Prepare a question before the earnings headline arrives
An earnings announcement is easier to evaluate when you know what you are trying to learn. Start with one business question, one profitability question and one unresolved risk. The aim is to prepare a short reading plan that survives a busy release day, rather than predict the share price or collect every available number.
This tutorial uses NVIDIA’s public ThesisMemo preview as a starting point. Open the NVDA company page, then keep your preparation in a separate document or spreadsheet. The worksheet described here is a manual method: it does not assume that ThesisMemo stores your notes, sends earnings reminders or automatically scores a company against your expectations.
You can complete the preparation before knowing the next results. Leave the future-results column empty. A blank space is useful because it separates what was known before the announcement from what you will learn afterward.
Verify the event date and identify the reporting period
Look for the next earnings information on the company page. Treat a calendar entry as a planning aid and verify the announcement through the company’s investor-relations website. ThesisMemo’s methodology explains that provider dates can change and are not company-confirmed unless explicitly identified that way. If no date is available, retain that uncertainty.
In your worksheet, separate the expected announcement date from the fiscal period the release will cover. Add the announcement time and time zone only when the issuer provides them. A morning announcement in one country may fall on another calendar date for you; preserve the original zone alongside your own reminder.
Do not copy a previous year’s release date into the new calendar. Instead, bookmark the issuer’s events or earnings page and check it again before relying on the schedule. If you create a reminder in your own calendar, label an unconfirmed date as provisional so that the reminder does not turn an estimate into a fact.
Build a small baseline from the previous disclosure
Open the previous relevant company disclosure from the available source links or the issuer website. Record its title, publication date and reporting period. Select only the figures needed for your three questions. For a revenue question, that could mean consolidated revenue and the particular segment you intend to follow, with their units clearly distinguished.
Create columns for metric, scope, period, accounting basis, prior value and source location. Leave another column for the upcoming result. This structure helps you avoid comparing a quarter with year-to-date results or treating a segment as the whole company. Store the page number or table heading beside the link so you can find the same evidence again.
The SEC’s guide to reading a 10-K explains where business descriptions, risk factors, management discussion and financial statements appear in an annual report. Use those sections for background when a short earnings release leaves a term unexplained. This is a document-navigation aid, not a claim that an annual report is the newest quarterly announcement.
Turn the company factors into three answerable questions
Read Current key factors and its review date on ThesisMemo. Choose the parts relevant to your preparation, then inspect their underlying sources. An earlier assessment can suggest a question, but it cannot already contain results that have not been published. Keep the assessment date visible beside your baseline.
Make the business question concrete: what disclosure would help establish whether a previously described development has become a reported business result? For a profitability question, identify the margin or expense relationship you need to inspect. For a risk question, name the specific uncertainty and the evidence that could reduce or increase it.
A useful question has a possible answer and a source location. “Will the company do well?” is too broad. “Does the release explain the change in the relevant segment, and does its reporting scope match the previous quarter?” tells you what to look for. Adapt the wording to the actual company disclosure; do not assume that every company publishes the same operating measures.
Keep guidance and your own assumptions in separate rows
If the previous release contained guidance, copy its period, metric, range and qualifications into a separate guidance row. Do not substitute an analyst estimate or your personal expectation for management’s statement. They may all concern the next quarter, but they come from different authors and answer different questions.
Here is an invented exercise, unrelated to NVIDIA’s actual results. Suppose a fictional company previously guided to revenue of $100–110 million and later reports $108 million on the same basis. The result is inside the range and $3 million above its $105 million midpoint, approximately 2.9%. It is not above the top of the range.
Before using that calculation, confirm that the scope and units match and that guidance was not revised in the meantime. Even a correct comparison does not explain a share-price reaction. Your worksheet should say exactly what was compared, then leave a separate space for business interpretation and any evidence still missing.
Use a release-day reading order that preserves context
When the announcement becomes available, open the issuer’s original document before filling your worksheet. Confirm the company, period and publication date. Read the headline figures, then the definitions and explanation surrounding the measures you selected. If an adjusted measure is prominent, locate its stated basis before comparing it with a reported figure.
Fill the future-results column with the actual disclosure and its source location. Mark each prepared question answered, partly answered or unresolved. Do not force all three into an answered category. A release can provide revenue while leaving a commercial milestone or an operating explanation uncertain.
Next, read management’s explanation and any outlook separately from completed results. Preserve attribution when describing what management expects. If an earnings call, presentation or filing supplies additional evidence later, append it with its own date. Keeping this sequence visible prevents a later explanation from being mistaken for information you had at the first release.
Handle missing or conflicting evidence without improvising
If a desired metric is absent, record not disclosed in the document reviewed. That statement is narrower and more useful than assuming the value is zero or that the company stopped tracking it. Search the relevant supplementary material before deciding whether the gap remains, and name the documents you checked.
When two figures conflict, inspect units, periods, company versus segment scope and whether one number is adjusted. For example, a quarterly value and a cumulative year-to-date value can both be correct while answering different questions. Do not average them or choose whichever supports your expectation.
If you still cannot reconcile the difference, preserve both citations and suspend that comparison. You can complete other parts of the worksheet while leaving this item unresolved. A source-backed partial answer is more reusable than a confident conclusion built on an unexplained mismatch.
Close the event with one change and one next check
After the initial review, write a short event note: what changed relative to your documented baseline, which question remains open and where you will look next. Avoid rewriting the entire company story because one headline looked surprising. Explain the particular evidence that changed your understanding.
Return to the company page to compare its research review date and sources with the new disclosure. A newer filing may appear before the assessment has been refreshed. Read any pending or missing-data notices, and retain your original source record while the product catches up.
Keep your preparation sheet and add the completed results beside it. At the next reporting cycle, review which questions were useful and which were too vague to answer. That feedback improves the next reading plan without turning the exercise into a forecast of returns or an automatic trading rule.
Product references
Questions & answers
Do I need an account to prepare with the NVIDIA preview?
No. The NVIDIA company preview is publicly readable. Keep the preparation worksheet in your own document or spreadsheet. Saving a company to an account-based watchlist is a separate action and is not required for these steps.
Is the earnings date on the page guaranteed?
No. Treat calendar information as provisional unless company confirmation is explicitly provided. Check the issuer’s investor-relations announcement, including its time zone, and revise your own reminder if the schedule changes.
How many metrics should I put in my first worksheet?
Start with the few measures needed to answer your three questions. A smaller set with clear periods, definitions and source locations is easier to review than a long list of numbers with an unclear purpose.
What if the earnings release arrives before ThesisMemo updates?
Read the original issuer disclosure and note its publication date separately from the research review date. Keep the earlier assessment as background; do not assume it has incorporated a newer announcement simply because you reopened the page.
Does an outcome above guidance mean the stock should rise?
This exercise only compares disclosed results with a properly matched baseline. It does not establish market expectations or a price response. Record the business comparison first and keep any explanation of market behavior separate until supported by evidence.