B2B Lead Generation Strategies: What Actually Fills a Pipeline

Last updated July 2026.

The honest problem with lead generation advice is that the strategy that works depends on a variable nobody asks about: how long you can survive without a meeting. A funded team with eighteen months of runway and a founder with four months are correctly given opposite advice, and almost every listicle gives both of them the same twenty bullet points.

So here are the strategies that actually work in US B2B, sorted by time to first meeting, with what each one costs and when it fails.

What are the best B2B lead generation strategies?

Which lead generation strategy should you start with?

Start with referrals, because they cost nothing and convert better than anything else you will ever do, and because they run out fast enough that you need the next channel running before they do. Ask every customer, investor, and former colleague for two specific introductions. Not "let me know if you know anyone." Name the company.

Then start outbound, because it is the only channel that produces meetings on a schedule you control. Cold email is the cheapest version of that: domains and mailboxes cost tens of dollars, the software costs low hundreds, and a well-targeted campaign to 400 companies produces conversations within a month. Cold calling produces them faster and costs several times more per meeting once you price the rep.

And while outbound runs, start the compounding channel. Nobody does this, which is why so many companies are still doing nothing but outbound at $5M in revenue with a cost per meeting that never improved. Publishing content that ranks for what your buyers search takes months to pay, which is the entire argument for beginning before you need it. If nobody on the team writes, this is now the sort of work you can hand to an agent that researches keywords and publishes the articles for you, which is a different economic proposition than hiring a content marketer at $90,000.

How does outbound B2B lead generation work?

Four stages, and teams fail at a different one than they think. Build a list of companies that match a specific profile. Find the right person inside each. Send them something specific enough that they do not feel processed. Follow up until they answer or clearly do not.

The failure is almost never stage three's copy. It is stage one. A campaign aimed at "SaaS companies, 50 to 500 employees" is aimed at nobody, and no amount of clever writing rescues it. A campaign aimed at "companies that posted a Series A in the last 90 days and are hiring their first SDR" writes itself, because you know something true about the recipient.

The mechanics of the rest are solved. Verify every address before you send, because bounce rate is what damages your sending domain rather than your open rate. Send from domains you bought for the purpose rather than the one your invoices go out on. Cap each mailbox at 30 to 50 a day. Stop the sequence the moment someone replies. Our guide to running a cold email campaign covers the sequence, and B2B lead generation software compares the tools by which stage they actually handle.

How many leads do you need to book a meeting?

Do the arithmetic before you commit a quarter to a channel, because it is unforgiving and most teams have never written it down. A well-targeted cold email campaign to a verified list gets a reply rate somewhere between 3% and 8%. Roughly a third to a half of replies are positive rather than a polite no. So 1,000 contacted prospects produce 30 to 80 replies and something like 10 to 30 conversations.

If your close rate from conversation to customer is 15%, that is two to four customers per thousand prospects. Now work backwards from your revenue target. If you need ten new customers this quarter and your contract value is $12,000, you need roughly 3,000 to 5,000 well-targeted prospects contacted, which at 40 emails per mailbox per day means several sending domains and a dozen mailboxes running for the whole quarter.

Two things fall out of that math. First, list quality moves the whole equation and copy only moves one term of it. Second, if the arithmetic says you need 40,000 prospects to hit the number, your contract value is too low for outbound and you should be building the compounding channel instead. Our cold email benchmarks has the current reply-rate distributions.

Is inbound or outbound better for B2B lead generation?

They answer different questions. Outbound answers "how do I get meetings this quarter." Inbound answers "how do I stop paying full price for every meeting forever." A company that only runs outbound has a cost per meeting that is roughly flat for its entire life. A company that only runs inbound has no pipeline for the first two quarters, which for most companies means no company.

The sequencing that works: outbound to survive, inbound to compound, and the outbound conversations tell you what the inbound content should be about. Every objection a prospect raises on a call is a question someone is typing into a search box. Writing the answer costs you nothing you have not already learned, and it is the reason content written by people who do sales outranks content written by people who do content.

What is the cheapest B2B lead generation strategy?

Referrals, at zero, and they do not scale. After that, cold email, because the marginal cost of the thousandth email is nearly zero and the fixed costs are small. The full stack for a team sending a few thousand emails a month is two or three domains at around $12 each per year, six to nine mailboxes at roughly $6 each per month, a sending platform, and verification credits. We put the full arithmetic in what cold email software actually costs.

What makes it expensive is doing it badly. Burned domains, unverified lists, and generic copy all convert the cheapest channel into an expensive one, because the cost per meeting rises even though the cost per email does not.

What B2B lead generation strategies stopped working?

Worth saying plainly. Buying a 50,000-row list and blasting it stopped working when Google and Yahoo began enforcing authentication and complaint thresholds in February 2024, and the practice mainly serves to burn a domain now. Merge-tag personalization stopped working because every prospect receives eight emails a day that use their first name, and the recipient's pattern matcher is better trained than the sender's. Open rate as a success metric stopped meaning anything once Apple Mail Privacy Protection began fetching images on delivery, which means a meaningful share of your "opens" are machines.

What replaced them: smaller lists with a real filter, a message written from something true about the specific company, and reply rate as the only metric that a machine cannot fake. That is the whole reason AI email personalization matters now, and why it means writing from an enriched record rather than filling a token.

Putting it together

Ask for two named introductions from everyone who owes you a favor. Stand up two or three sending domains, warm them, and run a tightly targeted cold email campaign to a few hundred companies you can describe in one specific sentence. Publish the answers to the questions your prospects ask you on calls. Then, once outbound is producing, start the partnership or integration that will still be producing leads in three years.

Do the arithmetic before each channel, not after. The strategy that fills a pipeline is rarely the clever one. It is the one whose math you checked.