Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

E-commerce Marketing › Intermediate

Segmenting a Catalogue for Shopping Ads at a Bike Parts Store

You split an invented bike parts retailer's catalogue into segments by margin, price and sales history, choose the feed labels that carry them and decide which campaign type each segment belongs in. The exercise turns one store-wide shopping campaign into a structure where every product group is judged against its own break-even line.
ExerciseIntermediate levelModule 1: Shopping, feeds and catalogue ads5 minUpdated

Why this matters

A campaign can only treat products differently if the feed lets it tell them apart. The shopping and catalogue ads guide explains why one store-wide target subsidises thin-margin products with fat-margin ones, and the product feeds guide shows how custom labels carry the facts a campaign needs. Here you decide which facts matter, turn them into labels and pick a campaign type per segment with the Performance Max versus Standard Shopping comparison.

The store and its catalogue

Spokewright Cycle Parts is an invented online retailer selling own-brand consumables alongside branded drivetrain parts, wheels and tyres. Its feed is clean on the basics: identifiers, categories and product types are filled and disapprovals are rare. It has no custom labels, so its single Standard Shopping campaign bids on everything as one product group against one target return on ad spend (ROAS). Conversion tracking records real order values and has been verified, and the store has good packshots and a few workshop videos.

Illustrative example — not a real result

Spokewright Cycle Parts: last quarter by product line

Margin is gross margin after product cost, delivery, payment fees and expected returns.

Product lineProductsTypical priceMarginOrdersShopping spendShopping revenueReported ROAS
Own-brand inner tubes40£862%2,400£1,100£3,9003.5
Own-brand brake pads60£1455%1,900£900£5,2005.8
Own-brand tools and workshop kit45£3248%500£700£3,1004.4
Own-brand lights (launched last month)20£5550%40£300£1,1003.7
Branded tyres150£3830%1,100£2,500£11,0004.4
Branded wheels70£26026%140£2,600£12,4004.8
Branded chains and cassettes180£4522%1,300£4,200£17,6004.2
Branded groupsets25£42014%60£3,000£9,8003.3
Clearance (discontinued models)300£209%200£1,000£3,6003.6
Products with no sales last quarter1,800mixedabout 35%0 from ads£1,600£9000.6

Total Shopping spend £17,900, revenue £68,600, blended ROAS 3.8 against a store-wide target of 4.0. Brake pads lose most eligible impressions to rank, not budget. The owner wants each segment to clear its own floor plus a profit of about 5% of revenue.

Choosing the segments

Start from the decisions the campaigns must make. Margin decides the target, so it is the first label. Sales history decides whether a segment has enough conversions for automated bidding, and separates proven lines from new launches and dead stock. Price tier matters because cheap and expensive items can cost the same per click while earning very different amounts. Stock status isolates clearance. Keep each label to a few documented values, populated from the stock and order systems rather than by hand.

Task

  1. Work out the break-even ROAS (Break-even ROAS) for each product line from its margin, then mark which lines clear it today and which do not.
  2. Define the custom labels you would add, with their values, the rule that sets each value and the system it comes from.
  3. Group the product lines into campaign segments and set a target return for each, using break-even plus the owner's required profit.
  4. Choose Performance Max, Standard Shopping or no advertising for each segment, and give the reason from the comparison page.
  5. Explain how the segments avoid competing for the same products, and what you would read first.

Worked answer

Illustrative example — not a real result

Spokewright: break-even, targets and segments

Break-even is 1 ÷ margin. Target is 1 ÷ (margin − 0.05).

LineBreak-evenReportedClears it?
Inner tubes1.63.5Yes, comfortably
Brake pads1.85.8Yes, and held back
Tools and kit2.14.4Yes
Lights2.03.7Yes, on little data
Tyres3.34.4Yes
Wheels3.84.8Yes
Chains and cassettes4.54.2No
Groupsets7.13.3No, by a wide margin
Clearance11.13.6No
No-sales products2.90.6No

Labels:

  • Margin band: high (45%+), mid (25% to 44%), low (under 25%), from the stock system's cost price.
  • Sales velocity: best-seller, steady, new (listed within 60 days), no sales (none in 90 days), from orders.
  • Price tier: under £15, £15 to £100, over £100.
  • Stock status: core or clearance, from the buying team's discontinued flag.

Segments and targets:

  • High margin (tubes, pads, tools): blended margin about 56%, target about 2.0. Performance Max.
  • New launches (lights): target 2.2, own Standard Shopping campaign with a fixed budget until it has history.
  • Mid margin (tyres, wheels): blended margin about 28%, target about 4.3. Standard Shopping.
  • Low-margin best-sellers (top chains and cassettes): target about 5.9. Standard Shopping; rest of band excluded.
  • Groupsets, clearance and no-sales products: no paid shopping this quarter. Clearance goes to email.

The £5,600 spent last quarter on groupsets, clearance and no-sales products is freed for the high band.

Why Performance Max for the high band. It has the most conversions, the widest room above break-even and strong inputs. Its Impression Share lost to rank shows the store-wide target has been holding it back; a target from its own margin lets it bid for volume it can afford.

Why Standard Shopping for the rest. Mid and low bands have thin room above break-even, so the store needs search term reports and negatives to cut unprofitable queries. The new lights need the same visibility while they build history.

Keeping segments apart. Performance Max takes priority over Standard Shopping on shared products, so its listing groups hold only high-margin core products and each Standard Shopping campaign filters on labels the high band excludes.

The first month. Read profit after ad spend by margin band, impression share on the high band and whether the low-margin best-sellers clear their target. Check label values against the stock system after the first price changes.

Common mistakes

  • Segmenting by category. Chains and tyres are both wear parts, yet their margins need different targets; split by what changes the decision.
  • Rewarding revenue. Chains and cassettes earned the most Shopping revenue and still lost money.
  • Overlapping campaigns. A product in both campaign types serves from Performance Max, so the Standard campaign's bids never run.
  • Hand-keyed labels. Margin bands typed into a spreadsheet go stale at the next price change, and every campaign built on them drifts.

Next, take the feeds and shopping checkpoint or return to the intermediate e-commerce level.

Pages this lesson applies

Terms used on this page

Also related

Related