Use the words you expect to see in the documents
Our search matches your wording against the text of the source documents. The more of the right terms you include, the more relevant the evidence we retrieve. Name the metrics, section titles, instruments, and time periods you are after.- Instead of “how is their debt”, write “debt service coverage ratio, direct debt outstanding, debt service schedule, fiscal 2024”.
- Where a term varies by issuer or accounting framework, include the variants, for example “statement of net position (balance sheet)” or “DSCR (debt service coverage)”.
There is no penalty for naming terms that turn out not to appear. Extra relevant terms only help the search find the right pages.
Name the documents you want to focus on
Our search reads more than the page text. Each page also carries its document’s type, name, and publish date, so you can steer retrieval toward the sources you trust by naming them in the prompt.- “Use the official statement and the most recent audited financial statements.”
- “Pull the debt service schedule from the 2023 official statement, not the interim figures.”
- “Prefer audited financials over the budget book where both exist.”
Keep each prompt focused
Shorter, targeted prompts outperform long ones, and the system favors them. When a single prompt blends two unrelated analyses, the search has to satisfy both at once and the evidence for each is diluted, so both answers suffer.- If you want a debt profile and a separate governance review, run them as two prompts.
- If you want five years of one metric, that is one focused analysis and belongs together.
- A good test: if you could hand the two halves of your prompt to two different analysts with no overlap, split it.
Follow-ups stay in the session
In a Deep Dive session, a follow-up question is not searched on its own. Each new question is sent to the search together with everything you have already asked and the answers so far in that session. That is what makes follow-ups work. After a debt service coverage answer you can ask “now break that down by fiscal year” and the search still knows which coverage figure you mean, because the earlier turns travel with it. The trade-off is that an unrelated question asked in the same session drags that earlier context along, which can pull the search toward off-target pages. So treat a session as one line of analysis: keep building on it with follow-ups, but when you move to a genuinely new question, start a fresh session so it gets a clean search.State scope, periods, and definitions
Spell out the boundaries so the analyst does not have to guess.- Scope: what to include or exclude, for example “water and sewer enterprise only, exclude the general fund”.
- Period: which years to cover, for example “fiscal years 2022 to 2024”.
- Definitions: generic metrics differ by sector and accounting framework, so pin them down, for example “DSCR = net revenues available for debt service divided by annual debt service”. See Pin metric definitions for more.
Describe the output you want
Tell the analyst the shape of the answer. Prompts that fix the structure are easier to read and to compare across entities.- Sections and their order, or the exact table columns.
- Formatting rules, for example “round monetary values to the nearest USD thousand”.
- How to handle gaps, for example “use null for missing years rather than dropping the row”.
Putting it together
A focused, well-scoped prompt might read:Operating revenues and expenses for fiscal years 2022 to 2024, water and sewer enterprise only, exclude the general fund. Report debt service coverage as net revenues available for debt service divided by annual debt service. Return a table with columns: fiscal year, operating revenue, operating expense, net position, debt service coverage. Round to the nearest USD thousand and use null for missing years.For authoring, versioning, and sharing these as reusable prompts, see Prompt Templates.