Case studies / BabyCoo
Six lists of products, six gifts, and never a cart total.
BabyCoo hangs each gift on its own list of products, and the lists ladder from a A$22 median item up to a A$145 one. Six campaigns, no cart value anywhere in the setup, and 4,538 orders that carried a gift.
Visit babycoo.com.au ↗
01 / The problem A catalogue too broad to shop by budget
BabyCoo sells to parents across a range that runs from rattle socks under A$6 to magnetic tile sets near A$200. A shopper who came in for a teething toy is nowhere near the thing they need next.
The basket is not small because the shopper will not spend. It is small because nothing on the page connects to what is already in it, and asking for another A$30 only sends them hunting for filler.
Why a spend threshold was the wrong tool here
- Spend thresholds
- A threshold asks for any A$30 at all, so a shopper who is A$30 short goes hunting for filler. Not one of the six campaigns running today reads the cart total. babycoo.com.au
- Price discounts
- Money off a feeding set comes out of margin and still says nothing about which second item belongs beside it. AOV.ai analysis
- One gift for the whole store
- The same gift cannot suit a A$20 order and a A$200 one. It overpays on the first and goes unnoticed on the second. AOV.ai analysis
So if the gift is not attached to a number, what is it attached to?
02 / The campaign The trigger is a list of products, not a number
All six live campaigns work the same way. Buy one item the store has attached to that campaign, and the gift goes into the cart on its own. No threshold, no code, nothing to claim.
Look at what each rule stops from happening. One qualifying item is enough, so the gift rewards choosing that product rather than filling a cart. Each campaign carries exactly one gift, and the dearest of the six sells for A$22. None of them carries an end date, a usage cap or a once per customer rule. They are shelf furniture, not a promotion.
Read the buy column and a ladder appears that no campaign name mentions. Each list is a price band. Against the median item on its own list, the pencil gift is worth forty five per cent, the bib thirty three, the tote twenty eight, the rattle socks twenty five, the art book twenty two. The Connetix campaign is the outlier at eleven.
Five of the six lists are typed out product by product. Only the Connetix one points at a collection.
03 / The outcome The counter is a floor, not a total
Orders that carried a free gift, across the six live campaigns.
The counter moves once per order, not once per gift. Three of the six multiply the gift by how many qualifying items are in the cart, so a basket with two trigger products ships two gifts and counts as one. The real gift count is higher than the counter shows and cannot be worked out from outside.
It also covers only the six campaigns switched on today, and the storefront says nothing about what ran before them.
One campaign carries three fifths of the total. The three in the middle are not dead weight: together they are 1,752 orders, thirty nine per cent of everything. Only two sit near zero, and the pencil campaign is the newest of the six by nearly two years. The Connetix one is genuinely flat, eight months at well under half an order a day.
Line the two figures up and one row refuses to fit. The five campaigns whose gift runs between a fifth and a half of its list carry 4,508 orders between them. The sixth, priced at eleven per cent and sitting on the dearest list in the shop, carries 30.
Do not read that as cause and effect. The thin gift sits on the dearest list in the shop, so price and product move together and nothing here separates them. Claim volume says the offer works. It cannot say what any single setting did.
The number usually attached to this store is 27 per cent on average order value, from an earlier AOV.ai write up. It gives two windows for that figure, three months in the summary and four in the body, and it describes cart value thresholds, which are nowhere in the setup today.
04 / What this case shows Sort the catalogue by price, then pick a gift per band
The useful part is the method, not the gifts. Sort what you sell by price, draw two or three bands, and give each band a gift worth a real fraction of that band's typical item. A parent choosing between two feeding sets around A$66 takes the one with the bib. That same bib against a A$20 order is an expensive giveaway, and against a A$200 order nobody notices it.
Two questions then answer themselves. Which of your products get bought in a straight comparison against something similar? And what do you hold that costs little and belongs beside that product? If the second answer is nothing, this has nowhere to stand.
There is a catch, and it is the upkeep. This is one campaign per band, each with a list to be picked again whenever the range moves, and the longest of the six runs to fifty seven products. A narrow catalogue, or an obvious second purchase, is better served by one threshold at a tenth of the upkeep. And if the gift sits too far below the basket, as the Connetix one does, a campaign can stay live for eight months and produce almost nothing.
