What is the observation, implication, question framework for cold outreach?

What are the three parts?

Swipe to see more →

Each part covers a gap the others leave. A fact with no implication reads as trivia. An implication with no fact reads as a guess. Either one with no question gives them nothing to do.

Put your offer after the question, in a single line. If the first three sentences don't hold up, no offer will rescue them. Our guidance: keep O, I and Q to about 45 words together, so the whole email still fits the short format in how long a cold email should be.

What makes a good observation?

Three tests. It has to be verifiable, specific and about them.

• Verifiable. The prospect could check it in under a minute, and you could name where you saw it. A job posting, a pricing page, a press release, a conference agenda, a public post. If you can't point to the source, don't write it.
• Specific. A number, a date, a place or a named product. "You're growing" fits everyone. "You launched an estate-planning practice in August" fits one firm.
• About them. Not about your product, your industry or the economy. The subject of the sentence is their company or their role.

Pick recent facts: a two-year-old announcement shows you skimmed a profile. Stay with what a company or person has published for business purposes, never personal posts. Drop compliments: they add nothing the fact doesn't, so state the fact without the praise. Where to find good facts is covered in sales trigger events, and how much time to research a prospect tells you when to stop looking.

How is the implication different from a feature pitch?

An implication is about their operation. A feature pitch is about your product.

Run the strip test: delete your company name and product category from the sentence. If it still makes sense and still matters to the prospect, it's an implication. If it collapses, it was a pitch.

Swipe to see more →

To draft it, picture the month after this fact became true for a company like theirs: what got harder, and for whom? Hedge on purpose ("usually", "tends to"): you're describing a pattern, not accusing them, and you have no statistic to quote. If nothing comes to mind, stop drafting and go learn the segment. Ask five customers what went wrong in the month after a similar change.

Keep it about teams, not about them. "Your onboarding is probably a mess" is an insult. "Teams at this stage often find onboarding lands on the veterans" is a pattern.

How do you write the question?

One question. About their situation. Answerable in one sentence by someone reading on a phone.

Good questions share three traits. They ask about facts inside their own operation, not your opinion of the market. They're narrow: "Who owns X?" beats "How are you thinking about growth?" And either answer helps you: "Yes, it's a mess" opens a conversation, and "No, it's covered" tells you to stop.

Avoid "Does that resonate?", "Thoughts?" and "Open to a quick chat?" The first two are vague. The third asks for their time before you've earned any. In a first email, the question can be the whole call to action: a reply is the goal. Once they reply, or when the stakes in your implication are clear, follow with a specific time, as laid out in how to write cold emails that actually work.

What does OIQ look like in practice?

Three examples we made up. Every company and person below is fictional, and each situation is invented to show the shape.

Example 1: cold email to a head of operations at a regional trucking company (fictional: Harlow Freight).

Our count: observation 12 words (greeting not counted), implication 21, question 7. That is 12 + 21 + 7 = 40, inside our 45-word guide, and the whole email, subject and offer included, is 66 words.

Example 2: cold call opener to the managing partner of a law firm (fictional: Halberd & Lowe).

On a call the question does the same job: it hands the conversation to them. For more call openers, see cold call openers.

Example 3: cold email to a finance lead at a software company (fictional: Bellwether Analytics).

What do bad openers look like, and how do you rewrite them?

Swipe to see more →

When should you not use it?

Skip it when:

• You can't find a real observation. Don't invent one and don't stretch a weak one. Send a segment-level message that states the problem for the whole group instead, and say plainly that it is a message to a group.
• You can't write the implication. Without a known problem behind the fact, you're guessing.
• The problem is the same for every account and your offer is strong. One shared message to a tight segment can beat a hand-built opener. Use a test to decide: how to A/B test cold emails.
• They already replied or came to you. Inbound leads and warm replies need an answer to what they asked, not a performance of research. For the wording of a time ask, see what words get a prospect to put a time in the calendar; for a brush-off, see how to reply to "not interested" or "send me info"; for a vendor objection, see what to say when a prospect says "we already have a vendor".
• You're writing to hundreds with a merge field. A scraped "observation" pasted into a mail merge is the same irrelevant outreach you were trying to avoid. In a Gartner survey of 632 B2B buyers (fielded August to September 2024, published June 2025), 73% said they actively avoid suppliers who send irrelevant outreach. Relevance takes real research per account.