Then choose one channel, collect your first proof by helping a few buyers, and review what you learn after 90 days.
How to market a B2B business from scratch is a different question from how to grow an established one. You have no brand people recognise, no case studies and no audience, so the usual advice about scaling does not fit yet.
This guide gives seven steps in the order we would do them, and a realistic first 90 days. It is written for founders and marketing heads in India, the UK and the Netherlands who must show progress without a large budget.
How do you market a B2B business from scratch?
Start narrow, speak plainly, and learn from real conversations before you spend on scale. These are the seven steps.
- Choose one type of customer, defined by industry, size and role.
- Write their problem in their own words.
- Turn it into one clear offer and one sentence of value proposition.
- Pick one first channel and commit to it for a set period.
- Create your first proof by helping a small number of buyers well.
- Run founder-led conversations and publish what you learn.
- Measure a few things, then decide what to keep after 90 days.
The sections below take each step in turn.
Who exactly are you selling to?
Step one is a short, specific description of one buyer. Without a brand, a narrow target is easier to reach than a broad one.
Write one sentence in this form: "We help [role] at [type of company] who [situation]." Add two checks. Can you name 30 real companies that fit? Can you reach the role on LinkedIn or by email?
If you cannot name them, the target is too vague. Illustrative example: "Heads of logistics at mid-sized food distributors in the Netherlands who are replacing a manual routing process" is workable. "Companies that need efficiency" is not.
Account-based marketing, which focuses effort on a named list of companies, suits this stage well when your target list is short.
What problem do you solve, in their words?
Step two is to find the words buyers use, and use them. Talk to people before you write.
Ask five to ten people in your target group three questions: what is the problem costing you, what have you tried, and what made you start looking? Write down their phrases exactly. Those phrases become your headlines, your search topics and your first outreach messages.
If you have no customers to ask, ask people who have the problem and do not yet know you. Most will talk if you say plainly that you are researching, not selling.
What is the best way to market to B2B businesses at the start?
The best way at the start is direct, specific and helpful. You are buying attention with usefulness, because you have no reputation to borrow.
Step three is the offer. Write one value proposition sentence: who it is for, the problem solved, the result they can check, and why you can do it. Then make the first step small and clear, such as a short assessment or a trial with a defined scope.
Test the sentence on five people from step two. If they ask "how does it work?" you are close. If they ask "what is it?" go back and simplify.
Which first channel should you choose?
Step four is to pick one channel and give it a fair test. Spreading a small effort across five channels gives you no signal from any of them.
The table is our view of how the main options compare when you have no audience. It is a starting point, not a benchmark.
| Channel | Fits when | Watch out for |
|---|---|---|
| LinkedIn, in the founder's own voice | Buyers are professionals who use LinkedIn, and the founder can talk about the problem | Slow to build; needs weekly consistency |
| Direct outreach by email or message | You have a named list of 30 or more target companies | Local rules on email, and the cost of sounding generic |
| Search content | Buyers search for the problem by name | Takes months before it brings enquiries |
| Events and communities | Your buyers gather in a few known places | Time cost; value comes from follow-up |
For outreach, check the rules first.
In the UK, the ICO says PECR does not apply to emailing corporate subscribers, but sole traders are treated as individuals and UK GDPR can still apply. In the EU, GDPR applies. Take local advice for India and for your own case.
How do you get your first proof when you have no case study?
Step five is to earn proof, not to wait for it. Proof can be small and still useful.
- A short pilot with one buyer, agreed in writing, with a defined result to check.
- A written assessment you give a prospect, with their permission to describe it anonymously.
- A quote from a buyer about the process, even before the result.
- Your own working method, shown step by step.
Anonymise carefully and ask permission in writing. Under NDA, you may not be allowed to name the client at all. TBG's published cases are anonymised for the same reason.
We use the same approach with experts. In one published TBG case, a clinical-stage biotech turned monthly interviews with its chief scientific officer into LinkedIn content and technical briefs, as published on our site. Read the full case: /results/biotech-executive-content
How does founder-led marketing work in practice?
Step six is to have the founder or a subject expert talk to buyers and share what they learn. At zero awareness, a named person is more believable than a brand.
In practice, that is four weekly habits: send a few personal messages, post one useful note, reply to buyers' questions in public, and write down every objection you hear. The objections become next month's content.
Keep the notes in one place. They are the cheapest research you will ever do.
How do you create a B2B market when nobody knows the category?
Category creation means defining a new market and positioning your company as its leader. G2 describes category design as setting up a new market where you have a better chance of dominating, rather than competing for share in an existing one.
The same G2 article warns that many companies should not attempt it, and that a niche within an existing category may be a better path. Be honest with yourself about which you are doing.
If buyers already search for something close to what you do, position against that. If they do not, start by naming the problem in their language rather than inventing a label.
This takes budget and patience, and in our view it rarely works as a first move for a small team.
What does a realistic first 90 days look like?
This is an order of work, not a promise of results. Results depend on your market, offer and effort.
| Period | Focus | What you finish |
|---|---|---|
| Days 1 to 14 | Customer and problem | Target sentence, 5 to 10 conversations, phrases list |
| Days 15 to 30 | Offer and first channel | Value proposition tested on five people, one channel chosen, a simple page that explains the offer |
| Days 31 to 60 | Outreach and content | Weekly messages and posts, a growing target list, objections log |
| Days 61 to 90 | Proof and review | One or two pilots started or finished, a review of what produced real conversations |
At day 90, decide what to keep, what to change and what to stop.
How do you know it is working?
Track conversations first and traffic last. At the start, a few real conversations tell you more than a hundred visits.
- Conversations with people in your target group, per week
- Replies and meetings from outreach
- Enquiries, and how many fit your definition of a good lead
- Pilots started
- Where each person says they heard of you, including AI assistants
If after 90 days you have conversations but no pilots, the offer is probably unclear. If you have no conversations, the target or the channel needs another test.
What changes by market: India, the UK and the Netherlands?
Check four things before you copy a plan across borders:
- Who signs, and who else is in the buying group?
- Which language and platform does the buyer use to research suppliers?
- How do buyers expect first contact: email, phone, a form or a messaging app?
- Which data-protection and marketing rules apply to your outreach?
We do not state rules for each country here. Ask a local adviser and test with a few buyers.
