What a good marketing team replaces open rate with

Retiring open rate is an easy argument on paper. In practice it is hard because the dashboard has to say something. The CMO looks at it on Monday morning and wants a number. The weekly report needs a headline figure. The copywriter wants to know which subject line worked.

This piece is the practical replacement kit. Four metrics, each tied to a specific question, each reasonably resistant to proxy pre-fetch and scanner contamination. Deploy the stack once, retire open rate for good.

Replacement 1: Inbox placement rate

Inbox placement is the percentage of your sent messages that land in the recipient's primary inbox folder, as opposed to promotions, spam, or updates. It is the metric open rate was pretending to be a proxy for — did the message reach the human at all?

Measurement is done by seeding. You add a small set of seed mailboxes to each campaign (one per provider: Gmail, Outlook, Yahoo, Apple iCloud, Mail.ru, Yandex, ProtonMail, etc.). A service then inspects each seed inbox after send and records which folder the message landed in.

Why it works: it does not use pixels. Folder placement is determined by inspecting the mailbox directly. There is no proxy, no scanner, no pre-fetch. The number is ground truth for the seed accounts, and the seed accounts are a reasonable proxy for the rest of your list if you pick them carefully.

What "good" looks like

• Transactional: 95%+ primary-inbox rate across all major providers.
• Opt-in newsletter: 85%+ primary / 10% promotions / <5% spam.
• Cold outbound: 60%+ primary is excellent; 40–60% is acceptable; below 40% means your reputation is damaged.

Replacement 2: Reply rate

Humans reply. Pixels do not. If you measure the percentage of your sent messages that received a human reply within 7 days, you have a metric that is essentially immune to pre-fetch and scanner distortion.

Reply rate is the gold standard for outbound and relationship marketing. For broadcast newsletters it is less applicable, because most newsletters do not invite replies. You can still proxy it with "reply-to-sender" rates on send-from addresses that accept mail.

Measurement

Most ESPs now track reply rate natively if the reply-to address routes back through their systems. If you use a custom reply-to, you need to pipe the mailbox back to the ESP via API or webhook. Both approaches are well-supported as of 2027.

Benchmarks:

• Cold outbound: 1–5% reply rate is healthy.
• Warm outbound (prior interaction): 8–15%.
• Broadcast newsletter: 0.1–1% reply rate (most readers do not reply to newsletters even when they love them).

Replacement 3: Decorated-link click-through

Raw click rate is partially polluted by security scanners that "pre-click" every link to check for malware. You can filter most of that noise by using link decorators that require a two-step interaction: the link hits a redirector, the redirector sets a cookie or session token, and the final destination requires that token.

Security scanners follow the first redirect but usually do not complete the full flow. Humans do. The decorated click rate is therefore a cleaner signal than raw click.

Implementation

Most ESPs and link-wrapping services support a "strict mode" or "verified click" option that implements the two-step flow. If yours does not, you can implement it in about 20 lines of Node.js:

// 1. Wrap every link as https://link.example.com/r?t=<token>
// 2. On GET /r, set a short-lived cookie and redirect
// 3. Landing page requires the cookie; missing cookie = bot

app.get('/r', (req, res) => {
  const { t } = req.query;
  res.cookie('vk', t, { maxAge: 60_000, httpOnly: true });
  res.redirect(lookupDestination(t));
});

app.get('/', (req, res) => {
  if (!req.cookies.vk) return res.redirect('/nocookie');
  // render landing page and count as a verified click
});

Replacement 4: Revenue per thousand sent (RPM)

The bottom-line metric. Total attributed revenue divided by messages sent, multiplied by a thousand. It is harder to game because it terminates at a cash transaction, and it is immune to every distortion above the call-to-action.

RPM is less useful for top-of-funnel and brand campaigns where attribution is muddy, but for any campaign with a direct-response CTA it is the cleanest success signal available.

Why RPM beats revenue in absolute terms

Revenue alone rewards sending more, even at the cost of list health. RPM normalises for send volume and reveals when you are degrading the asset. A campaign that earns $100K from 1M sends is healthier than one that earns $120K from 2M sends.

Putting the stack together

Here is the minimal weekly report that replaces the open-rate-led version:

Weekly Email Performance — Week of Jan 18, 2027
————————————————————————————————————————
Campaigns sent:          8
Messages delivered:      412,000
Inbox placement rate:    83% (Gmail 88%, Outlook 71%, Apple 92%)
Verified click rate:     4.2% (vs 3.8% prior week)
Reply rate:              0.3% (newsletters) / 2.1% (outbound)
Revenue per thousand:    $87 (vs $92 prior week)
Unsubscribe rate:        0.08%
Complaint rate:          0.01%
Notes: Outlook placement dip under investigation.

The report tells the reader whether mail arrived, whether humans engaged, which campaigns resonated, and whether the revenue held. Open rate adds nothing to that picture that is not already captured.

What to expect when you switch

1. The headline number will drop. Your old "open rate" of 42% will be replaced by something like "83% inbox placement, 4% click, 0.3% reply". Executives will need briefing.
2. Campaign rankings will change. Campaigns that won on open rate will not necessarily win on the new stack. That is a feature — it means the old ranking was partly an artefact of pre-fetch.
3. You will catch delivery problems faster. Inbox placement reacts within hours to a blocklist event. Open rate takes days to drift.
4. You will spend less time arguing about subject lines. With reply and click as the signals, subject-line A/B tests converge on content that produces action, not content that produces pixel fires.

FAQ