B2B prospecting without guesswork: start before the first line

A B2B prospecting message that gets a reply starts before its first sentence. It starts with the decision on which account to approach at all, and with what has specifically happened to make now the right moment. The wording is the last, shortest link in that process. When it is the only link, prospecting turns into guesswork with decorative personalisation. This article describes how to get out of guesswork: how to choose an account, tell a signal from a pretext, build a hypothesis about the problem, gather the minimum context, and close every contact with an explicit status, whether or not a reply comes.

Imagine a salesperson opening a spreadsheet of companies and adding another column: ‘compliment’. In row three it says ‘nice website’. In row seven, ‘like the name’. The list grows, because the feeling grows that if something has been noticed, it must be worth writing. On Monday forty emails go out based on these notes. By Friday two people have replied, one asking to be removed from the list. The uncomfortable conclusion: most of the work was done at the moment of sending, not before.

This is the crux, and it is why the rest of this article runs in order. The message is the shortest link in the chain, not the first. Each earlier link has to be in place: an account worth approaching, a dated signal, a hypothesis about the problem, and enough context to write without guessing.

Signal versus pretext: one distinction that organises the work

Suppose two messages go out on the same morning. The first: ‘I saw you’re opening a new warehouse. Congratulations.’ The second: ‘Account: building materials distribution. Board change 14 days ago. Previously opened a second site.’ Both look similar in the recipient’s inbox. The difference lies elsewhere. The first is written by a sender looking for a reason to write. The second is written by someone who has noted a fact.

A signal has three traits. It is public: anyone can check it in a source. It is dated: you know when it happened, and therefore whether it is still current. And it exists independently of the sender’s intent. A board change, a job advert for a new role, a new site, a tender, an expanded range: these happen even when nobody intends to sell anything. A pretext only appears when someone needs a reason to send a message, which is why it always comes after the wish to make contact. Personalisation without a reason for contact is decoration. It shows that the sender noticed something, but not that anything follows from it.

The line is sometimes blurred, and that is no shame. A job advert for a document specialist might signal growing paperwork, or simply cover for someone on holiday. The aim is not never to be wrong, but to make errors visible. A signal can be checked and challenged. A pretext cannot even start a conversation, because it has no content.

Account before wording: why choosing the target is half the answer

Before anyone starts writing, they make several decisions that matter more than style. Which company to write to, and why us in particular. Which people in it genuinely influence the area the hypothesis concerns. And whether it makes sense to start this quarter or the next. This order is often reversed: first a list of addresses is produced, then reasons are sought. That is where guesswork begins.

A practical filter is to move from account to step through five questions: who are we writing to and why us (account), what happened (signal), what probably follows from it (hypothesis), what minimum information we need so we are not firing in the dark (context), and who is writing, and in what role (sender). Only at the end is the first sentence written, and it is usually short, because the real work has already been done. A well-run message does not need to impress. It only needs to name precisely one thing that connects sender and recipient in that particular week.

Industry data suggests the work is worth doing properly, because the alternative is costly. The Instantly benchmark covering campaigns from 2025 gives an average cold email reply rate of 3.43%, and 58% of all replies come from the first message. In other words, the quality of the first contact, not the volume of sends, drives most reactions. Woodpecker’s 2026 report shows the same mechanism from the other side: campaigns with several follow-ups reach a reply rate of 8.3%, against 4.1% without follow-ups, and advanced personalisation roughly doubles results. These numbers are not promises. They are conditional figures. They suggest that context and consistency take prospecting out of the lottery; they do not guarantee replies.

The model: account, signal, hypothesis, context, message, reply, next step

This sequence is not a textbook method. It describes what salespeople who get replies already do. It looks like this:

1. Account. The company you are writing to, and the reason it is us who are getting in touch. Without this step, every later step is decoration.
2. Signal. A public, dated event that answers the question ‘why now?’. Without a signal there is no reason to make contact, only a pretext.
3. Problem hypothesis. One sentence on what is probably going wrong at this company in the area we would address. Written down explicitly, so it can be challenged.
4. Minimum context. Enough information about the account and the signal that the recipient does not have to guess what it is about, without the sender flooding them with detail. Usually three or four items.
5. Message. Short, with one reason for contact and one proposed next step. The first sentence is written here, not earlier.
6. Reply or exception. Every possible outcome has an explicit response, including the most common one: no reply at all.
7. Next step and status. The matter ends with a named state, not silence: who leads it, what happens next, and when you return to the topic.

The model is deliberately flat, because flat processes are easy to turn into a system. If someone on the team runs twenty threads at once, each of these seven elements can be a column in a spreadsheet or a field in a CRM, provided it is filled in honestly. A dated signal, a one-sentence hypothesis, and a status that tells the truth about the case. This alone does not generate replies, but it removes the most common fault in prospecting: the situation where nobody, on either side, knows what is happening.

Reply, exception and a follow-up that is not just chasing

There are several kinds of reply, and each deserves a different response. Interest leads to a conversation and an agreed next step with a date. A question about detail calls for an answer to that question, without attaching an offer nobody asked for. A refusal closes the matter but not the relationship: note what it concerned, and do not return to the same thread with the same pretext.

The harder cases are ambiguous. ‘Get in touch in six months’ is not a refusal. It is a date to record, along with its source. Passing the matter to another person in the company changes the addressee but not the context: the new recipient inherits the matter rather than starting from scratch. An ambiguous role, such as a contact who may not decide on the area in the hypothesis, is a sign that the account needs clarifying, not a reason to send the same message five times.

Then there is no reply at all, which is the outcome most contacts end in and the one everyone asks about. A follow-up is worth only as much as the new content it adds. A good follow-up brings context the first message did not have: a different signal, a sharpened hypothesis, a different proposed step, or finally an honest close to the thread. Mechanical repetition, such as ‘just bumping this to the top of your inbox’ or ‘did you have a chance to look at our offer?’, adds nothing, and the recipient knows it. After the second or third attempt, silence is an answer, and it is worth closing the matter with an explicit status: ‘no response; we return in six months if a new signal appears’. That is not a failure. It is a record that lets you come back with a reason, instead of starting from zero.

The latest Salesforce report (a 2025 study of 4,050 salespeople) shows the scale of the challenge. The average salesperson spends 40% of their working time selling, and 48% of salespeople say they lack the capacity to make enough contacts, even though they spend close to a full working day a week on prospecting. An earlier 2022 edition put actual selling at 28% of the week. The direction is encouraging, but the numbers still say the same thing: sales work is drowning in administrative tasks. Structured prospecting with an explicit case status does not add hours to the day. It means the same working time goes on contact with the account, rather than on reconstructing from memory who was written to, and why.

Where automation stops, and the legal position

Automation has clearly visible limits in prospecting. Collecting public signals, tracking statuses and reminding yourself to ‘get in touch in six months’ are tasks a system will do more reliably than a person, because it does not forget. But the decision ‘is this a signal or a pretext?’ and the judgement of whether a hypothesis makes sense require common sense, which cannot be written down as a rule. An automated tool that qualifies signals and sends messages on its own is not a prospecting system. It is a machine for generating pretexts. Sooner or later someone will notice, and the reputation of the whole company will pay for it.

The second limit is legal, and it concerns bulk sending. In the UK, electronic direct marketing is governed by the Privacy and Electronic Communications (EC Directive) Regulations 2003 (PECR), alongside UK GDPR where personal data is involved. The Information Commissioner’s Office (ICO) publishes guidance on direct marketing. In broad terms, emails to corporate subscribers (such as limited companies and public bodies) can generally be sent without prior consent, provided the sender is identified and the recipient can easily opt out. Emails to individual subscribers generally need consent, or must fit the narrow ‘soft opt-in’ conditions for existing customers. Sole traders and most partnerships are generally treated as individuals. A one-to-one message about a business matter is treated differently from a mass campaign, but the position depends on the type of recipient and the content, so this article does not settle it. Before planning a campaign, read the ICO guidance, take your own legal advice, and see sitco’s privacy policy.

The third limit is human. The process described here does not guarantee meetings or replies, because too many factors influence those to promise anything. What it does offer is something more modest and more useful: confidence that every contact has a reason, every exception has a handler, and every matter has a status to return to. It is a good starting point for one process to start automating from.

If your ‘to write’ list grows faster than the number of replies, it is worth tidying it up together: sitco.uk/contact.

Sources cited in the text: Instantly Cold Email Benchmark Report 2026, Woodpecker Cold Email Statistics, Salesforce State of Sales 2026 and Salesforce State of Sales 5th edition, and the Privacy and Electronic Communications (EC Directive) Regulations 2003. These are publicly available research and legislation. The vendor figures (Instantly, Woodpecker, Salesforce) describe those companies’ own data, not results from sitco’s services, and are not a promise of outcomes.