Thought Leadership · Outbound Strategy · Diagnostic
Why your outbound isn’t working — the real reasons.
Most outbound fails for the same seven reasons. None of them are the ones sales leaders spend their time fixing. It is not the subject line. It is not the call-to-action. It is not the number of follow-ups. The problems are upstream — and by the time you are tweaking copy, the damage is already done.
Cameron Evans · Founder & CEO, Haku
19 June 2026 · 9 min read
I spent years as a top-performing sales rep before founding Haku. And in that time I watched good salespeople — people who knew their product, believed in it, worked hard — run outbound programmes that produced almost nothing. The standard response was always the same: rewrite the email, change the subject line, try a different cadence, hire better reps.
None of that ever fixed it. Because the problem was never the email.
Outbound fails at the infrastructure level — the targeting, the data, the timing, the deliverability — before the rep ever types a word. By the time a sales leader is staring at a 2% reply rate and wondering what to change, they are usually optimising the wrong variable entirely. The ship has a hole in the hull. They are rearranging the deck chairs.
Here are the seven real reasons your outbound is not working — and what to do about each one.
3.43%
Average cold email reply rate in 2026 — per Instantly's benchmark of billions of emails
10.7%
Reply rate achieved by top-performing teams — same channel, same tools, different inputs
28%
Of the working week the average B2B rep spends actually selling — the rest is research and admin
1
Your ICP is too broad — you are targeting anyone who might buy, not who does buy
The most common outbound mistake is not a bad email. It is a bad list built on a vague ICP. When the ICP is defined as "B2B companies with 50 to 500 employees in the UK" — which is roughly 60,000 companies — the messaging cannot be specific enough to resonate with any of them. The email is generic because the audience is generic. A list of 5,000 loosely matched contacts will always underperform a list of 500 precisely matched ones, because the only way to write a relevant message is to know exactly who you are writing to and why right now.
Fix: Go back through your last 10 closed-won deals. What did those companies have in common beyond size and industry? What was the business context at the point they bought? Build your ICP from evidence, not assumptions. Then disqualify at least 70% of your current list against it.
2
Your data is stale — you are paying for contacts that no longer exist
B2B contact data decays at 25 to 30% annually. That means one in four contacts on a list you bought twelve months ago has moved company, changed role, or had their email address deactivated. You are not generating a 15% bounce rate because of your copy. You are generating it because a significant proportion of your list does not lead anywhere. Every hard bounce also signals to Gmail and Outlook that you do not manage your list responsibly — which compounds into inbox placement problems that affect every subsequent campaign from that domain.
Fix: Stop trusting the verification timestamp on your data provider's database. Data verified six months ago can decay significantly by the time you use it. Validate every contact through live SMTP checks at the point of outreach, not at the point of purchase. Target a bounce rate below 2%. Above 3% means your data problem is actively destroying your deliverability.
3
Your outreach has no timing trigger — you are reaching out at an arbitrary moment
This is the one that separates the 3% reply rate from the 10%. A company that raised a funding round last week is in a buying window. A company that just hired a new VP of Sales is in a buying window. A company that just posted three SDR roles is actively building outbound infrastructure and probably needs the tools to support it. A company that simply matches your ICP on paper — right size, right industry, right title — might buy eventually. Reaching out to the second group with the same urgency as the first is what produces reply rates that make you question whether outbound works at all.
Fix: Every outreach should be triggered by a specific, verifiable signal — not by a weekly sequence cadence. Build a signal monitoring workflow that surfaces accounts when a buying trigger fires. Funding events, leadership hires, headcount growth, tech stack changes. If you cannot name the specific reason this company should hear from you this week, they should not hear from you this week.
4
Your deliverability is broken — your emails are landing in spam
Here is the most uncomfortable truth in outbound: you can have a 95% delivery rate and a 40% deliverability rate simultaneously. Your ESP reports that 95,000 of 100,000 emails were accepted by receiving servers. What it does not tell you is that 27,000 of those went directly to spam. Nobody sees them. Nobody marks them as spam. Your reply rate is low and nobody understands why. Gmail, Outlook, and Yahoo are scoring your sender reputation, domain age, authentication records, and engagement history on every send. If SPF, DKIM, or DMARC are misconfigured, if your domain was never properly warmed, if your bounce rate has been climbing — the emails are going to spam and the copy quality is irrelevant.
Fix: Run an inbox placement test before assuming your copy is the problem. Check SPF, DKIM, and DMARC on every sending domain. Ensure every mailbox has been warmed for at least 21 days before cold sends. Cap sending at 30 to 50 cold emails per mailbox per day. Keep warmup running indefinitely at 20% of daily cold volume. Never send from your primary domain.
5
Your messaging is generic — prospects can identify AI-generated outreach in two seconds
The irony of AI in outbound is that it has made generic personalisation worse, not better. Every rep is now using the same tools to generate the same "I noticed your company recently [insert trigger]" openers, the same three-line value propositions, the same "would love to connect" closes. Prospects in 2026 identify AI-generated outreach instantly — the cadence of the sentences, the structure of the opener, the generic specificity of the personalisation. It reads like it was written by someone who has never spoken to a human. And the reply rate reflects that.
Fix: Specificity is the only thing that reads as human. Reference the actual funding round by name and amount. Reference the specific role they just hired for and why that matters to you. Reference the tech stack change and what it signals about their current priorities. Generic personalisation is worse than no personalisation — it signals that you did the minimum and hoped nobody noticed.
6
Your reps are spending most of their day not selling
Salesforce's 2026 State of Sales report found the average B2B sales rep spends just 28% of their working week in actual selling activities. The remaining 72% goes to prospect research, data entry, internal meetings, email admin, and CRM updates. A rep spending 70% of their time on administrative tasks is not underperforming because they are bad at selling. They are underperforming because they are barely selling. The problem presents as a conversion problem. It is actually a workflow problem — and fixing the workflow has more impact than any rep coaching or messaging workshop.
Fix: Map where your reps' time actually goes — not the idealised version, the real one. Then identify which of those tasks require human judgment and which do not. Research, data validation, list building, and CRM data entry do not require a human. They require a system. Remove them from the rep's workflow. Give the time back to conversations.
7
Your volume is too high relative to your list quality — you are burning your domain
When reply rates are low, the instinct is to send more. More emails, more dials, more sequences. This is exactly backwards. Sending higher volume to a poorly targeted list with stale data does not produce more replies. It produces more bounces, more spam complaints, and a sender domain reputation that takes months to rebuild. A domain that accumulates a 15% bounce rate from a campaign does not recover quickly — and the reputational damage affects every future send from that domain, including warm outreach, customer emails, and transactional messages.
Fix: When outbound is not working, reduce volume before increasing it. Tighten the list, improve the data, fix the deliverability infrastructure, add a signal filter. Send to 200 highly qualified contacts rather than 2,000 loosely matched ones. Prove the motion works at small scale before scaling the problem.
The diagnostic before you change anything
Before rewriting a single email or changing a sequence structure, run through this checklist. The answer to where your outbound is broken is almost always in here — and it is almost always upstream of copy.
Outbound diagnostic — check these before touching the copy
“Every outbound problem that looks like a copy problem is actually a targeting problem. Every targeting problem that looks like an ICP problem is actually a data problem. Start at the data, fix it, and the rest improves automatically.”
Outbound works in 2026. The teams hitting 10%+ reply rates are using the same channels and the same tools as the teams hitting 2%. The difference is not talent or effort. It is the quality of the inputs — the precision of the ICP, the freshness of the data, the specificity of the timing, and the discipline to fix the infrastructure before scaling the volume.
Fix what is broken upstream. The copy will take care of itself.
References
- Instantly. (2026). Cold Email Benchmark Report 2026. Instantly. instantly.ai
- Salesforce. (2026). State of Sales Report, 7th Edition. Salesforce Research. salesforce.com
- Mailgun. (2026). State of Email Deliverability Report. Mailgun. mailgun.com
- Google. (2024, updated 2025). Email Sender Guidelines. Google. support.google.com
- LinkedIn Sales Solutions. (2026). State of Sales Report. LinkedIn. linkedin.com