Jorge LewisIf you are a founder who has never done outbound, you can get a LinkedIn campaign draft from one sentence about who you sell to. You give the sentence to your own AI. Your AI writes targeting criteria from it, GTM Bud checks each one and saves the draft, and you see real matching prospects before any message is written.
This guide is for the founder of a B2B service or software company who starts outbound for the first time. You have customers, you can say who they are, and you have never built a campaign in an outreach tool. It covers one job: get from that sentence to a draft you trust. It runs through the GTM Bud connector in Claude, ChatGPT or a coding agent.
If you want the wider playbook for doing sales as a founder, read founder-led sales. This page stays with the first sentence.
Why is one sentence enough for a first campaign?
One sentence is enough because the only thing a first campaign needs from you is a clear picture of the buyer. Finding the people and screening them is work the platform does.
A first-time founder usually has the opposite worry: “I do not know the filters.” You do not need to. A good audience sentence answers four questions in plain words.
| Question | Example answer |
|---|---|
| Who is the person? | Heads of customer support |
| Where do they work? | E-commerce brands |
| How big is the company? | 20 to 200 employees |
| Where? | The UK and Ireland |
Put together: “Heads of customer support at e-commerce brands with 20 to 200 employees in the UK and Ireland.”
A rule for a first campaign: write the sentence for the customer you would most like to have again, not for every company that could buy.
How do you write the sentence when you have never done this?
Start from a real customer, not from a market. Think of the last customer who bought without a long argument, and describe that person to a friend who does not know your industry.
Founders who start from a market write weak sentences. Founders who start from a person write strong ones.
| Weak first sentence | What is wrong | Stronger sentence |
|---|---|---|
| “Online shops” | No person, no size | “Heads of customer support at e-commerce brands with 20 to 200 employees in the UK and Ireland” |
| “Companies that need better onboarding” | A need is not visible on a profile | “HR managers at engineering firms with 50 to 250 employees in Germany” |
| “Decision makers in fintech” | “Decision maker” is not a job | “Chief compliance officers at payment companies with 30 to 300 employees in the Netherlands” |
Then ask your AI: “Create a GTM Bud campaign draft for heads of customer support at e-commerce brands with 20 to 200 employees in the UK and Ireland. Show me the criteria check.”
What does the criteria check look for?
The criteria check is a review that GTM Bud runs on the targeting criteria before it saves the campaign draft. Your AI writes concrete targeting criteria from the sentence. GTM Bud checks each criterion for one thing: can the requirement be read from a LinkedIn profile, a company page or public web content? For each criterion that fails, it gives a message and a suggested rewrite. Read the check before you do anything else, because it is the cheapest moment to find a wrong audience. For a first-time founder, this problem is common. You know something true about your buyer, such as “they sell in several countries”, and you assume a tool can see it. Often it cannot. The check tells you so and suggests wording that it can use. The check does not judge how wide the audience is or whether a word has two meanings. Your AI asks you about those points, and the prospect preview shows them. Ask your AI to explain each criterion in plain words. If an explanation does not match what you meant, rewrite the sentence. The goal is a short list of criteria that a person could confirm by reading a profile.
Three results are possible:
- The check passes every criterion. Go to the preview.
- The check flags a requirement it cannot read. Accept the suggested rewrite, or remove the requirement and judge it by eye in the preview.
- The check flags a criterion that repeats built-in sourcing. Remove it. The next section names the two cases.
Your AI can also ask a question of its own, for example whether “e-commerce brands” includes marketplaces. Answer it in the chat, and your AI updates the targeting on the draft.
What should you leave out because GTM Bud already screens for it?
Leave out two filters that new founders often add by hand. GTM Bud sourcing already applies both, so they are not criteria, and the criteria check flags a criterion that repeats either one.
- Recent LinkedIn activity. GTM Bud already screens for people who were active recently. “Active on LinkedIn” in your sentence adds nothing.
- People you already contacted. GTM Bud already skips them. You do not need “exclude my past contacts” as a rule.
Keep the sentence about the buyer: role, company, size, region. The rest is handled.
How do you judge the prospect preview?
Judge the preview by one test: would you be glad if this person answered? The prospect preview shows real people who match the campaign. You mark each one Good fit or Not a fit, and the marks tell you if the sentence works.
- Ask: “Start a prospect preview for this campaign.”
- Read each person. Look at the role first, then the company.
- Mark each one. Tell your AI the reason for each Not a fit.
- If the wrong people pass, make the sentence stricter.
- If the right people are missing, ask which criterion removed them.
- Change the sentence and preview again until you do not need to excuse a result. A preview is one list for each run, and each campaign has a daily preview limit, so change the audience before you ask for a new one.
A message to the wrong person cannot be fixed by better copy. That is why the preview comes before any writing.
A worked example: a founder’s first campaign
The founder in this example runs a six-person company that sells a tool for multilingual customer support. She has eleven customers and has never sent a cold message. The names are invented.
The audience sentence. “Heads of customer support at e-commerce brands with 20 to 200 employees in the UK and Ireland that sell in more than one country.”
What the criteria check said. The check passed role, size and region. It flagged “sell in more than one country” as a requirement that cannot be read from a profile, and it suggested “e-commerce companies with international shipping”. The founder accepted the rewrite and planned to judge the unclear matches by eye in the preview. She had also typed “and who are active on LinkedIn”. The check flagged that as a repeat of built-in sourcing, and her AI removed it, because GTM Bud already screens for recent activity.
Two previewed prospects.
- Aoife B., Head of Customer Support at a skincare brand in Dublin with about 85 employees that ships to twelve countries. Marked Good fit: the role, the size, and customers in many languages.
- Daniel W., Customer Service Manager at a furniture retailer in Leeds with 140 employees that delivers in the UK only. Marked Not a fit: one country and one language, so the tool has nothing to solve.
One value offer. “Your ten most used support replies, translated into two languages of your choice and checked by a native speaker.” GTM Bud gave its instructions, the AI read the company website and wrote the offer, and GTM Bud checked and saved it. The founder chose it because she can deliver it in an afternoon.
One reply draft. A lead wrote: “We use a translation plugin in our helpdesk. Is this different?” The AI wrote, for review:
“A plugin translates each message when it arrives. We work on the saved replies your team sends most, so the answer is already correct before the ticket comes in. The ten translated replies are yours to keep with or without a call. Which two languages cause the most trouble?”
The founder approved the exact text, and her AI sent it.
The results question. “How is my first campaign doing as of now? How many people replied, and what did the Not interested replies say?”
Five prompts for a first-time founder
- “I have never run outbound. Ask me five questions about my best customer, then write one audience sentence from my answers.”
- “Create a campaign draft for heads of customer support at e-commerce brands with 20 to 200 employees in the UK and Ireland that ship abroad. Explain each criterion in plain words.”
- “Preview prospects for this draft. I will say Good fit or Not a fit for each one, and you tell me what my answers have in common.”
- “Set the seller profile of this draft from my website and my LinkedIn profile, then write two value offers that I can deliver alone in one afternoon.”
- “Write the message sequence, then run the launch readiness check and list what is missing in the order I should fix it.”
Three objections first-time founders raise
“I do not know my ICP well enough to write one sentence.” If you have a few customers, you know more than you think. Write the sentence from one of them. If you have no customers and cannot name a buyer, a campaign is too early. The criteria check tells you which parts of a sentence it cannot read. It cannot invent the buyer.
“One sentence is too simple. I will miss good prospects.” You will miss some. A narrow first campaign reaches fewer people, and those people fit. That is the trade. When the first campaign is launched, duplicate it and change the sentence for the next audience. Two clear campaigns teach you more than one wide one.
“I am afraid it will send something before I am ready.” A draft sends nothing. GTM Bud launches a campaign only after you confirm a preview, and launch needs a connected LinkedIn account and an active plan or trial. After launch, your AI shows you each reply for review, and GTM Bud instructs it to send only after you approve the exact text. The honest limit is that the review is your work, up to 40 replies a day for each LinkedIn account.
From draft to launch, in short
After the preview looks right, stay in the same chat. Your AI sets the seller profile, writes the value offers and the message sequence with GTM Bud instructions, runs the launch readiness check, and shows a launch preview with the first message of each channel. Connecting LinkedIn accounts, choosing sending accounts and the sending schedule are done on gtmbud.com.
Setup is one step: add the connector with the instructions on GTM Bud for Claude, GTM Bud for ChatGPT or GTM Bud for coding agents. The full capability list is on the GTM Bud connector page.
Frequently asked questions about a first campaign from one sentence
How long should my audience sentence be?
One sentence, two at most. Name the role, the company type, a size range and a region. If you need a third sentence, you are probably describing two buyers, and each buyer needs its own campaign. For deeper work on the buyer, read how to build an ICP for outbound that converts.
I have never done outbound. Do I need a prospect list first?
No. GTM Bud finds prospects for the campaign from the targeting criteria that your AI writes from your audience sentence, and the connector does not take a list upload. You check the match with a preview of real prospects before any message is written.
What if I sell to two different types of buyer?
Start with the buyer who decides fastest. When that campaign is launched, ask your AI to duplicate it and change the audience sentence for the second buyer. Then run the criteria check and the prospect preview again for the copy.
What if most previewed prospects are Not a fit?
Change the sentence, not the messages. Ask your AI which words in the sentence let the wrong people through, rewrite that part, and start a new preview for the changed audience. A first sentence often needs two or three rounds, and each campaign has a daily preview limit.
Can a first-time founder launch from the chat on the same day?
Often yes, if the LinkedIn account is connected. Your AI builds the campaign in the chat and GTM Bud launches it after you confirm a preview. Connecting your LinkedIn account, choosing the sending account and setting the schedule are done on gtmbud.com, and launch needs an active plan or trial. When replies start, follow the routine in manage LinkedIn replies from Claude or ChatGPT.
Write the sentence, then look at the people
A first LinkedIn campaign draft from one sentence works when the sentence describes one real buyer. Your AI writes the criteria from it. GTM Bud tells you which ones it cannot check, screens for recent activity and past contacts without being asked, and shows real prospects for you to mark Good fit or Not a fit. Write the sentence from your best customer and read the preview before you write a word of copy. If you run a software company with no sales hire, outbound sales for startups without an SDR covers what comes after the first campaign, and LinkedIn outreach automation shows how GTM Bud runs the sending.