Engagement scoring: the full calculation
This is the reference version, setting out the calculation in full. It explains exactly how a contact's engagement score is calculated, what moves it, and why the timings you see may differ from the numbers quoted elsewhere. If you only need the practical summary, start with Understanding contact engagement scoring.
What we'll cover
- The score is made of four parts
- Recency, 40%
- Frequency, 30%
- Quality, 20%
- Consistency, 10%
- Why the timings change with how often you send
- New contacts and the grace period
- Levels, lifecycle stages and trend
- A worked example
- Common questions
- Related articles
1 The score is made of four parts
Every contact's engagement score is a number from 0 to 100, calculated by combining four separate measures. Each is scored from 0 to 100 in its own right, then blended:
| Part | Weight | The question it answers |
|---|---|---|
| Recency | 40% | How long ago did they last engage? |
| Frequency | 30% | How often do they engage, across recent messages? |
| Quality | 20% | How meaningful was their most recent action? |
| Consistency | 10% | How many different ways do they engage? |
Recency carries the most weight deliberately. Someone who clicked yesterday is a better prospect than someone who clicked ten times last year, and mailbox providers take the same view.
Note: Journey emails count everywhere campaign emails do. A contact who never receives a broadcast but engages with your automated emails is scored on that engagement, not treated as silent.
2 Recency, 40%
Scored from the number of days since the contact last engaged with any of your email:
| Days since last engagement | Recency score |
|---|---|
| Today | 100 |
| Within 7 days | 95 |
| Within 14 days | 85 |
| Within 21 days | 75 |
| Within 30 days | 61 |
| Within 45 days | 45 |
| Within 60 days | 37 |
| Within 75 days | 25 |
| Within 90 days | 14 |
| Within 120 days | 8 |
| Within 180 days | 3 |
| Longer, or never | 1, or 0 if they have never engaged |
The curve is steepest between three weeks and two months, which is where a contact is most likely to be genuinely drifting rather than simply between sends.
3 Frequency, 30%
Measured across the contact's last five campaign emails and last five journey emails. Each action is worth a different amount, the total is divided by the number of messages, and the result is capped at 100:
| Action | Points |
|---|---|
| Reply | 50 |
| Click | 40 |
| Share | 30 |
| Forward | 30 |
| Open | 20 |
Because it is a rate rather than a total, a contact who receives one email a month and engages with all of them scores as well as one who receives one a week and does the same. Volume of mail does not inflate the score.
4 Quality, 20%
Based on the single most recent positive action the contact took:
| Most recent action | Quality score |
|---|---|
| Reply | 100 |
| Share | 90 |
| Forward | 90 |
| Click | 85 |
| Open | 50 |
| None | 0 |
Delivery is deliberately not a quality signal. A message arriving says nothing about whether the person wanted it. An open scores roughly half a click, because opens are the least reliable signal of the set and privacy features on some mail apps report them without the recipient doing anything.
5 Consistency, 10%
Rewards breadth rather than volume. A contact earns points once for each type of action they have taken across those recent messages:
| Has ever, in recent messages | Points |
|---|---|
| Replied | 30 |
| Clicked | 20 |
| Shared | 20 |
| Opened | 15 |
| Forwarded | 15 |
Someone who opens, clicks and replies is a more reliable prospect than someone who has opened twenty times and never done anything else, and this is the part of the score that says so.
6 Why the timings change with how often you send
This is the part most worth understanding, and the reason the day counts you see may not match the ones quoted in the summary article.
Thirty days of silence means something very different to a daily sender than to a quarterly one. If you send monthly, a contact who has not engaged for six weeks may simply have received two emails in that time. Treating them the same as a contact ignoring a daily newsletter would be wrong.
So every day-based threshold is multiplied by a factor derived from your actual sending rate to that audience, counting both campaigns and journey emails:
| How often you send | Multiplier |
|---|---|
| Weekly or more often | ×1 |
| About every two weeks | ×2 |
| About monthly | ×4 |
| Less than monthly | ×6 |
The multiplier is capped at ×3 where it applies to new-contact grace periods, and at ×4 where it applies to the decay thresholds. Without those caps a quarterly sender would have contacts sitting in an eighteen-month grace period, which helps nobody.
In practice, a contact goes At-Risk after 30 days if you send weekly, and after 120 days if you send monthly. Both are 30 days in your sending terms.
Note: The rate is calculated per audience, from your real sending history, and updates as your habits change. You do not configure it.
7 New contacts and the grace period
A brand-new contact has no engagement history, so scoring them on it would put every new signup straight into Inactive. Instead there are two tiers of protection, both scaled by the cadence multiplier above:
- First 30 days. The score cannot fall below 40, which keeps the contact at least Warm.
- Days 31 to 60. The score cannot fall below 30, which keeps them at least Cold.
These are floors, not fixed values. A new contact who engages immediately scores on their real behaviour and can be Hot on day one. The floor only catches contacts who have not engaged yet.
A contact is only marked Never Engaged after the onboarding period ends. Before that they stay New, because there has not yet been a fair chance to engage.
8 Levels, lifecycle stages and trend
The final score maps to a level:
| Score | Level |
|---|---|
| 70–100 | Hot |
| 40–69 | Warm |
| 15–39 | Cold |
| 0–14 | Inactive |
Lifecycle stage is a separate axis, based on time since last engagement rather than on the score, and also scaled by your cadence. The base thresholds are 30 days to At-Risk, 60 to Dormant and 90 to Lost.
Trend compares the current score with the previous calculation. A move of 10 points or more marks the contact as improving or declining; anything smaller is treated as noise rather than a direction.
9 A worked example
A contact who clicked 10 days ago, opened three of their last five emails and clicked one of them, and has both opened and clicked at some point:
- Recency: 10 days → 85. Weighted: 85 × 0.40 = 34.0
- Frequency: (1 click × 40 + 3 opens × 20) ÷ 5 messages = 20. Weighted: 20 × 0.30 = 6.0
- Quality: most recent action was a click → 85. Weighted: 85 × 0.20 = 17.0
- Consistency: opened (15) + clicked (20) = 35. Weighted: 35 × 0.10 = 3.5
Total: 60.5, rounded to 61. That is a Warm contact. To reach Hot they would need to engage more often or more deeply, not simply be sent more email.
10 Common questions
Why is a contact still Active when they have not engaged for two months?
Almost always the cadence multiplier. If you send monthly, At-Risk begins at 120 days rather than 30.
Why did a score drop when the contact did nothing wrong?
Recency decays with time, so a score falls on its own as the days pass. That is the intended behaviour: engagement from three months ago should not read the same as engagement from yesterday.
Does sending more email raise scores?
No. Frequency is a rate, not a total, and recency only responds to engagement. Sending more to an unengaged contact will not move their score, and it will hurt your sender reputation.
Do opens alone make someone Hot?
Very unlikely. An open is worth 20 in frequency against a click's 40, scores 50 in quality against a click's 85, and earns 15 of the consistency points. A contact who only ever opens tends to settle in the Warm band.
How often are scores recalculated?
On a regular schedule rather than in real time. Engagement from a campaign you have just sent appears at the next recalculation, not the moment somebody clicks, so a score you check minutes after a send will not yet reflect it.
Do transactional emails count?
No. Engagement scoring covers campaign and journey email, which the contact chose to receive. Transactional messages are triggered by the recipient's own actions and say little about their interest in your marketing.