Why B2B Lead Generation Is Getting Harder in 2026

Cold open

Michael CarterThe uncomfortable part isn't that lead gen got harder. It's that most teams can't tell you which part got harder.

Sarah MitchellMeaning?

Michael CarterThey see pipeline down twenty percent and they go buy a new tool. Nobody stops to ask whether the list got worse, or the inbox got stricter, or the buyer just stopped picking up. Those are three completely different problems. One tool fixes none of them.

Sarah MitchellAnd you'd argue most teams are solving the wrong one.

Michael CarterI'd argue most teams haven't identified one.

Introduction

Sarah MitchellWelcome to The B2B Data Playbook. I'm Sarah Mitchell. This show is about the unglamorous side of B2B growth. Contact data, prospecting, sales intelligence, the things that decide whether your campaign works before you write a single line of copy. Today's question is one I get asked constantly, usually with some frustration attached. Why is B2B lead generation harder now than it was two or three years ago? My guest is Michael Carter. Michael has spent his career on the revenue side, running sales and demand generation teams. He is practical, he is analytical, and he will tell you when he thinks you are wrong. Michael, welcome.

Michael CarterThanks for having me.

Sarah MitchellLet's start with the premise. Is it actually harder? Or does every generation of sales leaders think their era is the hard one?

Michael CarterBoth. Genuinely both. There's a nostalgia problem in sales where everyone remembers 2019 as easy. It wasn't. But no, I think something structural has changed, and I think it's measurable.

Sarah MitchellOkay. So make the case.

What actually changed

Michael CarterThere are four things happening at once, and the interaction is what makes it feel so brutal. First, the inbox got strict. This is the one I think is most underrated. The major mailbox providers tightened bulk sender requirements a couple of years back. Authentication, one-click unsubscribe, and crucially a hard spam-complaint threshold. That's public, it's documented, and it changed the economics of outbound overnight.

Sarah MitchellBecause volume used to be free.

Michael CarterVolume used to be nearly free. Bad list, low reply rate, who cares, send more. That doesn't work now. Now a bad list actively damages your ability to reach the good contacts on it.

Sarah MitchellThat's a real shift. It used to be that a bad list was wasted effort. Now it's wasted effort plus damage.

Michael CarterRight. The cost moved from your time to your infrastructure. And most teams haven't updated their mental model.

Sarah MitchellWhat's number two?

Michael CarterEveryone got the same tools at the same time. Generative AI made it trivial to produce personalised-sounding outreach at volume. So buyers are receiving a lot more mail that looks researched.

Sarah MitchellAnd it's all noise.

Michael CarterIt's worse than noise. It's noise wearing a suit. When the average buyer learns that "I loved your post about X" means nothing, they stop reading the ones that are genuine too. AI didn't just add volume. It devalued the signals that used to prove effort.

Sarah MitchellLet me push on that. Isn't the answer just, be better at AI than everyone else?

Michael CarterNo, and I think that's the trap. If your advantage is a prompt, your advantage lasts about a quarter. The durable advantage is knowing something true about the account that your competitor doesn't. That's a data problem, not a writing problem.

Sarah MitchellOkay. Three?

Michael CarterBuying committees. This one's well documented. Gartner's research has put the typical B2B buying group somewhere in the range of six to ten stakeholders, and it's been trending that way for years. I don't want to be precise about a number I can't verify for your specific market. But directionally, more hands on the decision.

Sarah MitchellAnd each one can say no.

Michael CarterEach one can say no, and only collectively can they say yes. That asymmetry is the whole story of the modern sales cycle.

Sarah MitchellAnd four?

Michael CarterDiscovery moved. Buyers used to research by searching and clicking. Now a meaningful share of that research resolves inside an answer engine. They ask, they get a synthesised response, they never land on your site.

Sarah MitchellWhich breaks attribution.

Michael CarterWhich breaks attribution, and it breaks the top of the funnel you were counting on to warm people up. So your outbound is now doing work that inbound used to do for free.

Sarah MitchellSo the pressure moves back onto outbound at exactly the moment outbound got harder.

Michael CarterThat's the squeeze. That's the whole episode, really. Four things, all pointing the same direction.

What "verified" should mean

Sarah MitchellI want to dig into the data quality piece, because that's where I see the most self-deception. Teams will say "our data is verified." And I've never once had someone be able to tell me what that word means in their contract.

Michael CarterYeah. "Verified" is the most load-bearing word in this industry and it's completely undefined.

Sarah MitchellSo define it. What should it mean?

Michael CarterIt should mean somebody checked a specific claim, at a specific time, against a specific source. All three. Drop any one and it's marketing.

Sarah MitchellGive me an example of dropping one.

Michael CarterEasy. "Verified email" that means the address is syntactically valid and the domain has a mail server. That's not verification, that's spellcheck. The address can pass that test and belong to someone who left the company eighteen months ago.

Sarah MitchellAnd the date matters because?

Michael CarterBecause contact data decays continuously. Not on a schedule. Somebody is changing jobs right now. A file verified in January is a different product in June, and vendors sell you the January verification with a June invoice.

Sarah MitchellSo the question isn't "is it verified", it's "when, and against what."

Michael CarterAnd "what happens when it's wrong." That third one is where you learn who you're dealing with. Ask a vendor what their replacement policy is on a bounced record. Not whether they have one. What it actually is. How long is the window, does it cover a wrong job title or only a hard bounce, and do you get a credit or a replacement. The specificity of the answer tells you how confident they are.

Sarah MitchellThat's a good tell.

Michael CarterVagueness is information. If somebody can't answer a specific question about their own product, that's an answer.

Why record counts mislead

Sarah MitchellLet me shift to record counts, because you get quite worked up about this.

Michael CarterI do.

Sarah MitchellWhy?

Michael CarterBecause counts are the easiest number to inflate and the hardest to check. Everybody leads with one. Fifty thousand, two hundred thousand, whatever. And nobody says what a record is. Is a record a company or a person? If a company has four contacts, is that one record or four? Does the count include contacts with no email? Does it include contacts you verified once in 2023 and haven't touched since? Same file, four different headline numbers, all technically defensible.

Sarah MitchellSo the number is almost meaningless without the definition.

Michael CarterThe number is meaningless with the definition, honestly. What matters is coverage of your segment. I don't care if you have two hundred thousand records. I care how many operations directors at mid-market manufacturers in the Midwest you have, because that's who I'm selling to.

Sarah MitchellAnd that's a question you can actually test.

Michael CarterThat's a question a sample answers in ten minutes. Which is why I think the sample is the only part of the buying process that matters.

How to run a data sample properly

Sarah MitchellLet's talk about that, because I think most people run a sample badly.

Michael CarterMost people don't run one at all. They eyeball fifty rows, the names look like names, the emails look like emails, and they sign.

Sarah MitchellSo how should you run it?

Michael CarterTwo things. First, request the sample scoped to your hardest segment, not your easiest. If you sell to CFOs at Series B fintechs in Germany, don't accept a sample of US software companies. Anyone can produce that.

Sarah MitchellAnd second?

Michael CarterCheck the fields you'll actually filter on, not the ones you'll read. Everyone checks the email. Almost nobody checks whether employee count is populated on every row, or whether the job titles are normalised, or whether "industry" is granular enough to segment. Those blanks are where campaigns die.

Sarah MitchellThat's the bit I'd underline. The failure isn't usually a wrong email. It's an empty column you assumed was full.

Michael CarterThat's exactly it. And you find out at ten thousand records, not at fifty.

Three patterns that repeat

Sarah MitchellLet's make this concrete. Not named companies, just the patterns you see repeat.

Michael CarterSure. Three patterns I'd say are near-universal. The first is the team that scales volume into a deliverability wall. Pipeline is soft, so they double send volume. Complaint rate ticks up, domain reputation drops, and now the well-targeted emails stop landing too. They diagnose it as a messaging problem and rewrite the sequence.

Sarah MitchellWhen the messaging was never the issue.

Michael CarterThe messaging might have been fine the whole time. They just can't see it, because reply rate is now measuring deliverability, not persuasion. Those two get conflated constantly.

Sarah MitchellWhat's the second?

Michael CarterSingle-threading into a buying committee. Rep finds one enthusiastic champion, works that relationship for three months, deal looks great in forecast. Champion leaves, or gets overruled by someone in finance nobody had ever spoken to.

Sarah MitchellAnd the deal evaporates.

Michael CarterAnd it was never a deal. It was one person's opinion. If the committee is genuinely six or more, one relationship is not coverage, it's a single point of failure.

Sarah MitchellThird?

Michael CarterBuying data for a segment you haven't defined yet. This is the most expensive one. You can't evaluate it, you can't scope it, and you end up buying broad because broad feels safer. Then you're paying for a hundred thousand records to reach the eight thousand you actually wanted, and your deliverability is carrying the other ninety-two.

Sarah MitchellSo the definition work has to come before the purchase.

Michael CarterAlways. And it's free. That's what kills me. The highest-leverage step costs nothing and everyone skips it.

Common mistakes

Sarah MitchellLet's do a quick round. Biggest mistakes, short answers.

Michael CarterTreating a list as a purchase instead of a subscription. You didn't buy an asset, you bought a snapshot of a thing that's decaying.

Sarah MitchellNext.

Michael CarterMeasuring the wrong end. Everyone tracks meetings booked. Almost nobody tracks what percentage of their file was even reachable. If forty percent of your list can't receive mail, your conversion rate is a fiction.

Sarah MitchellNext.

Michael CarterConfusing intent data with intent. A signal that a company is researching your category is not a signal that they'll buy from you. It's a tiebreaker for prioritisation. Teams treat it as a buying trigger and burn credibility reaching out too hot.

Sarah MitchellThat's a good one. Next.

Michael CarterPersonalising the wrong layer. Referencing someone's post is not personalisation, it's proof you can read. Personalisation is knowing something about their operating situation. What they run, what they're migrating off, what just changed.

Sarah MitchellLast one.

Michael CarterNot suppressing. Teams keep emailing contacts who opted out on a different campaign, or who bounced last quarter, because suppression lists live in three systems and nobody merged them. That's an unforced error and it's a compliance problem, not just a performance one.

Sarah MitchellAnd an entirely preventable one.

Michael CarterCompletely preventable. It's a Tuesday afternoon of work.

What to do this month

Sarah MitchellSo if someone's listening and their numbers are down, what do they do this month?

Michael CarterStart by separating the three failure modes I mentioned at the top. Reachability, relevance, and receptiveness.

Sarah MitchellUnpack those.

Michael CarterReachability is, can the message physically arrive. Check your authentication, check your complaint rate, check your bounce rate by source. That's a technical audit and it takes a day. Relevance is, is this the right person at the right company. That's your data and your segmentation. Receptiveness is, do they care right now. That's timing and message.

Sarah MitchellAnd most teams jump straight to the third.

Michael CarterEveryone jumps to the third, because it's the fun one. Rewriting copy feels like progress. But if your problem is reachability, better copy changes literally nothing.

Sarah MitchellOkay. What else?

Michael CarterNarrow before you scale. Pick one segment you can describe in a sentence. Build the list for that. Prove the motion works at small volume. Then scale. The instinct under pressure is to widen, and widening is what breaks deliverability.

Sarah MitchellAnd on the vendor side?

Michael CarterMake the sample do real work. Scope it to your hardest segment, check the filter fields not just the emails, and ask what happens when a record is wrong. If those three answers are good, you're probably dealing with someone serious.

Sarah MitchellAnd if the vendor won't show you records before you pay?

Michael CarterThen you have your answer and you can stop the conversation there.

Takeaways and close

Sarah MitchellLet's land it. Three things.

Michael CarterOne. A bad list isn't neutral any more. It used to waste your time. Now it damages your ability to reach everyone else. That changes the maths on buying broad.

Sarah MitchellTwo.

Michael CarterDiagnose before you buy anything. Reachability, relevance, receptiveness. They look identical in your dashboard and they need completely different fixes.

Sarah MitchellAnd three.

Michael CarterThe sample is the product. Everything else is a claim. Scope it hard, check the fields you'll filter on, and ask what happens when a record is wrong. If that's uncomfortable for the vendor, that's the useful part.

Sarah MitchellI'd add a fourth, if you'll let me. None of this is a tooling problem. Everything you've described is a discipline problem. And discipline is cheaper.

Michael CarterCheaper and less fun. That's why nobody does it.

Sarah MitchellMichael, this was genuinely useful. Thank you.

Michael CarterAnytime. And look, I don't want to be bleak about it. Harder isn't the same as worse. The teams doing this properly are having an easier time than they did five years ago, because the sloppy competition is being filtered out by the same forces.

Sarah MitchellThat's a fair point to end on. The bar went up, and that's an advantage if you can clear it.

Michael CarterThat's exactly it.

Sarah MitchellThat's the episode. If you took one thing from it, make it the diagnostic. Before you buy anything or rewrite anything, work out which of the three you're actually failing at. This show is produced by BizzContacts. If you want to pressure-test your own data against what Michael described, you can request a sample and run it through his checklist. Scope it hard. That's the whole point. Subscribe wherever you're listening, and we'll see you next episode.