Tabellab

Weekly Send Load: Week Numbers, Sends per Week, Emails per Subscriber and the Busiest Week

Lesson 1 of 20Lab

All lessons
  1. Weekly Send Load: Week Numbers, Sends per Week, Emails per Subscriber and the Busiest WeekLabFree
  2. Promotional vs Editorial Mix: Weekly Totals, Promo Share, a Cap Check and the Overall BalanceLab Locked
  3. Frequency Bands: Unsubscribe Rate, Revenue per Subscriber, Revenue per Email and the Average FrequencyLab Locked
  4. The Unsubscribe Curve: Rate per Test Arm, the Added Rate per Extra Send, a Steep-Step Flag and a Linear ForecastLab Locked
  5. The Fatigue Curve: Open Rate by Position in the Week, an Index Against the First Email and the First Fatigued EmailLab Locked
  6. Revenue per Send: Revenue per 1,000 Recipients, Contribution After Costs, Weighted Averages by Type and RankingsLab Locked
  7. The Marginal Send: Extra Revenue per Extra Email, Net Weekly Value After Unsubscribe Cost and the Best FrequencyLab Locked
  8. Mid-Course Project: A Cadence Plan by Segment with a Rule Lookup, Weekly Volume, Weekly Revenue and the List AverageLab Locked
  9. List Growth Under Two Cadences: Weekly Churn, a Closed-Form Projection with POWER, the Gap Between Scenarios and the Week It OpensLab Locked
  10. Best Day and Slot: Revenue per 1,000 by Send, Weighted Averages by Weekday, a Rank and the Strongest DayLab Locked
  11. Spacing Rules: Hours Between Sends, a Minimum-Gap Check, Hours to Move and a Colour Flag on the CalendarLab Locked
  12. Frequency Caps and Priorities: Ranking Competing Campaigns, Send or Hold Decisions and the Revenue the Cap Holds BackLab Locked
  13. Is the Difference Real? A Two-Proportion z-Test on Unsubscribe Rates from a 3 vs 5 Emails a Week TestLab Locked
  14. The Extra Send Test: Incremental Contribution per Week, Extra Unsubscribes, Cumulative Net Value and Value per Extra EmailLab Locked
  15. Peak-Season Calendar: Week Start Dates, Uplift Against the Baseline, a Peak Cap Check and the Week of the Big SaleLab Locked
  16. Engagement Tiers on the Move: A Transition Matrix, Next-Month Counts with SUMPRODUCT, Net Change and the Weekly Email LoadLab Locked
  17. A Sunset Policy: Cadence by Days Since the Last Open, Emails Saved, the Sunset Pool and Its RevenueLab Locked
  18. What-If Grid: Net Monthly Value for Every Frequency and Saturation Level, with Mixed References, the Best Cadence and a Colour ScaleLab Locked
  19. The Preference Centre: Choices Instead of Unsubscribes, Save Rate, Revenue and Emails per Choice, and the Annual Contribution KeptLab Locked
  20. Final Project: The Annual Cadence Scorecard with Quarters, Revenue and Unsubscribes per 1,000 Emails, a Health Verdict, a Flag and a ChartLab Locked

You are not signed in — the lab works, but your progress is not saved. Sign up

Lesson 1 of 20 is free. Unlock all 20 hands-on labs, the final exam and your certificate.

Buy course — $9.90
Loading the spreadsheet…

Why an email calendar needs numbers

If you run an online shop, a magazine, a club or any business that talks to its customers by email, you have probably felt the tension: more emails mean more orders today, but too many emails make people stop reading, mute you or leave. Everyone on the team has an opinion about the right number, and the opinions are usually loud rather than measured. The way out is to turn the calendar from a list of ideas into a table of numbers that everyone can read the same way.

A calendar is more than a list of dates. It decides how often each person hears from you, how much of that is selling and how much is useful content, which segments get which message and what the whole programme costs in unsubscribes. This free lesson starts with the foundation of all of it: the send load, the number of emails an average subscriber receives per week. The remaining nineteen lessons build on this number.

The example is Aldergrove Goods, a fictional online shop for pantry, home and garden products with 48,000 subscribers. All companies, campaigns and numbers in this article and in the course are invented for learning. Benchmarks are illustrative, and nothing here is a promise of results.

The core idea: sends per subscriber per week

Counting campaigns is the first instinct and the wrong one. Imagine a week with four campaigns: three go to the whole list and one goes to a small segment of customers who recently bought garden tools. A person outside the segment receives three emails; a person inside it receives four. The honest answer to "how many emails do subscribers get?" is therefore an average over the list:

  • Recipients of the week = the sum of the audience sizes of all sends in the week.
  • Emails per subscriber per week = recipients of the week ÷ list size.
  • A list-wide send adds exactly 1 to the average. A send to a quarter of the list adds 0.25.

The average is not what any single person feels, but it is the best single number for planning, because it adds up correctly: you can compare weeks, compare plans and compare it with the load of last year. Later lessons split the average into segments, so you can see who feels the most.

Only email people who have agreed to receive your messages, keep the unsubscribe link working in every message and honour an opt-out immediately. A calendar that respects consent is also the one that keeps its list healthy.

The data: 14 sends in four weeks

Here is the March calendar of Aldergrove Goods exactly as it appears in the lab above. The list holds 48,000 subscribers, week 1 starts on Monday 1 March 2027, and every send has a date, a campaign, a type (promotional or editorial) and a number of recipients. The last column shows the week number you will calculate in step 1.

Send dateCampaignTypeRecipientsWeek
Mon 1 MarSpring pantry refreshPromo48,0001
Wed 3 MarRecipe: wild garlic pestoEditorial48,0001
Fri 5 MarGarden starter offerPromo30,0001
Tue 9 MarGuide: kitchen herbsEditorial48,0002
Thu 11 MarFree-delivery weekendPromo48,0002
Sat 13 MarGarden tools flash salePromo30,0002
Mon 15 MarNew arrivalsPromo48,0003
Tue 16 MarComposting basicsEditorial22,0003
Thu 18 MarPantry bundle offerPromo48,0003
Fri 19 MarLast chance: bundlePromo20,0003
Tue 23 MarEaster table ideasEditorial48,0004
Wed 24 MarEaster baking offerPromo48,0004
Fri 26 MarOrder by Tuesday for EasterPromo48,0004
Sat 27 MarBehind the scenes: packing dayEditorial18,0004

The March calendar: one circle per sendMonTueWedThuFriSatSunPromotionalEditorialNumber = recipients in thousandsWeek 13 sendsWeek 23 sendsWeek 34 sendsWeek 44 sends4848304848304822482048484818Week 1 starts Monday 1 March 2027. Fictional data.
Figure 1. Fourteen sends in four weeks. The number in each circle is the audience in thousands of the 48,000 subscribers: a circle with 48 reaches everybody, a circle with 18 reaches a segment.

The picture already tells a story. The calendar is not evenly spread: week 1 has three sends, week 3 and week 4 have four each, and the circles in the last week sit close together on Tuesday, Wednesday, Friday and Saturday. Which of those weeks puts the heaviest load on the list? The numbers will say.

Step 1: turn dates into week numbers

To group sends by week you need a week number for every row. Excel has a WEEKNUM function, but it counts weeks of the calendar year, restarting every January and depending on regional settings. A planning period starts when you say it starts. The robust approach uses only subtraction, because Excel stores dates as serial numbers: 1 March 2027 is 46447 and 2 March is 46448.

Week = INT((send date − week 1 start) ÷ 7) + 1. Subtracting the start gives the number of days since the start. Dividing by 7 gives weeks as a fraction, INT cuts the fraction off and the final +1 makes the first week "week 1" instead of "week 0". With the start date locked in cell C3, the formula in E6 is =INT((A6-$C$3)/7)+1, copied down to E19.

Check it by hand once. "Free-delivery weekend" was sent on Thursday 11 March, which is 10 days after the start. 10 ÷ 7 = 1.43, INT gives 1 and +1 gives week 2. A Sunday 7 March would be 6 days after the start, 6 ÷ 7 = 0.86, INT gives 0, +1 gives week 1: correct, because the first week runs from Monday 1 March to Sunday 7 March.

Step 2: count the sends and add up the recipients

With a week number in every row, two functions do the grouping. COUNTIFS($E$6:$E$19,G6) counts the sends in the week named in G6, and SUMIFS($D$6:$D$19,$E$6:$E$19,G6) adds up their recipients. The results for the four weeks:

WeekSendsRecipientsEmails per subscriber
13126,0002.625
23126,0002.625
34138,0002.875
44162,0003.375

Notice that weeks 1 and 2 both have three sends and the same 126,000 recipients, while week 3 has four sends but only 138,000 recipients, because two of its four sends went to segments. The count of sends and the load on the list are related, but they are not the same thing.

Step 3: emails per subscriber per week

The last step divides the recipients by the list size: =SUMIFS($D$6:$D$19,$E$6:$E$19,G6)/$C$2. The list size in C2 is locked with dollar signs so that every row divides by the same number. The results are 126,000 ÷ 48,000 = 2.625 for week 1, 2.625 for week 2, 2.875 for week 3 and 3.375 for week 4.

Emails per subscriber per week012342.625Week 13 sends2.625Week 23 sends2.875Week 34 sends3.375Week 44 sendsAverage 2.875Emails
Figure 2. Emails per subscriber per week = recipients of the week ÷ 48,000. Week 4 is the busiest at 3.375 although it has the same number of sends as week 3: three list-wide sends and one segment send add up to more than the average week.

The average over the four weeks is 2.875 emails per subscriber per week, which sounds modest. The busiest week, found with =MAX(I6:I9), is week 4 with 3.375, and that is 17 % above the average. Subscribers do not experience averages: they experience the worst week, and unsubscribes and complaints tend to cluster there.

Why campaigns and load are different numbers

Consider week 4 in detail. The three list-wide sends (Easter table ideas, Easter baking offer and the Friday order reminder) already deliver 3.0 emails to everybody. The editorial "Behind the scenes" email reaches 18,000 people, which is 37.5 % of the list, and adds 0.375. A subscriber in that segment receives four emails in the week; a subscriber outside it receives three. The average, 3.375, sits between the two and describes nobody exactly.

  • Average load is what you plan, report and compare.
  • Individual load is what a person feels. It depends on the segments they belong to and is the subject of the cadence lessons later in the course.
  • Number of campaigns is what the team sees on the planning board. It is the least useful of the three, because audience sizes differ so much.
Add a column "Share of list" (recipients ÷ list size) next to the recipients. A glance down the column shows which sends are list-wide and which are targeted, and the weekly sums of that column are exactly the emails per subscriber.

Step by step in Excel

  1. Type the list size in C2 and the date of the first Monday in C3. Make sure the send dates in A6:A19 are real dates and not text: select them and press Ctrl + Shift + ~; real dates turn into numbers like 46447.
  2. Click E6 and type =INT((A6-$C$3)/7)+1. Lock the start date with F4 (Mac: Cmd + T) after clicking C3. Fill down to E19 with Ctrl + D.
  3. Type the week numbers 1 to 4 in G6:G9. In H6 enter =COUNTIFS($E$6:$E$19,G6) and fill down to H9.
  4. In I6 enter =SUMIFS($D$6:$D$19,$E$6:$E$19,G6)/$C$2 and fill down to I9. Format the column with three decimals.
  5. In I11 enter =MAX(I6:I9) for the peak week load.
  6. Cross-check: =SUM(D6:D19)/C2/4 must equal the average of I6:I9, which is 2.875. If it does not, a send has fallen outside weeks 1 to 4, usually because of a typo in a date.
  7. Highlight the busiest week: select I6:I9, choose Home › Conditional Formatting › Top/Bottom Rules › Top 1 Items and pick a light fill.

Common mistakes and how to spot them

  • Forgetting the +1. Without it the first week is week 0, the sends of the last week are labelled week 3, COUNTIFS for week 4 returns 0 and every count sits one week too early. Spot it when the weekly counts add up to fewer sends than the calendar has.
  • Using WEEKNUM for a custom planning period. The result depends on the year and the regional week system, so the weeks no longer start on your own Monday. Subtract the start date instead.
  • Counting campaigns instead of recipients. =H8 says that week 3 has four emails per subscriber, but two of the four sends went to segments, so the real load is 2.875.
  • Dividing by the wrong list. The list size changes during the year. Use the list size at the time of the send, or the average list of the period, and note which one you used.
  • An unlocked list size. /C2 copied down becomes /C3, then /C4, and the results jump around. Lock it as $C$2.
  • Averaging the weekly averages without weights. That works only if every week has the same list size. For long periods use total recipients ÷ list ÷ number of weeks.
  • Ignoring the peak. A comfortable average can hide one very heavy week. Always look at the maximum as well.

How to read the result

Is 2.875 emails per week a lot? There is no universal answer, which is why serious benchmarks must be illustrative. A weekly newsletter with an occasional offer may land at one or two; a shop with daily deals may plan five or more. What matters is how your own subscribers react as the load changes, and that is exactly what the next lessons measure: the unsubscribe rate at different frequencies, the drop in opens from the first to the sixth email, and the revenue each extra email brings.

  • If the peak week is far above the average, look at the calendar for clusters before cutting content.
  • If the average rises month after month without a plan, campaigns are being added faster than they are being retired.
  • If list-wide sends dominate, there is room to target more, which lowers the load for people who are not interested.

Treat the load as a control variable, like a budget. It is set deliberately in the calendar, reported every week and reviewed together with engagement and unsubscribes.

Quick reference: functions in this lesson

FunctionWhat it does hereExample
INTCuts off the decimals to give whole weeks=INT((A6-$C$3)/7)+1
COUNTIFSCounts the sends of a week=COUNTIFS($E$6:$E$19,G6)
SUMIFSAdds up the recipients of a week=SUMIFS($D$6:$D$19,$E$6:$E$19,G6)
MAXFinds the busiest week=MAX(I6:I9)
$ (absolute reference)Keeps list size and start date fixed when copying$C$2, $C$3

Real-world uses

  • Monthly calendar review. Export next month's planned sends with their audience sizes and check the weekly load before the plan is approved.
  • Agreeing a cap. Marketing, product and customer service all want to email the list. A weekly load number turns the argument into a shared budget.
  • Spotting hidden clusters. Two teams booking the same Thursday is invisible on separate boards and obvious in a weekly total.
  • Reporting. One line per week, emails per subscriber, goes into the monthly report next to open rates and unsubscribes.

What you will build in the rest of this course

Lesson 1 is free. Lessons 2 to 20 make up the paid course, each with its own live lab, and each builds on the weekly load you calculated today.

  • Promotional vs Editorial Mix. Promo share per week, a cap check and the overall balance, with a stacked column chart.
  • Frequency Bands. Unsubscribe rate, revenue per subscriber and revenue per email for subscribers who received one to six emails a week.
  • The Unsubscribe Curve. The added rate of each extra weekly send, steep-step flags and a straight-line forecast.
  • The Fatigue Curve. Open rate by position in the week, an index against the first email and the first fatigued email.
  • Revenue per Send. Revenue per 1,000 recipients, contribution after costs and weighted averages by type.
  • The Marginal Send. Extra revenue per extra email, net value after unsubscribe cost and the best frequency.
  • Mid-Course Project. A cadence plan by segment with a rule lookup, weekly volume and revenue.
  • List Growth Under Two Cadences. A projection with POWER and the week the gap passes a limit.
  • Best Day and Slot. Weighted revenue per 1,000 by weekday, ranks and the strongest day.
  • Spacing Rules. Hours between sends, a minimum-gap check and a colour flag.
  • Frequency Caps and Priorities. Ranking competing campaigns and the revenue a cap holds back.
  • Is the Difference Real? A two-proportion z-test on unsubscribe rates.
  • The Extra Send Test. Incremental contribution, extra unsubscribes and the net value per extra email.
  • Peak-Season Calendar. Week dates, uplift against the baseline and a peak cap.
  • Engagement Tiers on the Move. A transition matrix and next-month tier sizes with SUMPRODUCT.
  • A Sunset Policy. Cadence by days since the last open and the emails saved.
  • What-If Grid. Net monthly value for every frequency and saturation level, with mixed references.
  • The Preference Centre. Save rate, revenue per choice and the annual contribution kept.
  • Final Project. An annual cadence scorecard with quarters, a health verdict, a flag and a chart.

Summary

A good email calendar starts with the send load. Turn each date into a week number with =INT((A6-$C$3)/7)+1, count the sends of each week with COUNTIFS, add up the recipients with SUMIFS and divide by the list size to get the emails per subscriber per week. At Aldergrove Goods the four March weeks load the list with 2.625, 2.625, 2.875 and 3.375 emails; the average is 2.875 and the peak is week 4. Campaigns are not the same as load, so always use recipients, and always look at the peak as well as the average.

Now scroll up to the lab and build the sheet yourself. It takes about fifteen minutes, the Coach checks each step, and you will end up with a weekly load report you can apply to your own calendar.

Frequently asked questions

How many emails per week should I send to my subscribers?

There is no universal number. A useful starting point is to measure your current load in emails per subscriber per week, then test lower and higher frequencies and watch unsubscribes, opens and revenue. The right cadence is the one where an extra email still adds more value than it costs.

How do I calculate emails per subscriber per week in Excel?

Give every send a week number, add up the recipients of each week with SUMIFS and divide by the list size: =SUMIFS(recipients, week_range, week)/list_size. Lock the list size with dollar signs before you copy the formula down.

How do I turn a date into a week number counted from my own start date?

Use =INT((date - start_date)/7)+1 with the start date locked, for example =INT((A6-$C$3)/7)+1. This avoids WEEKNUM, which counts weeks of the calendar year and depends on regional settings.

Why is the number of campaigns not the same as the number of emails per subscriber?

Campaigns can go to different audiences. A campaign sent to a quarter of the list adds only 0.25 emails per subscriber on average, while a list-wide campaign adds 1. Always weight by recipients, not by the number of campaigns.

Should I look at the average week or the busiest week?

Both. The average is the number you plan and report, but subscribers feel the busiest week, and unsubscribes cluster there. Use MAX over the weekly loads to find the peak and compare it with the average.

Is it allowed to email everyone on my list every week?

You may email people who have agreed to receive your messages, and you must honour every opt-out. Rules differ between countries, so check the requirements that apply to you. This course teaches the arithmetic of cadence, not legal advice.

Keep going: Email Campaign Calendars & Send Frequency: Weekly Send Load, Promo vs Editorial Mix, Fatigue & Unsubscribe Curves, Revenue per Send and Segment Cadence

This free lab is lesson 1 of 20. Unlock the other 19 hands-on labs, the final exam and your certificate for a one-time $9.90.

  1. Promotional vs Editorial Mix: Weekly Totals, Promo Share, a Cap Check and the Overall Balance
  2. Frequency Bands: Unsubscribe Rate, Revenue per Subscriber, Revenue per Email and the Average Frequency
  3. The Unsubscribe Curve: Rate per Test Arm, the Added Rate per Extra Send, a Steep-Step Flag and a Linear Forecast
  4. The Fatigue Curve: Open Rate by Position in the Week, an Index Against the First Email and the First Fatigued Email
  5. Revenue per Send: Revenue per 1,000 Recipients, Contribution After Costs, Weighted Averages by Type and Rankings
  6. The Marginal Send: Extra Revenue per Extra Email, Net Weekly Value After Unsubscribe Cost and the Best Frequency
  7. Mid-Course Project: A Cadence Plan by Segment with a Rule Lookup, Weekly Volume, Weekly Revenue and the List Average
  8. List Growth Under Two Cadences: Weekly Churn, a Closed-Form Projection with POWER, the Gap Between Scenarios and the Week It Opens
  9. Best Day and Slot: Revenue per 1,000 by Send, Weighted Averages by Weekday, a Rank and the Strongest Day
  10. Spacing Rules: Hours Between Sends, a Minimum-Gap Check, Hours to Move and a Colour Flag on the Calendar
  11. Frequency Caps and Priorities: Ranking Competing Campaigns, Send or Hold Decisions and the Revenue the Cap Holds Back
  12. Is the Difference Real? A Two-Proportion z-Test on Unsubscribe Rates from a 3 vs 5 Emails a Week Test
  13. The Extra Send Test: Incremental Contribution per Week, Extra Unsubscribes, Cumulative Net Value and Value per Extra Email
  14. Peak-Season Calendar: Week Start Dates, Uplift Against the Baseline, a Peak Cap Check and the Week of the Big Sale
  15. Engagement Tiers on the Move: A Transition Matrix, Next-Month Counts with SUMPRODUCT, Net Change and the Weekly Email Load
  16. A Sunset Policy: Cadence by Days Since the Last Open, Emails Saved, the Sunset Pool and Its Revenue
  17. What-If Grid: Net Monthly Value for Every Frequency and Saturation Level, with Mixed References, the Best Cadence and a Colour Scale
  18. The Preference Centre: Choices Instead of Unsubscribes, Save Rate, Revenue and Emails per Choice, and the Annual Contribution Kept
  19. Final Project: The Annual Cadence Scorecard with Quarters, Revenue and Unsubscribes per 1,000 Emails, a Health Verdict, a Flag and a Chart

Buy course — $9.90 See the full course