Playbook · Chapter 6 · When it fails

When cold email doesn't work.

How to tell a programme that needs fixing from a channel that is genuinely wrong for your business.

The short answer

Cold email usually fails for one of four reasons: the deal size is too small for the economics, the market is too small to cover, the offer is not wanted, or nobody works the replies fast enough. Diagnose in this order: placement, reply handling, segment variation, the ask, and only then the offer.

We have run programmes that sent hundreds of thousands of emails, maintained healthy placement, produced reasonable reply rates, and generated no closed business at all. Publishing only the wins would make this playbook useless, so here is the failure pattern.

The four honest reasons it fails

1. The economics never worked

If your average deal value is under a thousand pounds, the arithmetic almost never closes. Full coverage has a real fixed cost: data, domains, mailboxes, tooling and the time to run it. You need deals large enough that a handful of them pays for the year. This is the most common reason we turn work away, and it is knowable before anyone spends anything. Run your numbers through the ROI calculator before you commit.

2. The market is too small

Under about 10,000 contacts, coverage is not a strategy, it is a list you will exhaust. Below that threshold you are better served by a small number of people doing genuinely researched, individually written outreach, plus referrals and events.

3. The offer is not wanted

This is the painful one, because volume makes it obvious in a way nothing else does. If you have covered a large market with several distinct angles, kept placement healthy, and still cannot generate positive replies, you have run a very large and very expensive market research study. The result is real information, even though it is not the information you paid for.

4. Nobody works the replies

We have watched genuinely strong reply rates convert into nothing because follow-up sat for four days. Cold email interest decays fast. If a positive reply is not worked within a day, most of its value is gone. The same goes for meetings that are booked and then never confirmed: show rates fall sharply without a reminder.

The diagnostic order

When a programme is underperforming, check in this order. Going out of order wastes months.

  1. Placement. Are the emails arriving at all? Nothing else can be assessed until this is yes. See chapter one.
  2. Reply handling. Are positive replies being worked within 24 hours, and are booked meetings confirmed?
  3. Segment variation. Is the weak average hiding one segment that works? See chapter four.
  4. The ask. Have you tested genuinely different asks, or twelve versions of one? See chapter two.
  5. The offer. Only after all of the above. This is the expensive conclusion, so earn it.

Symptoms and what they usually mean

What you seeMost likely cause
No replies at all, not even out-of-officeSpam placement
Auto-replies but almost no human repliesOffer, ask or targeting
Plenty of replies, mostly negativeWrong audience or an ask that irritates
Good positive replies, few meetingsSlow or weak reply handling
Meetings booked, low show rateNo confirmation or reminder process
Meetings happen, nothing closesThe message fits the inbox but not the sales process

That last row deserves a note. If positive replies turn into meetings that never close, the email is promising something the sales conversation cannot deliver. The fix is to make the email and the sales pitch tell the same story, not to send more email.

How long before you call it?

Give a properly built programme around three months. The first few weeks are infrastructure and warmup, the next month or so is testing, and the third month tells you whether the winners hold up at volume. Calling it after three weeks usually means calling it before a single real test has finished. More on timelines in how long cold email takes to book meetings.

A programme that proves the channel does not work for you has still saved you from spending the next two years assuming it would.