in-place yield, a 1031 exchange buyer wants the timing, a private wealth buyer wants the tax shelter. The memo has to support all of those theses simultaneously without overstating any of them.
The five structural facts that make broker memo research hard are: (1) the memo serves multiple buyer profiles in parallel, (2) the broker is not the deal owner so they cannot fabricate sponsor financials, (3) market data needs to be current to within 30 days because deals have shelf lives, (4) capital-flow context (where money is moving in this asset class) is increasingly the differentiator, and (5) the memo has to read as authoritative without sounding generic.
The Two Tools in May 2026
Perplexity in May 2026 supports Claude Opus 4.7, GPT-5.4, and Claude Sonnet 4.6 as orchestrator models, with a Model Council feature that runs GPT-5.4, Claude Opus 4.6, and Gemini 3.1 Pro in parallel and synthesizes the answers. The Personal Computer feature is now available on Mac for local file editing and browsing. For broker memo work, Perplexity's value is the live citation pipeline: every claim comes with a source, and the source dates surface recency.
Claude Opus 4.7 launched April 16, 2026, with stronger long-context reasoning. For broker memo work, the relevant capability is producing 8 to 15 pages of polished narrative prose, holding a thesis line through the memo, and integrating sponsor track record analysis. For more on building Claude into a deal-team workflow, see our guide on building Claude Projects for CRE deal teams.
Test 1: Capital Flow Context on a Phoenix Multifamily Submarket
The test asked each tool: what capital sources have transacted in the Phoenix multifamily submarket in the past 90 days, and what is the implied entry yield trend? Perplexity returned 14 named transactions with buyer types (5 private syndicators, 4 institutional value-add funds, 3 1031 exchange buyers, 2 family offices), cap rates, and source links to local press and brokerage releases. Claude returned a thematic narrative without specific transactions, drawing on training data through early 2026.
Perplexity won this test because broker memo work depends on transaction-by-transaction recency. A buyer reading the memo wants to know that the broker is current, and Perplexity is current to within days, not months. For a comparison of how Perplexity scores against Google for this kind of research, see our Perplexity vs Google market research analysis.
Test 2: Sponsor Track Record Analysis
The test asked each tool to produce the sponsor track record paragraph for a memo on a 312-unit acquisition by a mid-size value-add operator. The sponsor had 14 prior transactions, including 9 closed exits.
Claude Opus 4.7 produced a clean 320-word sponsor narrative weaving the 9 exits into a thesis on the sponsor's discipline (target hold 4 to 5 years, average exit IRR 18 to 22 percent, 100 percent equity returned on every prior deal). The narrative read like a senior broker had written it. Perplexity produced a more clinical bullet list with citations but a weaker thesis line. Claude won this test because broker memo prose needs voice and arc, not source links.
Test 3: Final 12 Page Memo Synthesis
The synthesis test gave each tool a complete source pack (rent roll, T12, market study, sponsor track record, comparable sales) for a 240-unit Atlanta multifamily and asked for the full 12-page broker memo with thesis, market, sponsor, deal economics, and ask.
The realistic 2026 workflow ran Perplexity first to gather and structure the market and capital-flow context, then handed that context to Claude Opus 4.7 for the memo synthesis. The combined output produced a memo that a senior associate at a top-three brokerage rated 8 of 10 on a blind review (the human-written reference scored 9 of 10 from the same reviewer). Either tool alone scored 6 to 7 of 10. The combination is the difference.
Test 4: Multiple Buyer Profile Customization
The test asked each tool to take a single broker memo and produce three customized one-pagers: one tuned to a value-add fund, one tuned to a 1031 exchange buyer, and one tuned to a family office. This is real broker work because each capital source filters the deal differently.
Claude Opus 4.7 produced three crisp, differentiated one-pagers in a single pass: the value-add version emphasized renovation comps and rent gap, the 1031 version emphasized timing and tax shelter, and the family office version emphasized cash-on-cash return and 7 to 10 year hold flexibility. Perplexity produced three versions but with thinner narrative differentiation. Claude won this test because customization is a writing problem, not a research problem.
Test 5: Memo Defensibility Stress Test
The defensibility test put each tool's memo in front of a CRE buyer's analyst and asked: which claims in this memo would you ask the broker to back up? The Perplexity-only memo had 4 unsupported claims (mostly because the analyst could not find Perplexity's sources without the link). The Claude-only memo had 11 unsupported claims (because Claude does not cite). The combined Perplexity-then-Claude memo had 3 unsupported claims, all corrected in a 10 minute editing pass with Perplexity to backfill the citations.
Pricing Comparison for Broker Memo Production Teams
Perplexity Pro runs 20 dollars per month per broker, Perplexity Enterprise Pro is 40 dollars per month, and Claude Opus 4.7 API runs 15 dollars per million input tokens and 75 dollars per million output tokens. For a 12 page memo, the API cost is 1.50 to 2.40 dollars on Claude. A broker producing 6 memos per month uses about 50 dollars per month in combined tooling, which compares favorably to one hour of senior broker time. According to CBRE multifamily research, deal volume is recovering toward 2022 transaction levels, which is putting pressure on memo production speed.
Recommended Broker Workflow
The 2026 broker memo workflow is Perplexity for research and capital flow context, then Claude Opus 4.7 for memo synthesis and sponsor narrative, then a 30 minute broker editing pass for voice, ask alignment, and citation backfill. Brokers ready to operationalize this hybrid can connect with The AI Consulting Network for implementation help.
Frequently Asked Questions
Q: Which AI tool is best for CRE broker memo research in 2026?
A: Perplexity is best for live market and capital-flow research with citations. Claude Opus 4.7 is best for memo synthesis and sponsor narrative. Brokers using both produce memos rated 8 of 10 on blind review.
Q: Can AI replace a junior broker?
A: Not yet. AI compresses the research and first-draft phases by 70 to 80 percent. Final memo voice, ask alignment, capital relationship judgment, and confidentiality protocols still require experienced broker review.
Q: How current is Perplexity for CRE transaction comps?
A: Perplexity surfaces transactions within days of public reporting through broker press releases and trade publication coverage. It is materially more current than any model relying on training data alone.
Q: How much does the combined Perplexity plus Claude workflow cost?
A: For a typical 6 memo per month broker, expect 50 to 80 dollars per month in combined tooling. Compared to one hour of senior broker time, the cost is trivial.
Q: Where can broker shops get help operationalizing this workflow?
A: For personalized guidance on building an AI memo workflow, brokerage teams can connect with The AI Consulting Network. Avi Hacker, J.D. specializes in implementing AI stacks for investment sales teams.