DateTime distribution: when competitors send

What this screen is

Send timing is the one lever you can pull without rewriting a single word of the email. This screen reads the send time of every analyzed newsletter in your account and counts how many landed in each hour of the day, on each weekday and in each calendar month. What comes out is a picture of when the brands you follow actually press send, rather than when the advice blogs say they should.

On top of those counts the screen makes exactly one judgement, in the Best time to send card. For each brand it looks at every weekday and hour slot that brand has actually used, averages the estimated open rate of the newsletters sitting in each one, and picks the highest average among the slots with at least three sends. A confidence badge on every recommendation tells you how thin the evidence is. High needs six or more sends in the winning slot, medium covers three to five, and Thin sample means nothing cleared the bar, so what you see is the brand's busiest slot instead.

The open rates behind that ranking are estimates. Newsletrix receives competitor newsletters as an ordinary subscriber, so it can never see another sender's opens or clicks. It reads the content of each email and puts two figures on it instead, an estimated open rate and an estimated click-through rate, both on a 0 to 100 percent scale. Where it has no figure to give, that newsletter carries no estimate and adds nothing to the averages. Send counts, the heatmap and the monthly volume line are hard facts. The rates that rank them are informed guesses.

No industry benchmark appears anywhere on this screen. That is deliberate, and a notice bar at the top says so, alongside the timezone every hour on the page is expressed in. Nothing is converted here, so that timezone labels the numbers without changing any of them.

One thing the screen cannot do is recommend a slot nobody has tried. Best time to send only ranks slots that already appear in your collected data, so if every brand in your set sends at 09:00 on Tuesday, the quiet Thursday evening that might actually work for you will never come up. Find that gap yourself in the heatmap, by looking for the pale cells rather than the dark ones.

Open it when

• You are scheduling next week's send and want to see the day and hour your competitors cluster around, so you can match the crowd or deliberately avoid it.
• A brand you track just changed cadence and you want to confirm whether it also moved its send slot.
• You need a starting hypothesis for an A/B test on send time, with an honest signal of how thin the evidence behind it is.
• You want to find the hours of the week that are dead air in your category, so a send there lands in a quieter inbox.
• You are checking whether newsletter volume across your tracked set is growing, flat or seasonal over recent months.
• You need a quick census of which brands dominate the sample behind everything else on the screen, and how long each one's tracked history runs.

Screen anatomy

The screenshot below comes from an earlier release. The annotated map after it follows the current layout from top to bottom.

1. Timezone notice bar A muted strip stating that no industry benchmark is shown for send timing, and naming the timezone every hour on the page is expressed in. Read it once, then keep it in mind for every hour you see below.
2. Leading recommendation tile The first tile in Best time to send carries a star with the tooltip Top performer in this set. The list is sorted by average estimated open rate, so the star means best among the brands currently filtered, nothing more.
3. Thin sample tile A grey badge instead of a coloured one means no slot for that brand reached three sends with an estimated open rate. The tile then shows the brand's busiest slot and says so, with the line busiest cell, only N sends, need 3.
4. Hour × day-of-week heatmap Seven weekday rows by 24 hour columns. Darker means more sends. Totals per weekday run down the right rail, totals per hour sit in a strip under the grid, and the single busiest cell gets a ring plus a Peak chip in the card subtitle.
5. Monthly timeline One point per calendar month showing how many newsletters your tracked set sent. Only volume is plotted, despite the subtitle mentioning performance.
6. Top brands by volume The census behind everything above: count per brand plus the date of the first and last newsletter collected. It only appears when no brand is selected, and it stops at 15 rows.
7. Page header The title Date & Time distribution with its subtitle, plus a round help icon on the right that opens a plain-language explainer called Timing Is Everything. It has no effect on the data.
8. Filter card Brand, From date and To date, side by side. There is no Apply button. Change any field and the whole screen reloads immediately.
9. Loading and empty states While the screen is loading, every card below the filters is replaced by a single centred spinner. When there is nothing to show, each panel prints its own empty message and the brand table is not drawn at all.

Every panel explained

Timezone notice bar

A bordered muted strip directly under the page header. It states that send timing is highly audience and timezone specific, so no industry benchmark is shown, names the timezone in a monospace font, and reminds you that filtering by brand compares one brand against the rest of your portfolio.

How it is calculated. The strip names one timezone, usually UTC, and that name is a label only. Every hour and weekday on the page is the one the email itself carries, with no conversion anywhere, so a different label would not move a single cell in the heatmap.

How to read it. Treat it as the unit on every hour figure on the page, the way a chart axis is labelled in dollars or percent. If a brand you track targets a different region, do the arithmetic in your head before you copy its recommended slot.

Best time to send

A card with one tile per brand. Each tile shows the brand, a confidence badge, the recommended day and hour with the timezone in brackets, the average estimated open rate, the estimated CTR when there is one, how many sends the recommendation rests on, and the brand's total sends in the window.

How it is calculated. Every analyzed newsletter counts toward the brand, weekday and hour slot it was sent in, and every slot carries a send count and an average estimated open rate. A slot can be recommended only if it holds at least three sends and has an open rate to average. Among the slots that qualify, the highest average wins. When none qualifies, the tile shows the brand's busiest slot instead and marks it Thin sample. Confidence comes from the winning slot. Six or more sends reads High confidence, three to five Medium confidence, and a fallback reads Thin sample. Tiles are ordered by average open rate, highest first, with brands that have no open rate at all pushed to the bottom, and the first tile gets the star.

How to read it. Read a tile as one sentence: this brand's highest-average slot in the selected window was Tuesday at 10:00, based on eight sends. The badge tells you how much of that sentence to believe. Two brands agreeing on a slot is a much stronger signal than one brand with a high confidence badge.

The empty state

Before enough newsletters have been analyzed, each panel prints its own message rather than an empty chart.

Hour × day-of-week heatmap

A grid of seven weekday rows by 24 hour columns. Each cell is the number of newsletters sent in that weekday and hour slot across the current filter window. Hovering a cell shows the slot, the send count and the average estimated open rate, or no open data when there is none.

How it is calculated. Sends are counted per weekday and hour, then laid out Monday first. Shading is relative rather than absolute. The busiest cell on screen sets the peak, every other cell is scored as its count divided by that peak, and the bands fall at 0.2, 0.4, 0.6 and 0.8, giving five shades from a 20 percent tint to full colour. Zero sends renders as flat grey. The single busiest cell also gets a ring, and its slot is repeated as a Peak chip next to the card title. The right rail is the sum of each row and the strip underneath is the sum of each column, printed only under every third hour and only when the total is above zero.

How to read it. Read a row to compare weekdays, a column to compare hours. Dark blocks are crowded slots; pale and grey blocks are the quiet ones, which is where the interesting question usually is. Hour labels appear every third column, so counting cells across is often faster than reading the labels.

Monthly timeline

A filled line chart with one point per calendar month, labelled Newsletter count, under the subtitle Newsletter volume and performance over time.

How it is calculated. One point per calendar month, counting the newsletters sent in it, oldest month on the left. Nothing is smoothed. A month with no sends is left out of the line entirely rather than drawn as a zero, so the line can step over a gap without showing you that it did.

How to read it. A rising line means the set you track is sending more. Treat the first and last points with suspicion, because a window that starts or ends mid-month produces a partial month that looks exactly like a collapse in volume.

Top brands by volume

One row per brand, with columns for Brand, Count, First newsletter and Last newsletter. It only appears when the Brand filter is set to all brands.

How it is calculated. A straight count of analyzed newsletters per brand inside the current window, plus the earliest and latest send date for that brand in the same window, formatted as a short date such as Mar 4, 2026. Brands are ordered by count, highest first, and the table stops at 15 rows.

How to read it. This is the census behind everything above it. A brand with a handful of sends contributes very little to the heatmap, so check here before you read a pattern as market-wide. A brand whose Last newsletter is weeks old has either gone quiet or stopped reaching your tracking address.

Metric definitions

Filters and controls

Some things are deliberately absent. There is no export or download of any kind and no reset button for the filters. No Apply button, no comparison against an industry benchmark, and no separate hour-of-day or day-of-week bar chart any more. Set a From date later than the To date and the screen simply keeps whatever it was already showing. Clear the fields by hand or set the dropdown back to All brands.

How to use it

1. Set the window before you read anything The default covers the last 90 days, which is usually three or four points on the monthly line and rarely enough for a high confidence recommendation. Widen From date to six or twelve months first, then read.
2. Read the heatmap rails, not the cells Start with the right rail to find which weekdays carry the volume, then the strip underneath for the hours. The individual cells matter later; the two totals give you the shape of your market in ten seconds.
3. Find the gap Look for the pale and grey blocks inside working hours. A quiet slot next to a crowded one is the cheapest experiment available to you, and it is the one thing Best time to send will never suggest, because nobody in your set has used it.
4. Take the recommendations badge first, number second Read the confidence badge before the day and hour. Discard every Thin sample tile as a timing recommendation, note the medium ones as hypotheses, and treat two brands agreeing on the same slot as worth more than one high confidence badge on its own.
5. Filter to one brand to separate the brand from the crowd Pick a single brand in the dropdown and watch the heatmap redraw. If its dark cells sit where the portfolio's dark cells were, it is following the category. If they move, that brand has made a deliberate choice worth studying.
6. Turn one slot into a test Choose one hypothesis, either the crowded slot or the gap next to it, and split your next two or three sends between it and your current time. Your own ESP measures the result; this screen only tells you where to point the test.

Limits and caveats

• Every open rate and click rate on this screen is an AI estimate derived from the content of the email. There is no ESP data and no tracking pixel behind them, and no competitor's real engagement is visible to Newsletrix.
• The time on a newsletter is the one its sender stamped on it. Nothing is re-based into your timezone or the recipient's, so a brand whose mail system works in its own zone shifts its whole row of the heatmap.
• Only analyzed newsletters count. One that has not been analyzed yet, or one that arrived without a usable send date, silently shrinks every cell on the page.
• Best time to send can only rank slots a brand has already used. It answers which existing slot correlates with the highest estimated open rate, never which untested slot would work best.
• Sample sizes count sends, not recipients. A brand sending twice a week reaches high confidence in the same slot far faster than a monthly sender, without being any more representative.
• The recommendation tiles and the brand table group by the name written on each email, while the Brand dropdown lists the brands you track, so inconsistent naming can split one sender across two tiles. A brand analyzed in the last ten minutes or so may not be in the dropdown yet, even though the charts already count it.
• Heatmap shading is relative to the current peak and the brand table stops at 15 rows without saying so, so neither is safe to compare across two different filter settings.

FAQ

Related screens