UpdatelyUpdately
Strategy·6 min read

The "My Guess Is" LinkedIn Opener: A Hypothesis-Based Cold Message

Updately Team·2026-10-31

Most cold openers ask the prospect to do work: a call request needs a decision, an open question needs an essay. A hypothesis opener makes one specific guess about their situation, and all the reader has to do is say "yes, actually" or "no, it's fine".

Short answer: send a two-line message: their first name, then one line starting "My guess is" that names the tool they probably use and where it falls short for a team like theirs. No pitch, no question, no link. If they react, answer that exact case in a sentence or two and ask one qualifying question. Updately's "My guess is…" opener template runs this as a campaign, with an AI agent that books a demo only when they ask for one.

Key Takeaways

QuestionAnswer
What is a hypothesis opener?A one-line guess about where their current tool falls short
What does it look like?"{FirstName}!" then "My guess is [tool] isn't a great fit for [their kind of team]."
Does it pitch or ask?No pitch, no question, no link, and your product isn't named
Who is it for?Users of a competitor's tool, and tight ICP lists
What's the goal?A reaction, one qualifying question, then a demo when they ask

Why a guess works better than a question

A question is easy to ignore because answering takes effort. A guess about someone's own setup is harder to leave alone: if it's right, they want to say so, and if it's wrong, they may want to correct it.

The guess also shows you understand their world. "My guess is your CRM was set up for a sales team twice your size" says more than a paragraph about your product. But a bad guess is worse than none: vague ones ("My guess is you have some inefficiencies") read like a template, and insulting ones start an argument. Aim for specific, plausible and fair to the tool.

Who to send it to

The template is built for two lists: competitor users, where the guess can name the tool, and tight ICP lists, where you know the category they must use and the guess names that instead. Signals that feed it:

  • Tech stack: companies using a given technology, fanned out to the right people there. The technology becomes {Intent}, so the opener can name it.
  • G2 or Trustpilot reviewers: people who reviewed a competitor have already said which tool they use.
  • Post engagement on a competitor's posts, or Keyword alerts for people posting about a competitor by name.
  • Profile search or a Sales Navigator list for the ICP version.

Score every lead against an ICP filter. Our guides on finding people using a competitor's tool and reaching competitors' users at scale cover building these lists.

The "My guess is…" sequence, day by day

Delays count from the previous step, messages after the invitation wait until it's accepted, and the sequence stops when they reply.

DayStepWhat it does
0View profileA light touch before the request
1Connection request, no noteSaves the guess for after they accept. Withdrawn after 14 days if not accepted
1 day after they accept"My guess is" openerAI-written in a fixed shape: "{FirstName}!" on its own line, then one line starting "My guess is" that names the tool they likely use (from {Intent} or what was observed about them, otherwise the category) and where it falls short for a team like theirs. No pitch, question or link, and our product isn't named. Under 160 characters
4 days after they acceptSecond guessIf they didn't reply: a different, equally specific guess about their setup, such as cost, a workflow gap or too many tools, in the same shape. One line
11 days after they acceptWhen to check backSays timing may just be off and there's no rush, then asks only "When would be a better time to check back in?"

Writing guesses that land

The AI writes each guess from the problem your product solves and what it knows about the lead, so describe that problem precisely in your business profile. Guesses in the right shape, with no brand names:

  • "My guess is your help desk was built for a support team three times your size."
  • "My guess is your reporting still lives in spreadsheets someone rebuilds every Monday."
  • "My guess is you're paying for three tools that each do a third of the job."

Each names a situation, not a feature. The second guess deliberately changes angle rather than rephrasing the first.

How the AI agent handles replies

The objective is to get a reaction to the guess, qualify with one question, and book a demo when they ask for it. The rules are short and strict:

  • They confirm the guess, or say what kind of team they run: one or two concrete sentences from your business profile about that exact case. No feature list.
  • Then one qualifying question about their setup, such as the tools around it or team size, and nothing else in that message.
  • They disagree: it agrees the tool is solid and asks what they'd change about it. It never argues or criticises the tool.
  • "Later" or "not this quarter": no push for a call. It agrees to check in later, asks when to reach out, and waits.
  • They ask for a demo: it accepts in one line and shares the booking link.
  • A price question: answered from Knowledge only, otherwise handed to you.
  • "Not interested": it thanks them and ends the chat.
  • Always: it matches their length. A one-line reply gets one or two lines back, never more than three short lines.

The standard hand-offs sit underneath: missing facts, or a prospect sharing an email, phone number or calendar link, mean HUMAN INTERVENTION. Once the demo is booked, the agent ends the chat. Note that it never proposes a demo on its own: the prospect pulls it forward.

Follow-ups after a reply goes quiet

If someone reacted to the guess and then stopped, the Follow-ups card sends two messages, with waits of 2 days and 5 days. The first picks up exactly where it stopped with one low-pressure question. The second adds one new, concrete reason to talk and ends with a yes or no question.

How to adapt it

  • Business profile: describe the problem you solve through the situations it shows up in. That's what the guesses are built from.
  • Knowledge: a one or two sentence answer for each kind of team you serve, and pricing if you want the agent to answer it.
  • Who you're reaching out to: name the competitor or category and how you found them ("teams using [tool], from a tech stack signal").
  • Booking link: your demo link, shared only when they ask.

Common mistakes: rewriting the opener into a pitch, adding a question to it (the guess is the question), guessing about the person rather than their setup ("My guess is you're overwhelmed"), and pushing after "later" when the template asks when to check back.

How to launch it in Updately

  1. Build a competitor or ICP list with a signal and an ICP filter, then create a campaign.
  2. On the AI Agent tab, click Start from a template and pick "My guess is…" opener under B2B SaaS & founders. It fills in the sequence, reply rules and follow-ups, and keeps your voice, Knowledge, files and booking link.
  3. Run Test Sequence on the Workflow tab on a few real leads and read the guesses. If they're vague, sharpen the problem description in your business profile.
  4. In the simulator, play someone who agrees, someone who defends their tool, "maybe next quarter", a demo request and a price question.
  5. Add leads and activate.

See it next to the other founder templates on the campaign templates page. For a broad cold list with no known tool, the cold outreach to your ICP template fits better, and our competitor review mining guide helps find reviewers worth guessing about.

FAQ

What is a hypothesis-based cold message?

A message that opens with a specific guess about the prospect's situation instead of a pitch or a question. The conversation starts from their answer.

Why doesn't the opener ask a question?

The guess already invites a reply, and agreeing or disagreeing is less effort than answering. A question would turn it back into a normal cold message.

Is it OK to name a competitor's tool in a cold message?

Yes, if the guess is fair. The template's guesses say where a tool isn't a great fit for a certain kind of team, never that it's bad, and the agent agrees the tool is solid if they push back.

Updately

Find people showing buying intent and message them on LinkedIn.

Signals surface people hiring for your use case, engaging with competitors or asking for recommendations. The AI researches each one and writes the message, then follows up and answers replies.

Watch: Updately AI SDR walkthrough

The cheapest cloud LinkedIn tool with buying signals built in

Updately$39/mo
Gojiberry$99/mo
Expandi$99/mo
Valley$199/mo
  • Sending only from $19/mo per account
  • 10 days free on every plan · cancel anytime

Run it from Claude

Add Updately as a custom connector in Claude (or ChatGPT, Cursor, any MCP client) and work in plain English. MCP setup guide

https://api.updately.ai/mcp

You ask Claude

“Find people at UK fintechs who are hiring SDRs, add them to a list and start my warm-intro campaign.”

“Updately gave us control of what we were doing as opposed to giving the keys and letting them go. But most importantly, the reason why we joined Updately was because it was a third of the price of the agency that we were using to do a similar type of outreach.”
SZSam ZSales Lead, 10X Management
“If you don't have an active, engaged LinkedIn pipeline and you have a clearly defined ICP and persona, you need to add Updately to your tech stack immediately to supplement your sales.”
JRJordan RichardsFounder @ Kaster
“I found myself looking at some of the messages that it's created and thinking like, wow, where did it find this? This went really deep to find information that you could never do in just a cursory search.”
TOTim O'NeilCRO, VP Sales & BizDev