Lessons · Lesson 3 of 3
Competitors, and the shops you cannot see inside
Compare a rival's range and not their products, work out what a price in their window can and cannot prove, and turn all three lessons into a buy, a drop and a test.
Lesson 3 of 3 · 36 min
The situation
A rival's price in a window is the easiest number in retail to react to. It is also one of the easiest to misread. It tells you what that shop will take for that garment. It does not tell you whether it is the same garment. Weigh the thing and read its label, and the two prices may describe different amounts of wool. This lesson turns shop-visiting into a method, and draws the line around what it can prove.
12 June. Two of Halgarth's buyers spend a Tuesday walking six shops in one city, four of them Skelmore and two Draycott. They come back with photographs, prices and three garments in a bag. Skelmore has 190 shops against Halgarth's 41. Draycott has 60 and sits above Halgarth on price.
One line in the report is going to decide something. Skelmore is selling a brown half-zip funnel neck at GBP 42.00. Halgarth's is planned at GBP 55.00.
The room's first instinct is to come down to GBP 45.00 or GBP 47.00. The bag on the table is what stops it.
Comparative shopping is a method, not an outing
An outing produces impressions. A method produces a file you can compare with the same file from four weeks ago, and the difference between the two is the whole value.
Halgarth's method is four visits a season: the same six shops, the same weekday, four weeks apart, plus the same three websites on the same morning. For every knitwear option, seven things are recorded. The price. The fibre composition off the label. The sizes actually on the rail. The shape. Whether it is on promotion, and how deep. The country of origin. The date. Once a season, one garment from each competitor is bought and taken apart.
That costs two buyer-days a month, charged at GBP 245.00 a day, plus GBP 1,400.00 a year of purchases. The total is GBP 7,280.00, against GBP 18,400.00 for the forecasting subscription. It is the cheapest information in lesson 2's table, and the only one that describes your own market rather than the market in general.
What a price in a rival's window actually tells you
It tells you what they are willing to sell that garment for. It does not tell you what the garment is, and until you know that, the price cannot be compared with anything.
Halgarth bought the GBP 42.00 half-zip, read the label and weighed it. It is 40% lambswool and 60% acrylic, 340 g in size 12, knitted at 12-gauge. Gauge is the number of needles per inch on the knitting machine, so a higher gauge means a finer, lighter fabric. Halgarth's HG-5120 is 100% lambswool, 415 g in size 12, at 7-gauge. Then the buyer asked her own supplier to price the Skelmore specification, which is the part most retailers skip.
| HG-5120 | The GBP 42.00 competitor garment | |
|---|---|---|
| Composition | 100% lambswool | 40% lambswool, 60% acrylic |
| Weight in size 12 | 415 g | 340 g |
| Yarn price | GBP 11.20 a kg | GBP 5.86 a kg blended |
| Yarn including 12% knitting waste | GBP 5.21 | GBP 2.23 |
| Knitting, linking and making | GBP 5.10 | GBP 3.90 |
| Zip, labels and packing | GBP 1.95 | GBP 1.30 |
| Factory overhead and margin | GBP 2.20 | GBP 1.55 |
| Free on board, or FOB — the price with the goods loaded at the port | GBP 14.46 | GBP 8.98 |
| Freight, duty and handling | GBP 1.94 | GBP 1.22 |
| Landed cost | GBP 16.40 | GBP 10.20 |
| Retail | GBP 55.00 | GBP 42.00 |
| Intake margin | 70.2% | 75.7% |
The competitor's column is an estimate, and it matters to say what kind. The yarn prices are Halgarth's supplier's. The making costs are what that supplier would charge for that specification. And Skelmore, buying for 190 shops, very probably buys better than that. So this column is not a claim about Skelmore's cost, and you should never present it as one.
What it does prove is the only thing you need. The two garments are not the same garment. The competitor is not undercutting Halgarth. It is selling a lighter, cheaper construction at a higher intake margin than Halgarth takes on wool.
That turns the price match into arithmetic. Selling HG-5120 at GBP 42.00 gives an intake margin of 61.0% against a category target of 67.0%, and gives away GBP 52,000.00 across the 4,000-unit buy. And it gives it away to a customer who, standing in front of both garments, is choosing between 415 g of lambswool and 340 g of a blend. Matching a price you have not costed is not competing. It is a donation.
The mistake nobody made
Last autumn Halgarth did exactly the thing this lesson is warning against, and everybody involved was doing their job well.
Skelmore's opening price in cotton knitwear was GBP 25.00. Halgarth put HG-4403 in at GBP 25.00 to hold the same entry price. It became the biggest unit seller in the range, with 6,800 units bought and 5,372 sold, and it was signed off in the post-season pack as a success. It was the correct action, taken for a real reason, with a bad outcome.
| HG-4403 at GBP 25.00 | HG-4471 at GBP 45.00 | |
|---|---|---|
| Units bought | 6,800 | 4,200 |
| Share of the category's units | 26.3% | 16.2% |
| Intake margin | 64.4% | 68.4% |
| Cash margin for the season | GBP 56,355.00 | GBP 116,411.25 |
| Share of the category's cash margin | 11.7% | 24.1% |
| Cash margin per unit bought | GBP 8.29 | GBP 27.72 |
| Residual units at the end of the season | 1,428 | 159 |
26.3% of the category's units returned 11.7% of its cash margin. It carried the only intake margin in the range below the 67.0% target, and it left 41.0% of the whole category's leftover stock. Skelmore can hold GBP 25.00 because it buys around 40,000 units of its entry crew against Halgarth's 6,800, and because 190 shops absorb an opening price differently from 41. The price was matchable. The economics behind it were not, and nobody checked which of the two they were copying.
That is also, exactly, the second half of the two complaints in lesson 1. The range was bottom-heavy because its bottom was set by somebody else's cost base, and the money that should have bought the empty band at GBP 55.00 was already spent at GBP 25.00.
Compare ranges, not products
A single price comparison is nearly always a trap, because you can only see the products they chose to put in front of you. The thing that carries real information is the whole range.
| Price band | Halgarth | Skelmore | Draycott |
|---|---|---|---|
| Under GBP 25.00 | 0 | 4 | 0 |
| GBP 25.00 to GBP 34.00 | 1 | 5 | 0 |
| GBP 35.00 to GBP 45.00 | 4 | 6 | 3 |
| GBP 46.00 to GBP 68.00 | 0 | 6 | 5 |
| GBP 69.00 to GBP 89.00 | 3 | 1 | 6 |
| Over GBP 89.00 | 0 | 0 | 4 |
| Total listed | 8 | 22 | 18 |
The empty band that lesson 1 found from the inside is visible here from the outside, and it is where both competitors concentrate: 6 options for Skelmore and 5 for Draycott, against nought. Two retailers with different strategies, agreeing on where the money is, is a stronger signal than either one alone.
Now the correction that most comparisons never make. Count bodies, then count colours. A body is one shape; a colourway is that shape in one colour. Of Skelmore's 22 listings, five are one roll-neck body in five colours and four are one crew body in four colours. So 22 listings are 15 bodies. Halgarth lists an option once and holds its colours underneath it, so its 8 options are 8 bodies in 24 colourways. Draycott's 18 listings are 17 bodies.
| Halgarth | Skelmore | Draycott | |
|---|---|---|---|
| Bodies | 8 | 15 | 17 |
| Colourways | 24 | 22 | 18 |
| Colours per body | 3.00 | 1.47 | 1.06 |
Counted raw, Skelmore's range is nearly three times Halgarth's. Counted properly, it is twice as broad in shape and less than half as deep in colour. Those are two different businesses, and the comparison turns around once you count the same thing on both sides. If you had bought against "they have 22 and we have 8", you would have added colourways to bodies you already own and left the shape gap exactly where it was.
The shops you cannot see inside
Sort everything the June report contains into three boxes before anybody argues from it.
| Reliable | Suggestive, and easy to over-read | Not observable at all |
|---|---|---|
| Price on the day | A "sold out" badge — you cannot see whether it was 40 units or 4,000 | Their cost, at any level |
| Fibre composition and weight | A shrinking size range across two visits | Their sell-through |
| Shape, gauge and construction | Review counts, whose conversion to units you do not know and which move | Their markdown plan and its budget |
| Sizes physically on the rail | Position in the shop, which may be a landlord's decision | Their intake and achieved margin |
| Promotion depth on the day | An option reappearing, which may be a repeat or a re-allocation | Their forward commitments |
| Option and body counts | Whether they are pleased with any of it |
The middle column is where careers are spent arguing. The closest honest guess at a rival's sell-through is to look twice: the same six shops, four weeks apart, recording which sizes are on the rail. If an option has lost sizes 12 and 14 on the second visit and never gets them back, either it sold or it was bought thin, and you cannot tell which. Write down both readings and act on neither on its own.
The line you must not cross
Everything above is observation of things published to the public, and it is entirely proper. Ringing a buyer at Skelmore to ask what they plan to charge in September is not, and the difference is not a matter of manners. It is competition law.
A supplier is the usual route by which this goes wrong, and the practical version is simple. A supplier who tells you what your competitor ordered is a supplier who is telling your competitor what you ordered. The value of that conversation is negative in both directions.
What to buy more of, what to drop, what to test
All three lessons converge here. Every line in this table names the evidence, and where the evidence is thin it says so.
| Option | Decision | Evidence |
|---|---|---|
| HG-4471 lambswool crew | Hold at 4,300 units, do not deepen | Clean-week rate of sale 3.1 against 2.6 for the band; 19.4% of its full-price units were overflow from the v-neck |
| HG-4472 lambswool v-neck | Same units, size curve corrected upward in 12 and 14 | Implied demand 2,858 against a buy of 2,600; ran out in week 11 |
| HG-4403 cotton crew | Cut from 6,800 to 3,800 units | 26.3% of units for 11.7% of cash margin; the only option under the 67.0% intake target; the price was matched to a cost base Halgarth does not have |
| HG-4486 wool-blend cable | Drop | 47.0% full-price sell-through; its band is better served by a new option at GBP 55.00 |
| HG-4494 merino roll neck | Hold, and treat its 6.8% return rate as no information | Three returns on 44 online units is not a measurement |
| HG-5120 brown half-zip | Test: 1,800 committed, 2,200 reserved | An empty band at GBP 46.00 to GBP 68.00 where two competitors hold 11 options between them, and a spring brown at 88.0% against neighbours at 71.0%, 64.0% and 59.0% |
What else goes into the empty band, whether it needs one option or three, and how the range balances around it are course 16.2's questions. This course hands 16.2 the input and the evidence for it. How the reserve on HG-5120 is negotiated, and whether GBP 0.85 a unit is a fair price for held yarn, belongs to 16.4.
A test that cannot fail is not a test
HG-5120 is described as a test. Most things called tests in retail are not tests, because nobody wrote down in advance what a failure would look like.
So it is written down in June, while nobody yet has a reputation attached to it. The buy needs 4.2 full-price units per selling point per week across the first ten weeks to reach 4,000 over the season. Allowing for a slower start, the gate on weeks 2 to 4 is 2.9. Week 1 is left out, because a new option sells to curiosity. Below 2.9, the reserve is not called off, whatever the room feels in October. How the 4.2 is spread across the season is course 17's work, and the figure comes from the merchandiser. What belongs to this lesson is that the number existed before the emotion did.
That is the discipline the whole course comes down to. Lesson 1: a reassuring number needs positive evidence, and sell-through on its own is not it. Lesson 2: a trend is a quantity and a date, and the right to change your mind has a price you can read off a quotation. Lesson 3: a competitor's price is evidence about a garment you have not weighed. In all three the mistake is the same one. Accepting a comfortable reading because nothing visible contradicted it.
Prompt · Turn a comparative shop into a range comparison and a costed teardown
The afternoon a comp shop comes back and somebody wants to match a price before anybody has read a label.
Act as a retail buyer who has been told to justify a price match and cannot. Comparative shop data: [FOR EVERY COMPETITOR OPTION — RETAILER, DESCRIPTION, PRICE, FIBRE COMPOSITION FROM THE LABEL, WEIGHT IF WEIGHED, GAUGE OR CONSTRUCTION, COUNTRY OF ORIGIN, SIZES SEEN ON THE RAIL, PROMOTION STATE, DATE AND SHOP]. My own range in the same category: [OPTION, DESCRIPTION, PRICE, LANDED COST, COMPOSITION, WEIGHT]. My supplier's prices for the inputs: [YARN OR FABRIC PRICE PER KILOGRAM OR METRE, MAKING COST, TRIMS, WASTE PERCENTAGE, FREIGHT AND DUTY]. My category intake margin target: [PERCENT]. Number of shops mine and theirs: [NUMBERS]. Do the following. First, build the range comparison by PRICE BAND rather than in total, and count bodies separately from colourways — fold multiple colour listings of one shape back into one body and say what that does to the headline count. Second, name every price band where I have nothing and they have something, and every band where I am deeper than both of them. Third, take the one competitor garment I care about and cost MY specification and THEIR specification side by side using my supplier's prices, and state clearly that the second column is an estimate of what that specification would cost ME, not a claim about their cost. Fourth, tell me what intake margin I would take if I matched their price with my specification, and what that gives away in cash across my proposed buy. Fifth, sort everything in the report into what is reliable, what only suggests, and what is not observable at all, and put a warning on anything in the middle box that somebody in the meeting is about to treat as a fact. Sixth, give me the range action: which options to hold, which to cut, and which single gap to test, with the evidence for each. Do not recommend matching a price on the strength of the price alone.
AI can make mistakes — check anything you act on.
Check yourselfA competitor is selling what looks like your garment for 20% less. You have thirty minutes before the range meeting. What do you do with them?Show the answer
Buy it, weigh it and read the label. If there is time, ask your own supplier what your specification would cost at their weight and composition. Thirty minutes is enough to settle whether it is the same garment, and that is the only question that has to be answered before the meeting. If it is lighter, blended, knitted at a finer gauge or made somewhere else, the two prices cannot be compared and the meeting has nothing to decide. If it genuinely is the same garment at 20% less, you have learned something much more serious. Somebody buys better than you do, or takes a lower margin on purpose. Neither of those is fixed by repricing one option this afternoon.
Check yourselfYour comparative shop shows a rival with 22 knitwear options against your 8. Your director wants the range widened. What do you check first?Show the answer
Whether 22 is 22 bodies. Websites list each colourway as a separate item, so a rival's 22 listings were 15 shapes once the repeated colours were folded back. Your own 8 options carry 24 colourways underneath them. Counted the same way, they are twice as broad in shape and less than half as deep in colour. That is a completely different instruction: add shapes, not colours. Then check where the extra shapes sit by price band rather than in total, because a range that is wide in a band you already own well adds cost and reaches nobody new. In this range the gap was a single empty band between GBP 46.00 and GBP 68.00 that both competitors were trading heavily.