Software companies get real conversations by targeting on a signal rather than on company size, writing about the reader's problem instead of the product's features, and treating email and LinkedIn as one coordinated sequence. In our own campaigns, switching from a firmographic list to a signal-based one moved positive reply share from 4.6 percent to 41.4 percent on a list one sixth the size.
Why most SaaS cold email fails
The usual diagnosis is that cold email is dead, or that inboxes are too crowded. Neither is right. Campaigns fail for three specific and fixable reasons.
The list is built on firmographics alone. "Series A SaaS companies, 50 to 200 employees, in the US" describes companies that could buy. It says nothing about whether they have the problem right now. That distinction is the single biggest driver of results.
The message is about the product. Most cold emails from software companies open with what the product does. The reader has not yet agreed they have a problem, so a feature list is an answer to a question they never asked.
Nobody reads the replies properly. Plenty of campaigns generate interest that dies because a reply sat unread for four days. The response is the point at which the value is created or lost.
The experiment that proved it
We ran two campaigns for an AI company selling into enterprise technology leadership, deliberately structured to test targeting against volume.
Campaign A targeted CIOs, CTOs and CDOs across the UK on firmographic criteria: 2,349 emails, 30 replies, 7 positive. A 28 percent positive share, which is respectable.
Campaign B targeted a hand-built list of people who had spoken at AI conferences. The thesis was simple: someone who stands on a stage to talk about a problem has already decided the problem matters. Just 668 emails, 29 replies, 12 positive. A 41.4 percent positive share from a list one quarter the size.
The counter-example matters just as much. A separate campaign targeted 4,051 recently funded startups on the assumption that fresh capital means fresh budget. Reply rate was healthy at 2.15 percent, so deliverability and data were both fine. But only four of 87 replies were positive, a 4.6 percent share.
Funding is a weak buying signal on its own. It tells you a company has money. It tells you nothing about whether it has your problem.
Build the ICP from closed-won
If you have twenty or more customers, your target list already exists inside your CRM. The work is finding the pattern.
Look at your best customers, not all of them. Rank by revenue, retention and speed to close, take the top decile, and ask what they had in common before they bought:
- What was happening at the company in the three months before the deal, such as a hire, a funding round, a product launch, a regulation, an outage
- Which tools they already ran, since a tech stack is a statement of priorities
- Who raised the problem internally, and whether that person signed
- What they were doing instead, because the alternative is usually a spreadsheet and a person
That last one is the most useful. If your best customers all replaced a manual process owned by one overworked person, then "who has that overworked person" becomes your targeting criterion, and it is far more precise than headcount.
With fewer than twenty customers, start from two or three hypotheses and test them against each other rather than committing to one.
Signals that beat firmographics
A signal is evidence that the problem is live right now. Ranked roughly by strength:
- Public statements about the problem. Conference talks, podcast appearances, published articles. The strongest signal there is, and the one that produced our 41.4 percent.
- Hiring for the problem. A job posting for a role that exists to do manually what you automate is close to a buying intent declaration.
- Tooling changes. Adopting or dropping an adjacent tool means the category is under active review.
- Community membership. Belonging to a professional community around the topic. This carried an 11,287-person campaign to a 32.9 percent positive share, our largest by volume.
- Regulatory or deadline pressure. A compliance date creates urgency you did not have to manufacture.
- Leadership change. A new CTO or head of data reviews everything in their first ninety days.
- Funding. Weak alone. Useful only when paired with something above.
Who to write to inside the account
Below roughly fifty employees, the founder is the buyer, the user and the budget in one inbox. Write to them, keep it short, and make the offer concrete.
Above that, the person who feels the problem and the person who signs are usually different. A VP of Engineering feels the tooling pain daily. The CTO owns the budget. Writing only to the CTO gets you a polite forward at best. Writing only to the VP gets you enthusiasm with no authority.
The strongest campaigns write to both, with different angles: the practitioner hears about the daily friction, the executive hears about cost, risk or speed. When both reply, the internal conversation happens without you.
For vertical SaaS, do not overlook the operational manager. In one campaign a practice business manager, not a physician or a CEO, sent the clearest buying signal we received:
"Tell me more about your offering with cost."
What the message has to do
A cold email has one job: earn a reply. Not explain the product, not book the meeting, not establish the brand. Earn a reply.
That means:
- Short. Under 120 words. It will be read on a phone between meetings.
- Specific to them in the first line. Not "I saw your website." Something only true of this reader.
- One problem, stated plainly. In their language, not your category's language.
- One clear ask. A question they can answer in a sentence beats a calendar link.
- Written like a person. If a line would embarrass you to send by hand, cut it.
Note what is missing: no case study block, no feature list, no three-paragraph value proposition. Those belong on the call.
And be honest about AI personalisation. Using tooling to research at scale is fine. Using it to auto-generate a paragraph of flattery about someone's homepage is transparent to the reader and actively damages your reputation.
Email and LinkedIn as one sequence
Run separately, these two channels compete. Run together, they compound. A workable cadence:
- Day 0: LinkedIn connection request, no message attached.
- Day 2: First email, referencing the signal that put them on the list.
- Day 5: LinkedIn message if the connection was accepted, different angle from the email.
- Day 8: Second email, new angle rather than a "just bumping this" follow-up.
- Day 15: Final email, short, explicitly closing the loop.
The point is that the name becomes familiar before the ask lands. Five touches from one channel is pestering. Five touches across two channels, each saying something different, reads as persistence.
What to measure
Track four numbers, weekly:
- Delivered rate. Below 95 percent means an infrastructure problem, and nothing else matters until it is fixed.
- Reply rate. 1 to 3 percent is normal. This measures list accuracy.
- Positive share. The percentage of replies that are actually interested. This measures message fit and is the number that predicts revenue.
- Meetings booked. The only one that pays for anything.
Ignore open rates. Since Apple Mail Privacy Protection they are close to meaningless, and any agency reporting them as a headline metric is padding a slide.
If positive share falls while volume rises, you are scaling the wrong thing. That is the moment to go back to the list, not to buy more domains.
Frequently asked questions
Why does SaaS cold email stop working?
Usually because the list is built on firmographics alone. Size and industry tell you a company could buy, not that it has the problem now. Signal-based lists consistently outperform.
What is a good positive reply rate for SaaS cold email?
Above 25 percent of replies indicates strong list and message fit. Our range has been 4.6 percent on a poorly signalled list to 41.4 percent on a tightly targeted one.
Should we email founders or department heads?
Below about fifty employees, the founder. Above that, write to both the person who feels the problem and the person who signs, with different angles.
How long before outbound produces pipeline?
Around two weeks for infrastructure and warm-up, campaigns live in week three, first replies within days. Enough data to optimise by the end of month one. See SaaS lead generation for how we run it.