## What this case shows

The most convincing review is behaviour: they came back and ran it again.

The role profile and messaging built in round one were reused in round two at no extra setup cost.

Scarce hybrid roles attract almost no inbound. Targeted outreach is the route that scales.

## The full funnel

### Production data. Absolute counts.

Contacted is the denominator for every percentage, and a person who appears under several roles is counted once.

Contacted 1,412  100%Opened (pixel) 879  62.3% · step 62.3%Replied 116  8.2% · step 13.2%Interested 101  7.2% · step 87.1%Resume 50  3.5% · step 49.5% Width uses a square-root scale so a 2 percent stage is still a slice, not a hairline. The count is the fact. Opened is hatched because it is a tracking pixel.

### Per 100 people contacted

- 4 resume
- 3 interested, no resume
- 1 replied only
- 92 no reply

One cell is one person out of 100 contacted. Cells are exclusive: a resume is not counted again as interest.

### Exact counts

| Stage       | Count | Of contacted | Step   |
| ----------- | ----- | ------------ | ------ |
| Contacted   | 1,412 | 100%         | -      |
| Opened      | 879   | 62.3%       | 62.3%  |
| Replied      | 116   | 8.2%        | 13.2%  |
| Interested  | 101   | 7.2%        | 87.1%  |
| Resume      | 50    | 3.5%        | 49.5%  |

Interview booking was not used by this account inside the window. The funnel stops at resume. All 28 confirmed interviews on the customers page belong to the largest account.

## 2 rounds, one flow

February proved it. April reused it as-is, one day from decision to launch.

1,412 contacted

Both rounds combined. Unique people.

101 interested

7.2 percent of everyone contacted.

50 resumes

49.5 percent of interested people sent a resume.

## How fast it happened

### Raw timestamps, from the first wave.

These are real intervals in this account, measured from the first outreach wave. They are not a 24-hour delivery claim, and they are not a fill-time claim.

1. **Day 0**  
   First wave of round one sent.
2. **+2 min**  
   First interested reply.
3. **+11 min**  
   First resume received.
4. **+2 months**  
   Round two for the expansion. Same flow, reused as-is.

## How much the AI wrote

### 4,136 personalized emails.

First touch plus automatic follow-ups, each written for that person, with no template blast. The hour count below is the only estimate on this page: it assumes 3 minutes to research and write one email by hand.

4,136  
emails written and sent by Mira

207  
hours at 3 minutes each

26  
working days at 8 hours

8.2%  
reply rate on 1,412 contacted

4,136 × 3 min = 207 hours ≈ 26 working days. Substitute your own minutes-per-email and recompute.

## FAQ

### How to read this account.

### Are the two rounds counted separately in the funnel?  
No. The 1,412 contacted, 101 interested, and 50 resumes are the combined extract for the account. The timeline is what shows they ran twice.

### What was reused in round two?  
The role profile and the outreach approach from round one. The board's claim is zero extra setup cost for that reuse, not zero work on the new candidates.

## Against the other four

### Reply rate in this extract.

Same scale as the hub chart, with this account marked. The 55-role account is the largest by volume and the lowest by reply rate, and nothing in the set is dropped.

0612  
percent

1. [Many roles at once](/content/customers/talent-agency/index.html)  
   7.3%  
2. [Hard-to-reach PMs](/content/customers/consumer-tech/index.html)  
   11.4%  
3. [No recruiter on staff](/content/customers/product-startup/index.html)  
   9.5%  
4. [Engineering outreach](/content/customers/recruiting-services/index.html)  
   9.2%  
5. [A role they ran twice](/content/customers/data-team/index.html)  
   8.2%

A dot, not a bar. Axis goes to 12 percent. This account is the filled mark.