Most teams blame copy when outbound underperforms.
Subject lines get rotated, frameworks get swapped, and new templates circulate in Slack. But if you're emailing the wrong accounts about the wrong problems, no copywriter on earth can save your calendar.
In this article, I'll break down how to distinguish a copy problem from a targeting problem, how to rebuild a working ICP using your own data, and how to prove to yourself (and your team) that list quality is the primary lever.
The Symptom: Flat Replies, Constant Copy Tweaks
When outbound isn't working, the first instinct is to change the words.
You tweak subject lines, rewrite openers, and run A/B tests on CTAs. Occasionally you see a small bump, but it rarely persists across different lists and timeframes.
That pattern is a tell.
If you can't move the needle meaningfully with copy changes across a large enough send volume, you probably don't have a messaging problem. You have a segmentation problem, or worse, a dirty data problem.
A simple test: compare reply and meeting rates between two groups — accounts that strongly resemble your top 10–20 customers vs a generic segment you've been blasting. If performance is similar across both, your definition of "right account" isn't doing enough work.
Targeting Diagnostics: Ask Your Own Data First
Before you look outward to new tools or vendors, interrogate your own CRM.
Start with your last 20–30 closed-won deals and tag them on four dimensions:
- Industry and sub-vertical (e.g., "B2B SaaS → Sales tech" vs just "Software")
- Employee band (10–50, 51–200, 201–1000, etc.)
- Commercial motion (inbound-heavy, outbound-heavy, PLG, channel-led)
- Key events in the 12 months prior to the deal (funding, key GTM hires, new tools adopted, major launches)
Then run three basic checks:
- Overlap: How many accounts currently in your outbound sequences share at least 3–4 of these traits with your wins? If that number is low, your outbound engine is mostly trained on non-ideal accounts.
- Noise: How many of your currently targeted accounts share 0–1 traits with your wins? If that number is high, you're burning cycles and domains on low-probability bets.
- Conversion delta: Compare conversion rates for high-overlap accounts vs low-overlap accounts (reply → meeting → opp → win). If there's no meaningful step-up, either your ICP story is wrong or your data is too incomplete to be useful.
You want your data to tell a clear story: "When we talk to lookalikes of our best customers, everything gets easier." If it doesn't, fix that before rewriting copy again.
From ICP Narrative to Queryable Filters
Most teams have a narrative ICP: "We sell to B2B SaaS between Seed and Series C." That's a story, not a filter.
You need to translate that into something your tools can enforce. For example:
From there, enforce a rule like:
An account must meet all 3 hard criteria (geo, industry, employee band) and at least 2 of 3 "intent" criteria (tech stack, hiring signal, recent funding).
That's the jump from "we target B2B SaaS" to an enforceable ICP that can actually be used by data and ops.
The List Quality Experiment: Strict ICP vs Legacy List
Instead of debating opinions about whether your list is "good," run an experiment.
- Create two cohorts: List A (Legacy): a random sample from your current "broad" target list. List B (Strict ICP): accounts that meet your fully defined hard rules and intent criteria.
- Hold everything else constant: Same sequence structure, same messaging and subject lines, same sending window and volume pattern, same sender domain(s) and signature format.
- Run both lists until each has at least 200–300 delivered contacts. Then compare bounce rate, reply rate, and meetings booked per 100 contacts.
If List B significantly outperforms List A, you've now got quantitative proof that list quality is a primary lever. At that point, continuing to focus on copy in isolation is just avoidance.
Operationalizing Targeting Inside Your Stack
Once you have a working ICP, bake it into your systems so targeting doesn't quietly decay.
- Enrichment: Append critical attributes (industry, headcount, tech stack, funding, hiring) to accounts as early as possible in your process. Prefer account-level signals over contact-only heuristics.
- Scoring and gating: Assign points for each ICP trait. Disqualify or deprioritize accounts below a score threshold from outbound enrollment. If reps can "just add" anyone into sequences, ICP will drift.
- Governance: Build sequence-level rules: some sequences should be ICP-only. Periodically sample enrolled accounts to check how well they match your declared ICP. Make list reviews a recurring agenda item in pipeline meetings.
If targeting lives only in a slide deck, it will be ignored the moment a rep is chasing a number on the 28th of the month.
Ready to Fix Your Targeting?
Messaging can always be improved. But messaging is a force multiplier, not a miracle worker. If your list is filled with low-fit accounts, tweaks to subject lines and frameworks will never produce the lift you want.
The reliable way to get more out of outbound is to make sure you're talking to the right people about the right problems at the right time.
Work With We Build Pipe