sales-email-tracking

What this skill does

# Email Engagement Tracking for Sales

Help the user set up, interpret, and act on email tracking data — from understanding how tracking works through interpreting engagement signals and timing follow-ups. This skill is tool-agnostic and applies to any email tracking platform (Yesware, Salesloft, Apollo, HubSpot Sales, Mailshake, Mixmax), with platform-specific sections where relevant.

## Step 1 — Gather context


If `references/learnings.md` exists, read it first for accumulated knowledge.

Ask the user:

1. **What's your tracking goal?**
   - A) Setting up email tracking for the first time
   - B) Understanding open/click data and what it means
   - C) Timing follow-ups based on engagement signals
   - D) Dealing with inaccurate tracking data (Apple MPP, pixel blocking)
   - E) Choosing a tracking tool
   - F) Tracking attachments and documents
   - G) Something else

2. **What tools do you use?**
   - A) Yesware
   - B) Salesloft
   - C) Apollo
   - D) Mailshake
   - E) HubSpot Sales
   - F) Mixmax
   - G) Mailtrack / other lightweight tracker
   - H) None yet

3. **What email client?**
   - A) Gmail
   - B) Outlook
   - C) Both

**If the user's request already provides most of this context, skip directly to the relevant step.** Lead with your best-effort answer using reasonable assumptions (stated explicitly), then ask only the most critical 1-2 clarifying questions at the end — don't gate your response behind gathering complete context.

## Step 2 — How email tracking works

### Core tracking mechanisms

- **Open tracking**: Invisible 1x1 pixel embedded in email. When email is opened, pixel loads from tracking server → open event recorded → notification sent. Works for HTML emails only — plain text emails cannot be tracked.
- **Click tracking**: Links rewritten to pass through tracking server. When clicked → redirect to destination + click event recorded → notification. More reliable than open tracking because it requires deliberate action.
- **Attachment tracking**: Attachments hosted as tracked links (or via DocSend/similar). When opened → view event recorded → notification. Often includes page-by-page analytics showing which slides/pages the prospect spent time on.
- **Real-time notifications**: Desktop/mobile push when prospect engages — opens, clicks, attachment views. Enables immediate follow-up when a prospect is actively engaged.
- **Thread tracking**: Track entire conversation thread, not just initial send. See when prospects re-open earlier messages — a signal they're reviewing the conversation before responding.

### How tracking signals map to follow-up actions

| Signal | What it means | Recommended action | Timing |
|---|---|---|---|
| **Single open** | Glanced at it | Wait — could be automated preview | Don't act yet |
| **Multiple opens (3+)** | Actively reviewing | Call or send follow-up | Within 1-2 hours |
| **Link click** | Researching your offering | Call immediately — hot lead | Within 30 minutes |
| **Attachment view** | Deep interest, evaluating | Call or send relevant case study | Within 1 hour |
| **Forward to colleague** | Sharing internally | Multi-thread — reach out to the forwardee | Within 24 hours |
| **No open after 48hr** | Didn't see it or ignored | Re-send with new subject line | After 48-72 hours |
| **Opened but no reply after 3 days** | Interested but not compelled | Follow up with new value angle | Day 3-4 |

### Priority framework

Rank engagement signals from strongest to weakest:

1. **Reply** (99% reliable) — they responded, follow up immediately
2. **Link click** (85-90% reliable) — they took action, follow up within 30 minutes
3. **Attachment view** (85-90% reliable) — deep engagement, follow up within 1 hour
4. **Multiple opens** (60-70% reliable) — repeated interest, follow up within 1-2 hours
5. **Forward** (high reliability when detected) — internal discussion happening, multi-thread within 24 hours
6. **Single open** (low reliability) — don't act on this alone

**Critical caveat — always factor in tracking data inflation.** Apple MPP inflates open rates (pre-fetches pixels even when the email is never read), and bot clicks from corporate email security tools (Barracuda, Mimecast, Proofpoint, Microsoft Defender) inflate click-through rates by auto-clicking every link to scan for malware. Bot clicks happen within 1-2 seconds of delivery from data center IPs and hit all links simultaneously — human clicks are delayed and selective. Before acting on any open or click signal, consider whether it could be fake. The only truly reliable engagement signals are **replies** and **meetings booked**. See Step 3 for full details on identifying and filtering these false positives.

## Step 3 — Privacy and accuracy challenges

**The two biggest accuracy threats to email tracking data are Apple Mail Privacy Protection (MPP) and bot clicks from security scanners.** MPP inflates open rates; bot clicks inflate click rates. Together they can make both of your primary engagement metrics unreliable. Always warn about both when interpreting tracking data, and default to reply rate as the most trustworthy signal.

### Apple Mail Privacy Protection (MPP)

Launched with iOS 15 / macOS Monterey (September 2021):
- Pre-fetches all email content (including tracking pixels) regardless of whether the user actually opens the email
- Result: 100% "open rate" for Apple Mail users — the data is meaningless
- Prevalence: ~50-60% of consumer email opens, ~30-40% of B2B email opens
- **Workaround**: Don't rely on open rate as a primary metric. Focus on click-through rate and reply rate instead. If your open rate is suspiciously high (above 80%), Apple MPP is almost certainly the cause.

### Bot clicks from corporate email security

Security tools like Barracuda, Mimecast, Proofpoint, and Microsoft Defender pre-scan every link in inbound emails to check for malware. This generates fake click events that inflate your click-through rates with false positives. The problem is especially severe when selling to enterprise or security-conscious organizations, where nearly 100% of emails pass through link scanners.

- **How to spot them**: Bot clicks typically fire within 1-2 seconds of email delivery, come from the same IP block, and hit every link in the email simultaneously. Human clicks happen minutes to hours later with varied timing.
- **Mitigation**: Enable bot click filtering in your tracking tool (most modern tools support this). Use delayed redirect tracking if available. Set a minimum time threshold (e.g., ignore clicks within 5 seconds of delivery). For enterprise prospects, weight replies over clicks as your primary engagement signal — replies cannot be faked by security scanners.
- **Prevalence**: Bot clicks can inflate click rates by 10-50% depending on how many of your prospects use corporate email security. If your click rate seems suspiciously high for cold outbound (above 15-20%), bot clicks are likely the cause.

### Privacy-conscious users

- Image blocking in email clients prevents pixel loading
- ~15-25% of B2B recipients block images by default
- **Workaround**: Track replies and clicks (more reliable than opens). Accept that open tracking will always undercount by 15-25%.

### Tracking accuracy benchmarks

| Metric | Accuracy | Notes |
|---|---|---|
| **Open tracking** | ~60-70% | Inflated by Apple MPP, deflated by image blocking |
| **Click tracking** | ~85-90% | Some bot inflation from security scanners |
| **Reply tracking** | ~99% | Most reliable signal — hard to fake |
| **Attachment view tracking** | ~85-90% | Reliable when using tracked links (DocSend, Yesware attachments) |

## Step 4 — Platform-specific guidance

For platform-specific tracking setup and configuration (Yesware, Salesloft, Apollo, HubSpot Sales, Mailshake, Reply.io, Woodpecker, Seismic, Mixmax, SendGrid, Postmark, Customer.io, Mailgun, Klaviyo, ActiveCampaign, Veloxy, Cirrus Insight), see references/platforms.md.

## Step 5 — Actionable guid