Is Cold Outbound Dead for B2B, or Just Harder?

Is cold outbound dead for B2B, or just harder?

Not dead, but the version that worked by volume is finished, and it was killed by infrastructure rather than by taste. Mailbox providers and regulators tightened the rules on bulk sending, and the economics of sending thousands of mediocre emails stopped working. Targeted outbound is fine.

We are not an outbound agency. We build the websites and systems that outbound points at, which gives us a clear view of one thing: what happens after the click. Most outbound debates ignore that half entirely.

So here is the honest version of the argument, with the rules that actually changed.

What actually changed, technically?

The floor moved. Google's sender guidelines, with requirements effective from 1 February 2024, now ask every sender for SPF or DKIM authentication, valid forward and reverse DNS records, TLS connections, RFC 5322 compliant formatting, and a spam rate kept below 0.3 percent in Postmaster Tools.

For anyone sending 5,000 or more messages a day, more applies: both SPF and DKIM, DMARC configured on the sending domain even at a policy of none, alignment between the From header domain and the authenticated domain, and one click unsubscribe on marketing messages, implemented through the List-Unsubscribe-Post and List-Unsubscribe headers alongside a visible link in the body.

Google's documentation also recommends keeping the spam rate below 0.10 percent rather than merely under the 0.3 percent ceiling. That recommended number is the one that ends volume outbound, because sending to a poorly researched list will exceed it quickly.

Why does a spam rate threshold kill volume outbound?

Because it converts a sending decision into a domain level risk. One bad campaign does not just perform badly, it damages the deliverability of every email your company sends afterwards, including your invoices and your product notifications.

The old model treated email as free and reply rate as the only constraint. If 0.5 percent replied, you sent more. That arithmetic assumed the cost of the other 99.5 percent was zero, and it is not zero anymore. It is charged to your domain reputation.

This is why the burner domain industry exists, and why it is a bad idea. Sending from throwaway domains to protect your main one is an admission that you know the sending is harmful, and it does nothing for the brand impression you are leaving on the recipients.

What does the law require?

More than most outbound teams have read. The FTC's own CAN-SPAM compliance guidance is explicit that the law covers all commercial messages and makes no exception for business to business email. There is no B2B carve out, which surprises people every time.

The requirements it lists are concrete. Accurate From, To, Reply-To and routing information. Non deceptive subject lines that reflect the content. Clear and conspicuous disclosure that the message is an advertisement. A valid physical postal address. A clear explanation of how to opt out, honoured within 10 business days, with the opt out mechanism working for at least 30 days after sending.

The FTC states that each separate email in violation is subject to penalties of up to 53,088 dollars, and that you cannot contract away your legal responsibility by outsourcing the sending. That last clause matters for anyone using an agency.

Does any of this mean outbound stopped working?

No. It means outbound stopped being cheap. A researched, relevant message to someone whose situation you actually understand still gets replies, because the volume collapse has made inboxes emptier of that kind of message, not fuller.

What has genuinely degraded is the middle. The email that is personalised enough to look automated and generic enough to be ignorable. It used to survive on volume and now it fails both the deliverability test and the attention test at once.

So the market split into two viable modes: genuinely researched outreach at low volume, and no outbound at all. The middle is where most programmes sit, and that is what people mean when they say outbound is dead.

What does researched actually mean?

That you can name the specific reason this company, this week. Not their industry and headcount. Something that happened: a hire, a launch, a change in their product, a page on their site that says what they are trying to do.

The test is whether the email would be embarrassing to send to anyone else. If you could swap the company name and it still reads fine, it is not researched, it is templated. The recipient applies this test instantly and unconsciously.

This does not scale to thousands, and that is the point. Twenty of these a week from one person is a real motion. Two thousand a week is a different activity that happens to use the same protocol. Our piece on choosing your first GTM channel covers when this is even the right thing to try.

Is AI making this better or worse?

Worse in aggregate, better for individual teams that use it honestly. Generation got cheap, so the volume of mediocre personalised email went up, which raises the bar for everyone including the people using AI well.

Where it genuinely helps is research rather than writing. Using a model to read a company's site, their recent posts and their job listings, and to summarise what they appear to be working on, is real leverage. Using it to write the email from a template is the thing that stopped working.

Keep a person in the loop on the send decision. The failure mode of automated research is confident invention, and an email confidently referencing something the company did not do is worse than no email. Our article on human in the loop AI workflows covers where that checkpoint belongs.

What should the destination page look like?

Like a continuation of the email, not like your homepage. This is the part we see fail most often and it is entirely fixable. Someone reads a specific message about a specific situation, clicks, and lands on a general product page that says nothing about any of it.

The page should restate the situation the email described, then answer the obvious next question. That is usually how it works and what it costs, in that order. Nielsen Norman Group's B2B usability research found price ranked first of 28 information types, 29 percent above product availability in second place, and that finding has aged well.

It does not need to be a bespoke page per prospect. One page per outbound campaign, matching the message, is enough and is achievable in an afternoon. Our piece on landing page design covers making that page carry its weight.

What about outbound that is not email?

The same logic applies with different infrastructure. Calls, physical mail and social messages all reward research and punish volume, they just punish it through human irritation rather than through spam filters. The feedback is slower and the lesson is the same.

Social outreach has the additional problem that it is publicly adjacent. A bad cold email is private. A bad cold message to someone who screenshots it is a marketing event you did not plan.

Calls have quietly become more viable as email has got harder, for the obvious reason that fewer people are calling. That is a channel arbitrage rather than a permanent shift, and it will close.

How should you judge whether it is working?

By qualified conversations per hour of effort, not by reply rate. A low volume programme will have a terrible looking total and an excellent looking ratio, and the ratio is the one that tells you whether to continue.

Track your spam rate in Postmaster Tools as a health metric alongside the commercial ones. If it is drifting toward Google's recommended 0.10 percent, the programme is borrowing from your company's future deliverability to hit this quarter's number.

Also ask what happened to the people who did not reply. Outbound leaves an impression on everyone it touches, and at volume that impression is the largest thing your programme produces.

So what is the honest verdict?

Outbound is a craft channel now rather than a scale channel. If you are willing to do the research, it works and there is less competition for attention than there was. If you were relying on volume, the infrastructure has taken that option away and it is not coming back.

The companies we see doing well with it have one or two people doing it properly, supported by good research tooling and pages that match the message. The ones struggling have a large sequence, a burner domain, and a deliverability problem they have not connected to their pipeline yet.

If you want help making the pages your outbound points at actually convert, or building the research layer behind a smaller and sharper programme, we are happy to walk through it. Find us at phoenix.studio.