A stock-market video can make a complicated business feel understandable in fifteen minutes. The harder task begins when you close the tab: deciding which parts of the argument are supported, which depend on assumptions, and whether the price already reflects the opportunity.

The most useful question is not simply whether a creator has picked a winner. It is whether you can reconstruct the reasoning, identify what would prove it wrong, and distinguish the creator’s words from the conclusions you have added yourself. A small research record can make that process much more disciplined.

Illustration: Checking the original video alongside dated research notes. AI-generated illustration supplied by They Said Buy.

Start with the statement, not the thumbnail

A bullish headline might introduce a discussion of a company rather than a recommendation to buy its shares immediately. A creator could like a business, consider its stock expensive, and still hold a position bought years earlier. Those statements can all be true at the same time.

Write down what was actually said. Was it a disclosed purchase, a valuation opinion, a price target, or a purchase conditional on a lower price? Record the company, publication date, video timestamp, and any stated time horizon. If the speaker does not supply a deadline or entry price, leave that field blank instead of inventing one.

This is where source-linked records can help. They Said Buy presents exact quotes with links to the relevant second of the original video. The useful next step is still to open that source and listen to the surrounding discussion. A quotation establishes the words; context helps establish their meaning.

Keep the date attached to the idea

A convincing argument from an older video may depend on information that has since changed. Earnings, guidance, debt, share issuance, and the market price can all affect the original proposition. The upload date matters as much as the confidence of the presentation.

Keep two separate questions in your notes: “Was this a reasonable argument using information available then?” and “Does it still describe the company today?” Mixing them makes it easy to reward a prediction for facts that became visible only later, or to dismiss an earlier concern after circumstances improved.

For example, when reviewing Chris Sain’s stock picks, start with a dated statement about one company and follow its source. Then look for subsequent discussion of the same business. Apply that sequence consistently, rather than selecting whichever clip makes the strongest case for an opinion you already hold.

Separate a stock call from a portfolio

A creator’s enthusiasm for a company does not tell you the size of their position, its purchase price, or how it fits alongside their other assets. Mentioning a stock repeatedly is not evidence that it represents a large percentage of a portfolio. Silence is not evidence that a position has been sold.

When someone discloses a holding, record precisely what they disclose and when. A screenshot of one account may exclude other accounts, cash, debt, or hedges. A purchase can be a small addition to an existing position. A sale can be partial. These distinctions matter before trying to copy a trade or describe someone’s overall strategy.

Keep portfolio disclosures in a different column from predictions. That simple separation prevents an attractive price target from turning, in your notes, into an imaginary trade with an imaginary profit.

Make performance comparisons comparable

Once you have a clear statement and date, you can examine what happened afterward. Choose the observation period before looking for the most flattering endpoint. If you compare a stock with a broad market benchmark, use the same start and end dates for both.

A hypothetical stock rising 15% while the chosen benchmark rises 20% over the same period tells a different story from a 15% rise during a falling market. Neither comparison reveals the creator’s personal return. That would require actual transaction prices, position sizes, cash flows, and other information a public video may not provide.

Also check what the numbers measure. Price return and total return answer different questions because dividends affect the latter. Splits can distort an unadjusted comparison. Write down the calculation convention and keep it consistent across the examples you examine.

Illustration: A research trail connects a source, its date, verification, and both positive and negative outcomes. AI-generated illustration supplied by They Said Buy.

Ask what the available record leaves out

Several correct observations about one stock are still observations about one stock. They do not establish an ability to select winners across different industries or market conditions. A short record also cannot support the same conclusions as a broad, consistently assembled sample.

The Jerry Romine stock-pick record offers a useful place to practise this distinction. Examine the dated material that is actually available, then describe the limits of that evidence. A public collection should never be treated automatically as an exhaustive account of everything a creator has said or done.

Look for losing ideas, cautious statements, and unresolved predictions as well as successes. Decide how to handle ambiguous remarks before calculating a success rate. Otherwise, it is tempting to count a vague comment as a successful forecast after a rally and ignore a similar comment when the stock falls.

Check the business behind the story

A video is an introduction to research, not a substitute for it. Take the central claim to the company’s financial reports and investor presentations. If the argument depends on expanding margins, inspect the margin trend. If it depends on a large cash balance, check the debt and cash-flow picture too.

Ask which assumption carries the investment case. Strong sales growth, a new product, and an appealing narrative do not by themselves establish an attractive valuation. Write one sentence explaining what evidence would weaken the thesis, and revisit it when new information arrives.

The SEC’s February 2026 investor alert on social-media stock-tip scams cautions against making investment decisions solely from social media. That principle is useful even when a creator is acting in good faith: an engaging explanation still needs independent verification.

Build a record you can revisit

A practical research note needs only a few fields:

• The original source, publication date, and exact timestamp.
• The claim, its conditions, and any stated time horizon.
• The company information that supports or challenges it.
• Any disclosed holding, kept separate from the prediction.
• A consistent comparison period and a date to review the idea.

The goal is a conclusion you can explain and revise. You may finish the exercise interested in the company, unconvinced by the valuation, or unable to judge with the information available. Each is a useful outcome. The research has done its job when the decision rests on identifiable evidence rather than the momentum of a persuasive video.