Promotions: discounts, offers and urgency

What this screen is

Promotions works from every discount mechanic and every pricing phrase Newsletrix found in your analysed newsletters. Two of the four readings describe the cohort as a whole: the mechanic it reaches for most, and the way it frames offers. The other two are per brand, covering how often a brand attaches an offer to a send and how deep that offer goes when a number is stated.

Three comparative views sit on top of that. The price war detector flags calendar weeks in which several tracked brands went deep at the same time. Lining communication tone up against discount depth, the tone matrix shows whether your market escalates copy urgency in step with the size of the cut. Last comes the price positioning matrix, which places each brand on a frequency against depth plane with the peer medians drawn in and sorts the cohort into promo-heavy, always on sale, scarcity-based and premium.

Everything here is read out of the email copy. Newsletrix receives competitor newsletters as an ordinary subscriber, so it never sees a competitor's revenue, redemptions, margin or real price list. A discount the analysis missed leaves that newsletter counting as non-promotional. Wording decides the rest: up to 60% off gives the screen a 60 to work with, while limited time only records the mechanic and no depth at all.

One naming note. The address bar says /pricing and an older release of the screen was titled Pricing Strategy Analysis, but nothing on it concerns your own Newsletrix billing. Your plan and invoices live in the account area.

Open it when

• Before locking a promo calendar, to see which weeks the cohort already crowds with deep discounts.
• When you need to know whether a competitor is an always on sale brand or a premium brand that rarely discounts, before you benchmark against them.
• When writing offer copy and you want evidence on whether deep discounts in your market travel with urgent or with calm tone.
• When a competitor just sent an aggressive offer and you want to check whether it is a one-off or their normal cadence.
• When preparing a quarterly competitive review and you need the ranked list of who leans hardest on promotions.

Screen anatomy

The real screen first, then a numbered map of the current layout. Everything stacks in one scrolling column, and several panels appear only when there is enough data to fill them.

1. Filter card Brand, Date from and Date to. Change any of the three and every panel below reloads. The header and this card stay on screen while the rest of the page redraws, so you can keep adjusting while it loads.
2. Price war detector An alert card listing calendar weeks in which two or more tracked brands each shipped a discount of 25% or deeper. It only appears when at least one week qualifies. The pill in the header shows the deepest discount in the newest flagged week, red once that reaches 50%.
3. Discount patterns (Top 10) A doughnut of the ten most frequent discount mechanics across the filtered set, labelled with the wording Newsletrix read out of the copy.
4. Pricing communication styles (Top 10) A bar chart of how offers are framed rather than how deep they are.
5. Key insights Up to five sentences summarising the panels around it. Every figure in them also appears somewhere else on the screen, so you can trace any line back to the panel it came from.
6. Discount patterns by brand One small card per brand, alphabetically, each listing that brand's five most frequent patterns with an occurrence count.
7. Tone × discount depth A shaded matrix of tone rows against four depth bands, with a per-tone average column and a pill naming the tone that leads on deep discounts.
8. Price positioning matrix One bubble per brand at discount frequency against average discount depth, sized by send volume, with dashed lines at the peer medians and four quadrant labels.
9. Promotional intensity statistics Every brand with its send count, its discount count and a colour-coded frequency badge, ordered by frequency descending.
10. Header and the More information button The screen title, its subtitle, and a question-mark button that opens the Pricing Strategy Insights explainer. The button changes no data.
11. Loading and empty states While the screen loads, a spinner replaces everything under the filter card. If discount patterns, communication styles and promotional intensity are all empty you get a single message: No pricing data available, with the advice to adjust your filters or check back later.

Every panel explained

Price war detector

A stack of week rows, newest first. Each row names a week, how many brands discounted deeply in it, the deepest cut anyone made, and per brand a depth badge, an example subject line, the pattern wording and the number of sends that week.

How it is calculated. A newsletter's depth is the deepest percentage stated anywhere in it. To count here it needs a depth of 25% or more, a send date, and a sender Newsletrix could put a name to. Weeks run Monday to Sunday, and a week is flagged once two or more different brands clear 25% inside it. Each brand appears once per week, represented by its deepest send, which is where that brand's badge, sample subject line and pattern wording come from. Brands are listed deepest first and weeks newest first. The card holds the 12 most recent flagged weeks and shows three of them until you expand it.

How to read it. Each row is a week when the market discounted together. A red row means at least one brand went 50% or deeper that week, which makes it a poor week to launch a 10% offer. Use the Monday dates to plan around the crowd, or to deliberately join it.

Discount patterns (Top 10)

A doughnut of the ten most frequent discount mechanics in the filtered set, labelled with the wording found in the copy.

How it is calculated. Identical wordings count together, and the ten most frequent get a slice. The palette runs to six colours, so a doughnut with more than six slices repeats one. Unlike the brand-based panels, this chart also counts newsletters whose sender Newsletrix could not name.

How to read it. The largest slice is the mechanic the cohort reaches for most often. Compare the relative sizes rather than the raw counts, because counts scale with how many newsletters you have collected.

Pricing communication styles (Top 10)

A bar chart of the ten most frequent ways offers are framed: percentage discount, fixed discount, bundle, free shipping, subscriber exclusive and whatever else the copy offers up.

How it is calculated. Styles are grouped on their exact wording and the ten most frequent are drawn. One newsletter can feed several bars, and the same framing twice over when its copy repeats it, so the bars add up to more than the number of newsletters. Like the doughnut, this chart includes newsletters with no identified brand.

How to read it. This tells you how offers are packaged, not how deep they cut. A tall free shipping bar next to a short percentage bar means the market is competing on delivery, not on price.

Key insights

A numbered list of up to five sentences. They give the share of the top discount pattern, the count of the top communication style, the average discount frequency across brands, how many brands sit above 50% frequency and how many brands are tracked in all.

How it is calculated. From the same numbers the panels around it already show. The top pattern share is that pattern's count divided by all pattern occurrences, to one decimal. Average discount frequency is a plain arithmetic mean over the brands in the intensity table, and the high intensity count is how many of those brands sit above 50% frequency. Brand count is the number of cards in the by-brand grid.

How to read it. As a summary of the four panels around it, nothing more. The percentages are shares of occurrences, not shares of newsletters, and the average frequency is unweighted, so a brand with 2 sends counts as much as one with 200.

Discount patterns by brand

A grid of small cards, one per brand, each listing that brand's five most frequent patterns with an occurrence count on the right.

How it is calculated. Each card holds that brand's five most common patterns, biggest count at the top. The cards run alphabetically by brand name, so their order says nothing about how promotional a brand is. Brands Newsletrix could not put a name to are left out.

How to read it. This is the per-brand promo playbook. A brand whose top patterns are all time-limited and last chance runs urgency; one whose top patterns are numeric percentages runs price. The price positioning matrix averages a brand's whole pattern list, not only the five shown here, so a brand can carry an average depth that none of these five explains.

Tone × discount depth

A matrix with one row per tone that actually occurs, in the fixed order Urgent, Excited, Calm, Friendly, Neutral, against four depth bands: 0-10%, 10-25%, 25-50% and 50%+. Each row shows how many sends it holds and how many of them had a readable discount, plus an average depth column. A pill names the tone that leads on deep discounts.

How it is calculated. Each newsletter sits in exactly one tone row. An emotional appeal of urgency puts it in Urgent. Excitement or playfulness puts it in Excited. Comfort, trust, aspiration, belonging, professional, formal or authoritative reads as Calm. Casual, friendly or conversational reads as Friendly, and anything left over is Neutral. A newsletter that answers to more than one of those takes the row highest in that list. Depth bands are cut at 10, 25 and 50, so below 10 is 0-10%, below 25 is 10-25%, below 50 is 25-50%, and 50 or more is 50%+. A newsletter with no readable depth still counts in that tone's total but lands in no band. The average column is the mean of that tone's readable depths to one decimal, and shows a dash when there are none. The deep-discount lead is the tone with the most newsletters in the 25-50% and 50%+ bands combined. Cell shading is relative to the busiest cell in the same row, with a floor so that any non-empty cell stays visible.

How to read it. Compare the Urgent row with the Calm row. If Urgent concentrates in the two deep bands while Calm sits in 0-10%, the cohort escalates copy urgency in step with the size of the cut, and matching that pattern is a defensible choice. Hover a cell to see up to three example brands.

Price positioning matrix

One bubble per brand plotted at discount frequency on the horizontal axis and average discount depth on the vertical, with bubble size showing send volume. Dashed lines mark the peer medians and split the plane into four quadrants. Scarcity-based sits top left, Promo-heavy top right, Premium bottom left and Always on sale bottom right.

How it is calculated. Frequency comes from the intensity table. Depth is worked out from that brand's full pattern list: each pattern is read for a percentage and the brand's depth is the count-weighted mean, so a pattern seen ten times counts ten times. A percentage range such as 25-40% is averaged to its midpoint, a single percentage is taken as written, and buy one get one or half off is treated as exactly 50%. Currency amounts such as $20 off yield no depth. Brands with no readable depth anywhere are dropped from the chart. The medians are the middle value of the remaining brands, the mean of the two middle values when the count is even, and they are only drawn when at least two brands stated a readable percentage. Bubble size follows send volume, scaled against the busiest brand on the chart. The scale is compressed, so a brand that sends ten times as much does not get a bubble ten times as wide. Each axis runs from zero to the next multiple of ten above the highest value plus another ten, capped at 100.

How to read it. Right of the vertical line means the brand discounts more often than its peers; above the horizontal line means it discounts deeper. Big bubbles send a lot, so their position rests on more evidence. Because the lines are medians, the quadrants move as the cohort changes: a brand can shift quadrant without changing its own behaviour at all.

When fewer than two brands have a readable numeric depth, the card swaps to an amber fallback that ranks brands by frequency alone and breaks down which pattern shapes it found.

Promotional intensity statistics

A table of every brand with an identified name: send count, discount count and a colour-coded frequency badge, ordered by frequency descending with no cut-off.

How it is calculated. Total newsletters counts distinct sends from that brand inside the filter window. With discounts counts the discount phrases attached to those sends, so a newsletter carrying three phrases adds three. Discount frequency is with discounts divided by total newsletters, times 100, to one decimal, which is why it can exceed 100%. The badge turns red above 50, amber above 25 and green otherwise.

How to read it. A high frequency means the brand rarely sends without an offer attached. Read Total newsletters as the confidence column: 100% over three sends means very little, 64% over sixty-four sends is a real pattern.

Metric definitions

Filters and controls

There is deliberately no export, no CSV download, no column sorting and no search box on this screen, and the group label you can attach to a tracking address is not applied here. The explainer behind the More information button lists Price Point Distribution, Urgency Tactics, Competitive Positioning and Seasonal Trends: none of those are panels here. Urgency only shows up indirectly through the Urgent tone row and through wording such as time-limited, and seasonality lives on the Campaigns screen. Return to a filter combination you were looking at five minutes ago and the screen shows the same figures it showed then instead of counting again, and nothing refreshes while the tab sits in the background.

How to use it

1. Set a window wide enough to hold weeks. Leave Brand empty and set a date range covering at least a full quarter. The price war detector works on whole Monday-aligned weeks, so short windows give it nothing to flag.
2. Rank the cohort in the intensity table. Read each frequency badge next to Total newsletters. Anything red over a decent send count is a brand that almost never ships without an offer; anything green over a decent send count is a brand you cannot beat on price signalling alone.
3. Mark the crowded weeks. Expand the price war detector and write the Monday dates into your own calendar. Red weeks are the ones where a modest offer will be invisible, so either move your send or go in with something structurally different, such as a bundle or early access.
4. Place yourself on the positioning matrix. Find the quadrant your own promo behaviour would land in. If you sit in Promo-heavy next to three competitors, discount depth is not a differentiator in your market. Note the two median values, because they are the numbers to beat next quarter.
5. Check the tone matrix before writing copy. If the deep bands belong to Urgent while Calm stays shallow, the market has trained its readers to expect urgency with a real cut, and calm copy on a deep discount will read as a mismatch.
6. Drill into one brand. Select a single brand in the filter, then read its pattern card together with its line in the intensity table. That gives you their mechanic, their framing and their cadence in one pass.

Limits and caveats

• Every number here comes from the AI reading email copy. There is no commerce data, no redemption data and no visibility into a competitor's real prices or margins. A discount the analysis missed makes the newsletter look non-promotional.
• Depth is read out of free text. Currency offers such as $20 off, free shipping, and vague copy produce no percentage at all, so they are invisible to the price war detector, the tone matrix and the positioning matrix while still counting in the two Top 10 charts.
• A discount mechanic never has to carry a number. Most of what turns up is worded, stackable or time-limited and the like, so a depth figure exists only where the brand itself wrote a percentage. Numeric coverage is patchy, and uneven between brands.
• Buy one get one and half off are treated as exactly 50%, and a phrase that mixes mechanics can score one depth on the bubble chart and a different one in the price war detector, on the same page.
• With discounts, and therefore Discount frequency, counts phrases rather than newsletters and can exceed 100%. Key insights compounds this by averaging frequency across brands without weighting for send volume.
• Most panels stop short of showing everything. The price war detector holds 12 weeks, the doughnut 10 slices, the bar chart 10 bars, a brand card 5 patterns, the fallback ranking 12 brands and a matrix cell 3 example brands. Newsletters whose sender could not be identified are excluded from the brand grid, the intensity table, the tone matrix and the price war detector.

FAQ

Related screens