Outbound
How to find companies that need your service
Short answer
Need is visible before someone starts shopping. It is simply never visible where people look for it, which is at the company level.
Every filter you have describes a container: size, industry, location, revenue band, funding, the tools they bought. Two forty-person agencies can be identical on all of those fields while in one of them an operations lead is losing two days a week to something you fix, and in the other nobody has ever thought about it. Nothing in the company record separates those two, because the thing that differs is somebody's week.
That is the entire mechanism behind a familiar disappointment: a filtered list of lookalike companies feels like progress, and then nothing comes of it.
Why the pages ranking for this question answer a different one
Search this phrase and you mostly get directories, review platforms, chambers of commerce, and exhibitor lists. Those are answers to how do I get found, which is the reverse of what you asked. They dominate the results because the organisations publishing them are in the business of being listed on, so they are the ones with a reason to write four thousand words about it.
Being findable is a real strategy and it is not this one. You asked how to go and locate need that has not yet turned into a search, which means you are looking for evidence that exists before anyone is shopping.
Companies do not have problems
A company is an accounting boundary. Nothing inside it experiences anything.
Problems live in a role. Somebody owns a process, the process is failing, and that person can usually name the month it started getting worse. Everything you would want to know about whether a business needs you is stored in that person's head and, occasionally, in something they typed.
Two consequences follow, and both are inconvenient.
Any list of companies has already discarded the evidence. The filtering happened at the level where need is invisible, so what survives is a set of containers ranked by how much they resemble each other.
The company has no inbox that reaches the person with the problem. A contact form goes to someone whose week is fine. This is why the reply rate on company-level outreach is what it is, and why sending more of it does not help.
The evidence ladder
Three rungs. The whole method is refusing to pretend the bottom one is a reason.
| Rung | What you have | What you can open with | How it fails |
|---|---|---|---|
| 1 · They said it | A dated, first-person statement of the problem | The sentence itself | Rare, and often written by someone selling the fix |
| 2 · They did something about it | A detailed reply, a question asked in public, a group joined, a talk attended | The action, phrased as a question | You are inferring intent from behaviour, so ask rather than assert |
| 3 · They resemble people who had it | Industry, size, tools, funding, job title | Nothing specific to them | It is true of several hundred companies |
Rung three is exactly where filtering by company attributes stops. There is nothing wrong with it as research. The mistake is treating the output as a reason to contact someone, when the only true sentence it supports is that they look like other people.
This is the same ordering as the tiers in how to prioritize sales leads, applied to sourcing rather than to sorting a list you already have.
The problem that will eat your first attempt
The most articulate description of any problem online is usually written by someone who sells the solution to it.
The mechanism is straightforward. If you sell the fix, describing the problem is your marketing, so you write about it every week in polished language. If you have the problem, it is one annoying thing among forty, and you mention it once, in a reply, badly.
So rung one carries a check: before you treat a post as evidence, look at what the author does for a living. A consultant describing the pain of messy client onboarding is doing their job, not asking for help.
The corollary is the useful part. The buyers are in the replies. The person who wrote three sentences about their own situation underneath a vendor's post is on rung one, and very few people read that far down.
A worked example, with the arithmetic
Say you sell fractional operations help to small agencies. The numbers below are invented to show the shape. They are not measured results.
Company first. You filter for agencies of twenty to sixty people in two cities and get 380 of them. Two plausible contacts each gives 760 names. Every single one sits on rung three, which means one message is equally true of all 760. When it fails, you learn nothing, because the claim was never specific enough to be contradicted.
Person first. Same month. You find eight posts where the problem is described in the first person. Say each draws around forty reactions, and the eight of them collect sixty comments between them, so roughly 380 people touched that material. Most of the reactors are peers, vendors, and people who liked the writing. Of the sixty comments, perhaps twenty say something about the commenter's own situation, and after checking what each of those people actually does, perhaps twelve are the person who owns the problem.
Twelve names, out of about 380 people, so a little over three percent.
Twelve looks like failure next to 760. It is roughly a month of outbound for one person working an hour a day, and each of those messages has its opening sentence already written by the recipient.
The ten-minute version
- Write the sentence your buyer would type, not the name of your category. Complaints, in their words: the report ate another Friday, we onboard clients three different ways, nobody knows which number is right.
- Search that phrase and sort by recent. Need is dated. A post from two years ago describes a problem that has since been solved, staffed around, or given up on.
- Read the replies, not the post. The author is often a vendor. The commenters are often not.
- Check what each person does. You want the person who lives with the problem, not the person who sells the fix, and not the person two levels above who hears about it in a summary.
- Copy the sentence into your notes. If you cannot copy a sentence, that name is rung three and should be labelled as such rather than promoted because your list looks thin.
Stop at ten minutes. The stopping rule matters more than the technique, for the reasons in how to research a prospect before a call.
What to do with the bottom rung, which is most of your list
Keep it, and give it the job it is actually good at.
Company filters are excellent for deciding where to look. Knowing the two hundred businesses whose shape fits you tells you which people are worth reading, which posts are worth opening, and which comment threads to check. That is real work and it makes the ten minutes faster.
What it cannot do is tell you who to write to today. A name with no reason is not disqualified, it is unripe. The correct action is to watch it and notice when something changes, not to contact it because the top of the list is small.
Contact the person, not the company
Once you have a rung one name, the message writes itself, and the most common way people ruin it is by generalising the specific thing back into a pitch.
Say the thing they said, accurately, with roughly when they said it. Then say what you would look at first. That is a sample of your judgement, which is the only thing a services buyer can evaluate before they hire you, and the reason it works is covered in lead generation for consultants and agencies.
Do not send it to the company. Send it to them.
Where this does not work
If your buyers do not publish, the top two rungs are empty. Trades, manufacturing, clinics, most of the public sector. No method fills a space where nothing was ever written. Your honest position there is that you are working from resemblance, and the reason to contact someone has to be built in conversation rather than found beforehand. Any page promising otherwise for those markets is selling you something.
If you need hundreds of conversations a month, the ladder caps out. The top two rungs are small by construction, and stretching the definition to fill a quota just relabels rung three. At that volume, a large clean list and a carefully built sending setup is the honest answer, with everything that implies about reply rates.
If a contact costs you nothing, selection is a cost rather than a saving. This page assumes every message spends something real: your time, or the goodwill of a person you can only approach once.
Where Trendle fits
This is what we built Trendle to do, so weigh the recommendation accordingly. It watches the posts your buyers are reading, reads the people who reacted and replied underneath them, and returns a daily shortlist where every name carries the sentence that put it there.
It is the wrong purchase if your buyers are not on LinkedIn, if you need volume rather than reasons, or if you currently have twelve good names you have not written to yet. The method above costs nothing but an hour a week, and software is only worth paying for once doing it by hand is the thing capping your pipeline.
The test before you build any list
Name one person, not one company. Paste the sentence that made you think of them, with a date on it. Then ask whether that sentence would be equally true of somebody else on your list.
If it would, you are on rung three, and the fix is not a better filter. It is reading somewhere that a person has spoken.
Questions people ask next
- How can you tell if a company needs your product before reaching out?
- You cannot tell at the company level, because the fields available to you describe a container rather than a situation. What you can tell is whether a person inside it has said something. The working check is whether you can name the individual and paste the sentence that made you think of them, with a date on it.
- How do I find people talking about a problem online?
- Search the phrasing someone with the problem would use in the first person, not the name of your category. Nobody types operations consulting into a post. They write that the monthly report ate another Friday. Then read the replies rather than the post itself, because the people who have the problem usually show up in someone else's comments rather than writing their own.
- Are funding rounds, headcount growth and tech stack useful indicators of need?
- They tell you about capacity to buy, which is a different thing and often confused with need. A company that just raised money can afford you and may have no reason to want you. Use those fields to rule out companies that could never pay, then go looking for a person who has actually said something.
- What if my buyers do not post anything online?
- Then the top two rungs of the ladder are empty for you and no method fills them. That is common in trades, manufacturing, clinics and most of the public sector. The honest position is that you are working from resemblance, and the reason to contact someone has to be built through conversation rather than found in advance.
- Should I use company filters at all?
- Yes, for where to look and never for who to message. Narrowing to the two hundred companies whose shape fits you is genuinely useful research and it makes the reading faster. It just never supplies a reason, so it cannot be the last step before you write.