A demand-gen director I work with sent me a screenshot of her HubSpot dashboard last fall. 184,000 contacts. 12% average open rate. She wanted me to tell her if the open rate was good or bad. I asked her a different question: how many of those 184,000 had ever opened a single email from her company. The answer, when she dug it out, was 71,000. The other 113,000 had been on her list for an average of 19 months and had never engaged with anything.
That conversation is why email engagement scoring matters in 2026 in a way it didn't five years ago. Not as a sophisticated lead-scoring exercise. Just as the basic act of distinguishing between contacts who have ever responded to you and contacts who have not. That single split changes how you read every other metric in your email program.
Strip away the sophistication. The fundamental question email engagement scoring answers is: has this contact ever shown me they're interested? Everything else — tiered scoring, decay weighting, behavioral models — is refinement on top of that one signal.
The binary version of email engagement scoring is the most important version. A contact is either:
That's it. You can layer recency and frequency on top of this for more granularity. But the binary split alone will tell you more about your email program than any single engagement-rate number on a dashboard.
Here's the asymmetry that nobody talks about: when you send the same email to an engaged contact and a never-engaged contact, you are not running the same campaign. You are running two campaigns whose outcomes look nothing like each other, and HubSpot's default reporting blends them together so completely that you can't see which one is doing the work.
In every B2B HubSpot portal we've audited, the engaged segment generates 85-95% of total email-attributed revenue while representing 30-50% of the list. That's not a marketing trope. It's a measurement artifact of how recipient behavior actually works: people who have shown intent in the past are dramatically more likely to show intent in the future.
Sending to never-engaged contacts isn't neutral. It's actively damaging. Mailbox providers measure your sender reputation partly by what percentage of your recent sends got opened, clicked, or replied to — Google's sender guidelines are explicit about this signal. Every email you send to a contact who never responds drags that ratio down. After enough of these sends, your reputation drops, and your engaged contacts start seeing your emails in spam.
That's the asymmetric loss: sending to never-engaged contacts doesn't just fail to convert them. It actively degrades your ability to convert the engaged contacts you actually care about. This is exactly the dynamic we describe in over-messaging, the silent deliverability killer — the volume that feels harmless is the volume that costs you.
When you look at a single "12% open rate" on the dashboard, you can't tell whether that's 30% from engaged contacts and 0% from never-engaged contacts (a healthy program with a list hygiene problem) or 12% across the board (an engagement collapse in your active segment). Those two scenarios call for completely different interventions. The aggregate number doesn't distinguish.
Here's the principle that's worth tattooing somewhere: any email metric reported across an unsegmented list is the average of two populations whose behaviors don't average.
HubSpot doesn't ship with engagement scoring turned on. You have to build it. The minimal version takes about an hour and consists of a custom contact property and an active list.
Pick a window that matches your typical sales cycle. For most B2B portals, 90 days is the right default. For very long-cycle products (12+ month sales cycle), 180 days. For high-velocity transactional products, 30 days. The window should be long enough to capture a real reactivation opportunity and short enough that the data still reflects current behavior.
Build a custom contact property called something like "Email Engagement Status" with values: Engaged, Never-Engaged, Lapsed. The default is Never-Engaged. The property gets updated by a HubSpot workflow that watches for any email open, click, or reply within your engagement window.
Three active lists. Engaged contacts (opened, clicked, or replied in last 90 days). Lapsed contacts (engaged at some point in their history but not in the last 90 days). Never-engaged contacts (no engagement on record, ever). Now every report you build, every audience you target, every campaign you ship can be filtered by this split.
This is where most marketers stop. They build the segmentation, look at it once, and then continue blasting everything to the full list anyway. The point isn't the score. It's changing what you do with the never-engaged segment. Stop sending them the regular cadence. Run them through a quarterly reactivation sequence. If they don't respond, suppress them.
The mechanical setup above works, but it has a problem: HubSpot's engagement tracking is based on the same opens that Apple Mail Privacy Protection inflated. A "Never-Engaged" contact who has Apple Mail and MPP enabled may have actually been opening your emails for two years — you just can't see it. And an "Engaged" contact may be a never-engaged contact whose Apple server pre-fetched the tracking pixel. The full damage to open-based logic is laid out in why your open rate doesn't mean what you think it means post-MPP.
That's why the most useful engagement signals in 2026 are click-based, reply-based, and form-fill-based, not open-based. If you have to choose one metric to anchor engagement scoring, choose clicks. A click is a deliberate action that survives the MPP problem.
And that's also why we built Activation Status Analytics on top of HubSpot in September 2025. It separates your contact list into engaged and never-engaged using the cleanest available signals, accounts for the MPP inflation by weighting clicks and replies more heavily than opens, and shows you the engagement split inside your existing HubSpot reports without requiring you to build the property scaffolding yourself. The point isn't to replace HubSpot's reporting. It's to give you the engaged-vs-never-engaged lens that HubSpot's native dashboards don't have.
This is the part of email engagement scoring that requires the most discipline and pays back the most. Once you can see your never-engaged segment as a distinct population, you have three legitimate options.
Build a quarterly reactivation sequence specifically designed for never-engaged contacts. It runs four times a year, lasts three emails, asks them directly whether they still want to hear from you, and includes a single high-value offer relevant to their original signup intent. Most never-engaged contacts won't reactivate. The 5-10% who do are recovered revenue.
If you're not willing to suppress them entirely, at least stop sending them the same cadence as your engaged segment. Drop never-engaged contacts to one send per month or less. The deliverability hit per send is smaller, the chance of accidental reactivation is preserved, and your engaged contacts benefit because your sender reputation improves.
The hardest option and often the right one. After 18-24 months of zero engagement and a failed reactivation sequence, suppress the contact from marketing sends. They're not coming back. Continuing to mail them is paying a deliverability tax for an audience that has chosen not to participate. Suppression isn't deletion. The record stays. You just stop sending to it. Our recycling framework for email list management walks through how to sort the suppress-now contacts from the still-recoverable ones.
Lead scoring measures total purchase intent across many signals (page views, form submissions, demo requests, fit attributes). Engagement scoring is a narrower question: is this contact responding to my email program. A high lead score with low engagement score means you have a hot prospect who's not seeing your emails. That's a deliverability or send-time problem, not a content problem.
90 days for most B2B contexts. 30 days for transactional or high-velocity programs. 180 days for very long sales cycles. The right window is the one that gives you enough data to see a pattern without being so wide that lapsed behavior looks active. Shorter windows are more responsive but noisier. Longer windows are smoother but slower to catch list decay.
Both, but weight clicks more heavily. Opens are inflated by Apple Mail Privacy Protection and aren't reliable as a sole signal. Clicks, replies, and form fills are deliberate actions that survive MPP. A reasonable weighting is: click counts as one engagement, open counts as one-third of an engagement, reply counts as five engagements.
Reframe it. You're not shrinking the list. You're consolidating volume toward the segment that's driving revenue. Show them the engagement-segmented revenue attribution. In most B2B portals, 90% of email revenue comes from 40% of the list. Suppressing the inactive 60% doesn't reduce revenue; it improves deliverability for the 40% who are paying for the program.
Email engagement scoring isn't sophisticated. It's basic hygiene that most HubSpot portals never set up because the default reporting view doesn't make the absence obvious. Once you can see your list split into engaged and never-engaged, every metric you've been chasing starts to make more sense, and several decisions that felt impossible (suppression, frequency caps, cadence changes) start to feel obvious.
If you want to see your engaged-vs-never-engaged split without building the property scaffolding manually, the free trial of Seventh Sense reports it inside your HubSpot portal in about 15 minutes. You'll find out what percentage of your list is real. Most marketers are surprised by the answer.
The list you have isn't the list you have. The list you have is the engaged subset of the list you have. Once you can see the difference, you can do something about it.