Reachability Funnel: How Many Accounts You Can Actually Email in 2026

A reachability funnel is the ordered chain that turns a market count into accounts you can actually email: ICP match, role identified, contact record found, address deliverable, nothing suppressed or already contacted. Each stage keeps most of the previous one, and in the worked example below the six multiply out to 5% of the starting count.

Every stage in that chain looks survivable on its own. That is why the compounded result surprises people. A filter that keeps 88% of what it receives reads like a rounding error, and five of them in a row do not.

Going from TAM to reachable accounts is not a haircut, it is a chain of multiplications. The number at the top and the number you can work differ by an order of magnitude. Only one of them supports a quota. Here is the chain, with a named source or a labelled assumption at every stage.

The reachability funnel in six stages

Reachability is the product of six survival rates, not a single coverage number.

1. Market count. Every company in your defined industry and geography, taken from an official establishment count rather than a vendor's database size.
2. ICP match. The share of those companies that clear your size, geography and industry filters.
3. Role identified. The share of ICP accounts where you can name the person who owns the problem you solve.
4. Contact record exists. The share of those named people for whom a work email can be found at all.
5. Address is deliverable. The share of found addresses that verify and are still live on the day you send.
6. Nothing excludes them. The share left after unsubscribes, do-not-contact entries, and accounts a predecessor already emailed.

Multiply all six. The product, not the first number, is your reachable set.

Worked example: 93,903 establishments, 4,720 reachable accounts

Run one real US industry through the chain and about 5% survives.

Compounded survival is 5.03%, about one account in twenty. The harshest single stage still kept 70%.

• Stage 1 is a published count. The Bureau of Labor Statistics reports 93,903 private establishments in NAICS 541613, Marketing Consulting Services, for 2024 (BLS, 2024). BLS defines that as an annual average of quarterly establishment counts, so it counts establishments, not unique corporate parents (BLS QCEW file layout).
• Stage 2 you look up, you do not assume. The Census Bureau publishes County Business Patterns establishment counts by industry, geography and employment size (Census Bureau, 2025), so the 12% here stands in for a number you can count, starting from how to define your ICP.
• Stage 4 carries the only vendor rate in the table. Clay tested 12 email-finding tools and reported 88% raw coverage for Findymail on an enterprise sample of more than 3,700 contacts, and 89% for LeadMagic on an SMB sample of 1,075 contacts (Clay, 2026). Those were the leaders in that test, so 88% is a ceiling, not an expectation.
• Stages 3, 5 and 6 are assumptions, labelled as such. No published figure covers role identification for your buyer, the freshness of your list, or your suppression history. Replace them with counts from your own pipeline.

Why a B2B email coverage rate does not transfer to your market

Published answers to how many decision makers have findable emails were measured on somebody else's sample, usually a well-covered one.

Anymail Finder's 2026 benchmark returned independently verified-valid emails for 86.4% of 5,000 B2B decision-maker contacts in the US, UK, France and Germany, at a 0.9% false-positive rate (Anymail Finder, 2026). The benchmark tells you itself not to generalize it: four Western markets, junior and entry-level seniorities excluded, and an explicit caution that the results may not carry to other regions, industries or seniority levels.

Clay's segment data makes the same point from the other side. Its best SMB tool and its best enterprise tool were different tools, on samples of 1,075 and more than 3,700 contacts. An email match rate from a B2B database describes one segment, not the database.

One number, two stages. Anymail Finder counts a contact as covered only when the address it found passed independent verification, so its 86.4% already contains stage 5. Use a figure like that in place of stages 4 and 5, never on top of them.

Apply an 86% headline to 40-person contractors, non-desk roles, or a market outside those four countries, and you have imported a rate measured on large Western firms into a market nobody measured. Treat a vendor figure as the best case for its best-covered segment, then measure your own, using the methods in how to find decision makers and their emails.

Suppression and prior contact are funnel stages, not admin

The accounts you are not permitted to email are a stage of the funnel, and planning models leave it out.

• The mailbox providers enforce unsubscribes now. Google requires one-click unsubscribe for marketing and subscribed messages from senders above 5,000 messages a day, and tells senders to keep spam rates below 0.10% and avoid 0.30% or higher (Google, 2024).
• Some suppression happens without you. Microsoft documents that invalid recipient addresses can be automatically suppressed from subsequent delivery (Microsoft Learn).
• Stage 5 expires. ZeroBounce reports average bounce rates of 1.33% for marketing and advertising and 1.40% for computers and electronics, calls anything under 2% healthy, and says at least 25% of a database decays every year (ZeroBounce, 2024).

Those ZeroBounce figures were measured on senders' own marketing lists, not cold prospecting lists, so read them as the floor a clean opt-in list reaches, not a forecast for a list you bought last week. Verified is not deliverable covers that gap.

The stage nobody models is prior contact. An account your predecessor emailed twice last quarter passes every data check and is still not reachable. Build the exclusion list beside the account list, as in how to build a B2B lead list from scratch, and stage 6 stops being a surprise.

Invert it: the market size a conversation target needs

Run the chain backwards and it tells you how many target accounts you can actually reach before you commit to a number.

Reachable accounts = market count
                     x ICP% x role% x record% x deliverable% x unsuppressed%

Required market count = target conversations
                        / your measured reply-to-conversation rate
                        / compounded reachability

At the 5.03% from the worked example, a target of 50 conversations needs:

Those rows are arithmetic, not benchmarks. Use your own measured reply rate; a borrowed one plants a fictional number inside a calculation whose purpose is honesty.

At a 2% reply rate, fifty conversations needs a market larger than half of every marketing consultancy in the United States. That is where the plan changes: widen the industry, add a geography, lift the reply rate, or cut the target. Making that call in week zero costs a spreadsheet.

✅ Good: 4,720 reachable accounts in NAICS 541613, from 93,903 establishments through six stated rates, four of them assumptions you can correct.

❌ Bad: A TAM of 93,903 companies with the target set at 10% of it. When the list runs dry in week three, nobody can say which stage lost the accounts.

A smaller honest reachable set beats a larger notional one, because it produces a plan that survives the first send. The compounding is not a defect of any particular tool. It is a property of the market and the chain, whether you buy data or do B2B prospecting by hand.

Why this matters for planning

A bottom-up reachable number is a test of whether you understand your own distribution. A team that says 4,720 reachable accounts and can name the six rates behind it is describing a business; one that says 93,903 is describing an industry. The difference surfaces a quarter later, when the quota built on the larger number turns out to have been fiction. Causo works the qualification end of that chain, narrowing a market to the accounts that actually fit before anyone pays to enrich them.

FAQ

What percentage of B2B contacts have findable emails? Published figures cluster in the mid-80s on decision-maker samples in well-covered markets. Anymail Finder's 2026 benchmark returned independently verified-valid emails for 86.4% of 5,000 B2B decision-makers in the US, UK, France and Germany, and Clay's test of 12 tools put its best performers at 89% on 1,075 SMB contacts and 88% on 3,700 enterprise contacts. Both samples were Western and mostly desk-based, so treat them as a ceiling, not a forecast.

How do you calculate reachable accounts? Multiply the market count by five survival rates: ICP match, role identified, contact record found, address deliverable, and not suppressed or already contacted. From the 93,903 US establishments BLS counts in NAICS 541613, a chain of 12%, 70%, 88%, 80% and 85% leaves 4,720 accounts, or 5.03% of the start. Every rate needs a named source or a written-down assumption.

Why is my addressable market smaller than my TAM? A TAM counts companies; reachability counts contactable people, with four or five filters in between. Official counts also overstate companies: BLS defines its establishment count as an annual average of quarterly establishment counts, not a count of unique corporate parents. Add role identification, contact coverage, deliverability and suppression, and single-digit survival is ordinary arithmetic.

What is a good contact match rate? One measured on your segment. Clay's test found different leading tools in its SMB and enterprise samples, and Anymail Finder's benchmark cautions that its results may not generalize to other regions, industries or seniority levels. A headline rate describes a vendor's best-covered segment, so test your own sample before planning on it.

How many accounts do you need to hit a pipeline target? Divide the target by your measured reply-to-conversation rate, then divide again by your compounded reachability. At 5.03%, 50 conversations needs about 2,500 reachable accounts at a 2% reply rate, which implies a starting market of roughly 49,700 companies. Halving any single stage doubles the market you need.

Related on the hub

• The 12-slide seed pitch deck that raised $187M: real teardowns — for when the playbook turns into a raise.
• Filter Databases vs Research Agents: One-Sentence ICP Test — Related cold outreach guide.
• Buying a Lead List vs Building One: Cost per Verified Contact 2026 — Related cold outreach guide.
• Best B2B Lead Generation Tools 2026: 15 for Founders — Related cold outreach guide.