Your Outbound Isn't Underperforming. Your ICP Is Too Broad.

Nikhai Jaysen · September 1, 2026

Founders rewrite their cold email copy three times before they question who they are sending it to. The copy is almost never the problem.

A founder rewrites their cold email three times, tries four subject lines, shortens it, adds a case study, removes the case study, and gets the same two percent reply rate. Then they conclude that outbound does not work for their business.

Almost always, the copy was never the problem. The list was. You cannot write a message that is specific to five thousand companies with nothing in common, so you write a message that is specific to none of them, and it performs exactly as well as that description suggests.

The arithmetic that makes a broad list feel safe

Broad targeting is appealing because it looks like more opportunity. Ten thousand contacts feels safer than four hundred. But volume and relevance trade against each other in a way that is easy to miss, because the two numbers move on different timescales.

Send volume is immediate and visible. Reply quality is delayed and hard to attribute. So teams optimise the number they can see. Six weeks later they have a full activity dashboard, a burned domain reputation, and a pipeline of people who took a call out of politeness.

The uncomfortable version: a list of four hundred accounts where you genuinely understand the buying trigger will outproduce ten thousand accounts where you do not, on every metric that ends in revenue.

What narrowing actually means

Most ICP documents narrow on the wrong axes. Industry, headcount, and geography are the easiest fields to filter on, which is precisely why they carry the least information — every competitor is filtering on the same three.

The dimensions that change reply rates are the ones that describe a situation rather than a category.

The trigger

What happened recently that makes this company need you now rather than at any point in the last three years? A funding round, a new hire in a specific role, a product launch, an office opening, a compliance deadline, a competitor switching. If nothing has changed, you are asking someone to create urgency on your behalf, which is the hardest sale there is.

The observable symptom

What can you see from outside the company that indicates the problem you solve is present? A careers page listing four support roles is a symptom. A demo form that takes two days to get a response is a symptom. A pricing page with no self-serve tier is a symptom. If you cannot observe the problem, you are guessing, and the prospect can tell.

The person, not the title

Titles are a rough proxy for who owns a problem, and in smaller companies they are a bad one. The head of operations at a forty-person company owns things that at a four-hundred-person company are split across three teams. Segment by who plausibly owns the outcome, then check the title fits.

The disqualifier

The most useful line in an ICP document is often the one describing who to exclude. Companies below a revenue floor where the economics never work. Industries with procurement cycles longer than your runway. Regions where you cannot support the timezone. Writing these down stops a well-meaning list build from quietly widening the segment every month.

The test

Here is a practical check that takes ten minutes. Write one opening line that is true for every account on your list and false for most companies outside it.

If you can write it, your segment is tight enough to personalise at scale. If the best you can manage is something that would work sent to any B2B company, you do not have a segment — you have a spreadsheet.

This is also the test that tells you how many segments you need. Most businesses we work with end up with three to five, each with its own opening line, rather than one list with a merge field.

What changes downstream

Narrowing the ICP is not only a targeting decision. It quietly fixes four other things.

Research becomes tractable. Finding a genuine, specific detail about four hundred companies is a job. Doing it for ten thousand is not, which is why broad campaigns fall back on merge fields that everyone recognises as automation.

Deliverability improves. Relevant mail gets opened and replied to; irrelevant mail gets deleted and reported. Mailbox providers read those signals. A tight list protects the sending infrastructure you spent weeks warming up, which we cover in the cold email work that happens before you write a single email.

Follow-up gets easier to write. Second and third messages are hard when you know nothing specific about the recipient. When the segment is defined by a shared situation, the follow-up can add a genuinely new angle instead of repeating the first message with more urgency.

Sales stops filtering. When the list is broad, the sales team becomes the qualification layer, spending their week on calls that should never have been booked. That is the most expensive way possible to do targeting.

Where founders get stuck

The common objection is that narrowing feels like leaving money on the table, especially early, when any customer is a good customer. It is a fair instinct and it is worth being honest about the limits: if you genuinely do not yet know who buys, a wider exploratory phase is the right call. Run it as research with a small list and real conversations, not as a volume campaign.

The failure mode is staying in exploration for a year because the dashboard shows activity. Once you have ten or fifteen closed customers, the pattern is usually visible in the data you already hold. Look at which of them closed fastest, renewed, and referred — not at which were largest.

We build the targeting layer, the research, and the sequences as one system for founders doing outbound seriously, across both LinkedIn outreach and AI email outreach. If your outbound is producing activity but not conversations, the list is the first thing worth auditing. Get in touch and we will look at your segments with you.

Frequently Asked Questions

What is an ICP in outbound sales?

An ideal customer profile is a description of the accounts most likely to buy, stay, and refer. A useful one is narrow enough that you could name the trigger that makes a company qualify this quarter rather than last year. A list of industries and headcount bands is a filter, not an ICP.

How narrow should an ICP be for outbound?

Narrow enough that you can write one specific opening line that is true for every account on the list and false for most companies outside it. If your opening line would work equally well sent to any B2B company, the segment is still too broad.

Does a narrow ICP mean fewer meetings?

Usually the opposite. A broader list raises send volume but lowers reply rate, and the replies you do get convert worse because the fit was never there. Narrowing typically reduces sends and increases both reply rate and the share of replies that become qualified conversations.

How do you know when the ICP is the problem and not the copy?

Segment your reply data by account attributes rather than by message variant. If one segment replies at several times the rate of the others on the same copy, targeting is the lever. If reply rates are uniformly low across every segment, then look at the message.