The AUR U Role-Fit Framework for Category Management

Let's Start With Something Familiar

Imagine you manage the Produce department. Sales are up 1.4% on last year and your boss is happy. But look underneath that number: units are down 4.2%, and each unit brings in 5.8% more than last year. Shoppers are buying less produce and paying more for what they do buy.

For a department shoppers choose the store for, that isn't a win. It's an early warning.

"Sales is the score. AUR and units show how the game is being played."

This framework does three things. It splits sales into its two drivers, Average Unit Revenue (AUR) and Units. It reads the direction a category moves on that graph as a shopper behaviour. And it judges that behaviour against the role the category plays in the store, so the same movement can be a success for one category and an emergency for another.

The two measures

AUR (Average Unit Revenue) = Net sales ÷ Units sold
(1 + Sales %) = (1 + Units %) × (1 + AUR %)

AUR = what the shopper actually paid per unit, after promotions and markdowns
Units = volume sold (equivalised for pack size where needed)
% change = compared with the same period last year

Where a category sits on the quadrant shows its position. The way it moves shows what shoppers are doing. Its role tells you how to respond.

The graph

1. The Quadrant Graph

AUR (Average Unit Revenue) = Net sales ÷ Units
(1 + Sales %) = (1 + Units %) × (1 + AUR %)

Growth: shoppers buy more and pay more per unit.
Premiumization: fewer units, each earning more.
Value Play: more units, each earning less.
Decline: fewer units, each earning less.
Sales-flat line: above and to the right of it, sales are growing. An arrow that crosses it has flipped the category between growth and decline.

Every % change compares the same period last year. Put the AUR line at category inflation and the Units line at market unit growth (0% if you have no market data), so the graph shows performance relative to the market.

A single point on the graph is a position. The arrow from an older window to a newer one is a movement. This framework makes its decisions on the movement, because the movement is what shoppers are doing right now.

Reading the arrow

2. Eight Movements, Eight Shopper Behaviours

Measure an arrow's direction from its tail, wherever the tail sits on the graph, and put it in one of eight 45° sectors. Each direction is a distinct shopper behaviour. If both axes moved less than the category's normal week-to-week noise, the arrow is Holding and needs no play.

Borderline arrows. The 45° sectors are a convention, not a law of shopper behaviour. If an arrow is within 10° of a sector edge, read both neighbouring plays and let the driver decomposition in Gate 2 decide between them. Never let a 1° difference choose the action.

Defining the arrow

3. Timeframes: Which Points Make the Arrow

The original theory drew one arrow, from the quarter average to this week. That arrow is still here, but it's one of three. Each arrow is tied to the meeting that acts on it, so each decision is made at the right speed.

Arrow	Tail → head	Used in	Answers	Levers it can trigger
Pulse	prior 13 wk → this week	Weekly trading meeting	Did something break this week?	Availability, price errors, order changes, responding to competitor moves
Tactical	13 wk before → last 4 wk	Monthly category review	Which way are shoppers moving now? The playbook reads this arrow.	Price, promotion, targeted offers, display
Strategic	52 wk before → last 13 wk	Quarterly business review	Is the category's position shifting for good?	Range, space, private label, supplier terms, role

The windows never overlap. The tail always ends where the head begins. If the last 4 weeks sat inside the 13-week tail, part of the arrow would compare a period with itself, and every real movement would look about a third shorter than it is.

Why the playbook uses 4 weeks, not 1

A single week swings with a front-page ad, a holiday shift or a heat wave. A 4-week window evens out one promotion cycle but is still recent enough to act on. Use the weekly Pulse arrow for alerts; don't re-plan a category because of one week.

How many weeks make a signal

1 week in a new direction: noise, unless the arrow is exceptionally long.
2 weeks in the same direction: a signal. Check the driver.
3–4 weeks: a trend. It shows on the Tactical arrow, so run the play.
Confirmed on the Strategic arrow: structural. Plan it into the next reset.

Reading the Tactical and Strategic arrows together

	Tactical toward the role's target	Tactical away from target
Strategic toward target	Sustain Both are healthy. Don't intervene; protect what's working.	Early warning A new problem on a good base. Find the driver now; a fast lever usually fixes it.
Strategic away from target	Recovery Earlier actions are working. Hold course, don't add new actions, and judge again next quarter.	Entrenched Fast lever now and a structural fix at the next reset. Put a role review on the agenda.

Seasonal categories use different windows

Fixed quarters cut through seasons. For Seasonal categories, the tail is season-to-date vs. the same window last year and the head is the last 2 weeks. The Strategic arrow compares whole seasons year on year. Outside the window, don't score the category.

Calendar rules

Use the retail calendar (4-5-4 or 13 periods) so weeks line up year on year.
Align moving holidays (Easter, the week of Thanksgiving) before comparing, and handle 53-week years.
Compare like-for-like stores only, excluding new and closed stores.
Set a minimum-volume floor. Tiny categories swing too much to plot.
An arrow is long when it moves more than about 2× the category's typical 4-week move. Calibrate per category, since produce moves more than canned goods.

Intent

4. Category Roles Set What "Good" Looks Like

The same movement means different things for different roles. Each role has a job in the assortment, a target zone on the graph, and a direction it most fears.

Role	Its job in the assortment	Target zone	Good directions	Feared directions
Destination ~5–10% of categories	Win the trip. Shoppers choose the store for these.	Growth; Value Play above the sales-flat line	E, NE, SE	NW, W, SW
Routine ~55–60%	Steady sales and margin; never a reason to shop elsewhere	Growth, or close to where the lines cross	NE, N, E	SW, W
Seasonal ~15–20%	Capture the moment; add excitement	Growth or Premiumization, during the season	NE, N, E	SW, W, SE (early in the season)
Convenience ~15–20%	Complete the basket at a healthy margin	Growth or Premiumization	N, NE	SE, S

Who sets the role

A cross-functional group (merchandising, pricing, finance and shopper insights) sets roles once a year, not the category manager whose results the role will judge. A role needs evidence: penetration, purchase frequency, the retailer's share against its average, and shopper research on reasons for choosing the store. Changing a role mid-year needs the same evidence. That stops a role being relabelled to excuse a bad quarter.

Assign roles at category level, not bucket level

A bucket like Dairy holds milk and eggs, which behave like Destination items, alongside yogurt, which is Routine. Averaging them hides both. Assign roles and run the playbook at category or subcategory level, and use broad store buckets only as a summary. The role shares per role are a widely used convention, not a target to hit.

Before any play

5. Four Gates the Arrow Must Pass

The quadrant measures revenue per unit and units. It doesn't directly measure price, profit, or whether a move is big enough to matter. A category must pass these four gates before it gets a play. A category that fails a gate gets fixed data or a different diagnosis, not an action.

Gate 1Clean the axes

Use equivalised units (per ounce, per count, per serving) wherever pack sizes change. Otherwise shrinkflation reads as Price pushback and multipacks read as Trading down. AUR is measured net of promotions. Deflate it with the retailer's own like-for-like price index, not a national CPI.

Gate 2Decompose AUR

Split the AUR move into like-for-like price + promotion + mix. Direction alone isn't a diagnosis: a northward move led by mix is shoppers trading up, while one led by shelf price is the store taking price. They need opposite plays.

Gate 3Check profit

Look at gross margin $ net of vendor funding, and at margin rate. If margin dollars move against sales (for example, a Growth arrow with shrinking margin because costs rose faster than AUR), flag a revenue–profit split and raise urgency one tier. Sales growth that loses money doesn't count as Aligned.

Gate 4Benchmark and size it

Compare with market data, or at minimum with total-store like-for-like units, so a category isn't blamed for falling along with the whole store. Then size it: $ at risk = annualised sales gap vs. benchmark. Below the materiality floor, the category is logged and not actioned.

Store clusters. A chain-level arrow can average away two opposite stories. When store clusters (by format, region or competitor set) point in different directions, plot the arrow per cluster and play each one separately.

If X and Y, do Z

6. The Movement Playbook

Pick the category's role, then click the direction its Tactical arrow points. Every one of the 32 combinations is also listed in the tables below.

The tail is drawn at the centre for reading direction. On the real graph it sits wherever the category's 13-week point is.

When to act

7. Urgency and Timing

Every play has a base urgency tier. Adjust it using the escalators below, then match the action to a lever that can take effect within that window. Fix the problem in two moves: stabilise with a fast lever now, then fix the cause with a slow lever at the next reset.

Act nowwithin 1–2 weeks

Price corrections, availability fixes, order and allocation changes, pulling or adding a promotion that's already set up.

This periodwithin 4–6 weeks

The next ad and promotion cycle, targeted loyalty offers, secondary displays, prices on less price-sensitive items.

Next reset3–6 months

Range, space and planogram, private label, pack sizes, supplier terms, next season's buy.

Monitornext routine review

Sustain what works. Re-check the Tactical arrow next month.

Raise urgency one tier when

The category is a Destination. Lost trips cost the whole basket.
The arrow crosses the sales-flat line downward.
The arrow is long (more than ~2× the category's typical 4-week move).
The Pulse arrow has pointed the same bad way for 3+ weeks.
The Tactical and Strategic arrows are both moving away from target (Entrenched).
A Seasonal category is inside its selling window, where every week is a large share of the year.
Gate 3 flagged a revenue–profit split.

Lower urgency one tier when

It's a single-week move with no confirmation.
A known one-off explains it (holiday shift, weather, a supply issue already fixed) and it has already reversed.
The Strategic arrow shows Recovery: earlier actions are still working through.
It's a low-share Convenience category with margin dollars holding.

Judge each lever on the timeframe it works in

Lever	Time to execute	Judge the result on	Common mistake
Availability / out-of-stocks	Days	Pulse (weekly)	Blaming price for a units drop that is really empty shelves
Shelf price	1–2 weeks	Tactical (4 weeks)	Reversing a price move after one week, before shoppers have adjusted
Promotion	4–12 weeks (ad calendar lock)	The promotion weeks plus 4 weeks after	Counting stock-up volume as a gain without netting the dip afterwards
Display / space tweak	1–4 weeks	Tactical	Taking the space from another category's target zone
Range / planogram reset	3–6 months	Strategic (13 weeks after the reset)	Judging a reset on its first month
Supplier terms	Quarterly to annual	Strategic	Improving margin while the price index slips
Role change	Annual plan	52 weeks	Changing a role to excuse a bad quarter

Priority order for the action queue: rank by urgency tier, then by $ at risk within each tier, with Destination breaking ties. Cap the Act-now list at what the team can actually carry out in a week (start with five). Anything beyond the cap is escalated to leadership, not quietly dropped. An action queue where everything is urgent has no priorities.

When to review a role: the Strategic arrow has pointed at a feared direction for 2+ quarters despite action; penetration or the retailer's share of the category has fallen for 2+ quarters; a competitor has changed the market structurally; or the store's strategy has changed. The category may be doing fine for a role it no longer has.

How to run it

8. The Routine, by Cadence

Weekly: check the Pulse. Flag categories whose Pulse arrow is long, crosses the sales-flat line, or points in a feared direction for their role. Fix availability and price errors the same week.
Monthly: pass the four gates. Clean the axes, break down AUR, check profit, and benchmark and size the move. Only categories that pass go on to a play.
Read the Tactical arrow. Classify its direction (both neighbouring plays if it's borderline), confirm the role, and look up the play.
Set urgency and rank. Start from the play's base tier and apply the escalators. Rank by $ at risk, within the weekly Act-now cap.
Test before scaling. Pilot This-period and Next-reset actions in a store group against a matched control group before rolling them out. Act-now fixes (availability, price errors, emergency price corrections) go chain-wide immediately.
Log the decision. Record the arrow, the gate results, the play chosen, the target position and the expected result. Next period, score the play against its target.
Quarterly: read the Strategic arrow and grade the playbook. Combine it with the Tactical arrow, feed structural fixes into the reset brief, review roles against the triggers, and check each play's hit rate from the decision log.

Worked example

9. Case Study: One Monthly Review at Northfield Market

Illustrative only. Northfield Market is a fictional 42-store regional grocer. Every figure is invented to show how the framework works. None of them are real results or benchmarks.

It's the Week 36 category review. The Pulse flagged three categories in the weekly meetings. Each one shows a different part of the framework: Fresh Produce is a real alarm, Snacks is a false alarm caught by a gate, and Health & Beauty is a profit leak settled by a pilot.

Both axes are shown relative to their benchmark lines: units against the market, and AUR after category inflation. That lets three categories share one graph. The hollow dot is the tail (the 13 weeks before) and the solid dot is the head (the last 4 weeks).

Category	Tail	Head
Fresh Produce	+1.5, +0.5	−3.4, +2.9
Snacks (raw)	+0.2, +0.3	−5.6, +5.8
Snacks (equivalised)	+0.2, +0.3	−0.4, +0.2
Health & Beauty	+0.4, −0.3	+5.2, −4.0

Coordinates are (units vs. benchmark, real AUR), in percentage points.

Fresh Produce

DestinationAct now

Pulse (weekly)	Pointed up and to the left for 3 straight weeks, and crossed the sales-flat line in Week 35.
Gate 1: Clean the axes	Produce is sold by weight and count, so units are already equivalised. Category inflation is 2.9%, so nominal AUR +5.8% becomes real AUR +2.9.
Gate 2: Decompose AUR	+5.8% = like-for-like price +4.6 + promotion +1.4 (fewer deals) + mix −0.2. The move is price-led. The berry and avocado cost increases were passed through in full. The price index on the top 14 key value items (KVIs) went from 101 to 106 against the main competitor, which held its prices.
Gate 3: Check profit	Nominal sales +1.4% and margin $ +0.8%. The category's own profit looks fine, which is why nobody noticed. After inflation and against the market, sales are down 0.5 pts, and a Destination category's job is trips, not category margin.
Gate 4: Benchmark and size	Market units −0.8% against Northfield −4.2%, a gap of 3.4 pts. That's about $4.3M a year of sales at risk (on roughly $126M of annual produce sales). Loyalty data shows trips by produce-buying households down 2.1%. The loss is spreading to the whole basket.
Direction	The arrow sits 26° above due west, within 10° of the line between two directions, so it's borderline. Read both plays: Buying less (W) and Price pushback (NW). Gate 2 shows the move is price-led, so NW, Price pushback is the one that applies.
Two arrows	The Strategic arrow is still heading toward the target zone, while the Tactical arrow heads away: Early warning. That points to a fast lever.
Play and urgency	Destination × NW is Act now at base. The escalators (Destination, crossing the sales-flat line, 3 weeks in a row) confirm it. It ranks #1 in the queue by $ at risk.
Action	Chain-wide price correction, with no pilot needed because this is an Act-now correction: the 14 KVIs roll back to a price index of 100–101. The margin is recovered by +2–3% on 40 less price-watched items (fresh herbs, exotic fruit) and by reopening berry cost negotiations. Logged target: units back at or above the benchmark within 8 weeks.
Result (8 weeks)	The head moved to +0.6, +0.8, back in Growth. Margin $ −0.4% vs. last year: some was given back, but trips by produce households recovered to −0.3%. Status: Recovery. It gets re-scored at the next quarterly review.

Snacks, Candy & Cookies

RoutineNo action

Pulse (weekly)	A long arrow up and to the left: units −6.1%, AUR +7.4%. It looked like severe Price pushback in an everyday category.
Gate 1: Clean the axes	Two leading chip brands cut bag sizes by about 10% at the same shelf price. Measured per ounce, units −0.4 pts and real AUR +0.2 pts. The arrow is inside the noise band: Holding.
Result	No play. What was avoided: the draft response was a price rollback on 60 items, worth about $0.9M a year in margin, to fix a volume loss that never happened. Logged as "false alarm stopped at Gate 1", and pack-size changes are now flagged automatically.

Health & Beauty Care

ConvenienceAct now+ pilot

Direction	Down and to the right: SE, Deal-driven. Units +5.2 pts, real AUR −4.0 pts. On the sales line alone it looked like a win (sales +1.0%).
Gate 2: Decompose AUR	Promotion-led. A supplier-funded buy-one-get-one (BOGO) offer on shampoo and body wash drove 80% of the AUR decline.
Gate 3: Check profit	Net of vendor funding, margin $ −7.5% while sales +1.0%. That's a revenue–profit split. Shoppers don't choose Northfield for health and beauty on price, so most of the discounted units would have sold at full price anyway.
Play and urgency	Convenience × SE is This period at base. The revenue–profit split escalator raises it to Act now. It ranks #2 in the queue.
Action	Act now: cancel the chain-wide BOGO repeat planned for the next ad cycle. Pilot: the replacement, regular price plus targeted digital coupons for lapsed buyers, in 12 stores against 12 matched control stores for 6 weeks.
Result (6 weeks)	Pilot stores: units −2.8% and margin $ +9.1% against control. The play worked, so it rolls out chain-wide. The result is measured against a control, not just an arrow that moved.

The Week 36 action queue

Rank	Category	Role	Play	Tier	$ at risk / yr
1	Fresh Produce	Destination	NW, Price pushback	Act now	$4.3M
2	Health & Beauty Care	Convenience	SE, Deal-driven (+ revenue–profit split)	Act now	$1.6M margin
3	Frozen Foods	Routine	W, Buying less (supplier shortage, out-of-stocks)	Act now	$1.1M
4	Breakfast & Cereal	Routine	S, Trading down (private label mix)	This period	$0.4M
—	Snacks	Routine	Stopped at Gate 1: Holding	No action	—
—	Floral	Seasonal	Outside its selling window: not scored	No action	—

Three Act-now items, within the weekly cap of five. Breakfast waits for the next ad cycle.

What the case shows

The role changes what the numbers mean. Produce's category margin was up, and that would have been a win for a Convenience category. For a Destination category it hid a loss of shopper trips.
The gates earn their keep. Gate 1 stopped a $0.9M mistake in Snacks. Gate 3 found a profit leak in Health & Beauty that the sales line hid.
The borderline rule worked. Produce's arrow could have been read as Buying less, which would have led to an out-of-stock audit. The price/promotion/mix split sent it to the right play.
Fast fix, then measured proof. The emergency price correction went chain-wide right away; the new promotion approach was proven against control stores before rollout.

Before rollout

10. Calibration: Prove It on History First

Every threshold on this page is a starting default: the 4-week head, the 2× long-arrow rule, the 10° borderline band, the urgency tiers and the Act-now cap. None of them is validated yet. Prove them before anyone makes a pricing decision with them.

Step	What to do	Pass mark
Back-test	Replay 2–3 years of weekly data through the framework. For each Act-now and This-period call, check what happened to sales and margin over the next 13 weeks.	Calls come before real deterioration more often than they raise false alarms
Tune per category	Set the noise band, the long-arrow multiple and the signal-week count from each category's own volatility.	No category produces an alert most weeks
Shadow run	Run the framework alongside the current process for one quarter, in 2–3 categories per role, without acting on it.	Category managers agree the calls are sensible, or can say why they aren't
Controlled pilot	Act on the plays in pilot stores against matched control stores.	Pilot categories beat control on sales and margin $
Ongoing grading	Track the hit rate of each role × direction play from the decision log.	Rewrite or retire plays that don't beat doing nothing

Read before use

11. Disclaimers and Limitations

What this framework is

Personal work. This framework is my own thinking. It doesn't use, and doesn't represent, the views, data or practices of any employer.
A diagnostic aid, not a predictive model. It organises signals and suggests where to look. It doesn't forecast outcomes or guarantee results.
A draft that hasn't been validated. The thresholds, urgency tiers, role shares and plays are proposals. None has been back-tested on real data yet. See Calibration.
A starting point for judgement. Category managers stay accountable for decisions. Local knowledge, shopper research and commercial context can and should override a play.
Built on illustrative examples. The case study and any example roles are generic, not a description of any particular retailer.

What it does not cover

Causation. Arrow movements show correlation. Only controlled tests show what an action caused.
Profit on its own terms. The graph is revenue-based. Gate 3 adds a margin check, but it isn't a full profitability model (labour, shrink, space cost, activity-based costing).
Basket and halo effects. Cross-category effects, especially for Destination price investment, need basket-level data this version doesn't use.
Data quality. Results are only as good as the inputs. Item master errors, pack-size churn, calendar misalignment and missing market data will all produce wrong calls.
Legal and contractual limits. Pricing and promotion actions must follow pricing and advertising laws, vendor and trade-funding agreements, and competition law. Competitor price checks must use legitimately obtained public information, never coordination with competitors.
Professional advice. Nothing here is financial, legal or investment advice.

The AUR U Role-Fit Framework
for Category Management

Let's Start With Something Familiar

1. The Quadrant Graph

2. Eight Movements, Eight Shopper Behaviours

3. Timeframes: Which Points Make the Arrow

Why the playbook uses 4 weeks, not 1

How many weeks make a signal

Reading the Tactical and Strategic arrows together

Seasonal categories use different windows

Calendar rules

4. Category Roles Set What "Good" Looks Like

Who sets the role

Assign roles at category level, not bucket level

5. Four Gates the Arrow Must Pass

6. The Movement Playbook

7. Urgency and Timing

Raise urgency one tier when

Lower urgency one tier when

Judge each lever on the timeframe it works in

8. The Routine, by Cadence

9. Case Study: One Monthly Review at Northfield Market

Fresh Produce

Snacks, Candy & Cookies

Health & Beauty Care

The Week 36 action queue

What the case shows

10. Calibration: Prove It on History First

11. Disclaimers and Limitations

What this framework is

What it does not cover

Want to Discuss This Framework?

In This Article

Masimba Ruwo

The AUR U Role-Fit Frameworkfor Category Management

Let's Start With Something Familiar

1. The Quadrant Graph

2. Eight Movements, Eight Shopper Behaviours

3. Timeframes: Which Points Make the Arrow

Why the playbook uses 4 weeks, not 1

How many weeks make a signal

Reading the Tactical and Strategic arrows together

Seasonal categories use different windows

Calendar rules

4. Category Roles Set What "Good" Looks Like

Who sets the role

Assign roles at category level, not bucket level

5. Four Gates the Arrow Must Pass

6. The Movement Playbook

7. Urgency and Timing

Raise urgency one tier when

Lower urgency one tier when

Judge each lever on the timeframe it works in

8. The Routine, by Cadence

9. Case Study: One Monthly Review at Northfield Market

Fresh Produce

Snacks, Candy & Cookies

Health & Beauty Care

The Week 36 action queue

What the case shows

10. Calibration: Prove It on History First

11. Disclaimers and Limitations

What this framework is

What it does not cover

Want to Discuss This Framework?

In This Article

Masimba Ruwo

The AUR U Role-Fit Framework
for Category Management