KC
AOV Investigation
Hyde & Hare
Why AOV Jumped
Last 7 days vs the previous period, and the single change behind it
Period: 13–19 Sep 2026|
Compared to: 6–12 Sep 2026|
Source: Shopify · Meta · Google Ads · Klaviyo|
Prepared 20 Sep 2026
Section 01
The Answer
Verdict
The AOV bump is the free seat pad offer. It is not a campaign, not a platform, and not email.
AOV rose +13.8% to £173.74. Every penny of that came from the
"free seat pad over £150" automatic discount, which went from 6.6% to 25.4%
of all orders in a week. Strip those orders out and AOV actually fell 1.9%.
What the data confirms
- AOV rose entirely on basket size: units per order +14.0%, average item value flat at −0.2%
- Seat pad GWP attach rate quadrupled, 6.6% → 25.4% of orders
- GWP orders run £253.65 AOV vs £146.52 for everything else
- Median order value hit a 6-week high at £157.00
What it is not
- Not one campaign. AOV rose on Meta, Google, direct and organic at once
- Not the Meta seat pad campaign. Its own AOV fell from £241 to £189
- Not email. Klaviyo AOV tracked the site (+12–14%) rather than leading it
- Not whales. The biggest order was smaller than the prior week's
Section 02
What Actually Moved
AOV can only rise two ways: people buy more things, or they buy
more expensive things. Splitting the change tells us which, and here it is unambiguous.
Six weeks of AOV and basket size
Shopify orders, excluding draft orders, Etsy, pending and cancelled. Weeks ending Saturday.
The prior week (6–12 Sep) was the weakest of the six. It fell in the gap
between the Summer sale ending and the new offer landing. That flatters the +13.8% headline. But the lift is
still real on its own terms: 2.11 units per order and a
£157.00 median are both six-week highs, ahead of every week of the Summer sale.
Where the orders landed
Share of orders by value band. The £150–500 bands are where the offer bites.
Prior 7d (6–12 Sep)Last 7d (13–19 Sep)
Orders in the £150–250 band grew from 31 to 54 and £250–500
from 17 to 35, while the £50–100 band shrank. That is the signature of a spend threshold pulling baskets
upward, not of a few large orders skewing the average.
Section 03
The Proof
The cleanest test is to split every order by whether it took the seat pad offer,
then compare the two groups across both weeks.
AOV with and without the seat pad offer
Remove the offer orders and the underlying AOV goes down, not up.
Prior 7d (6–12 Sep)Last 7d (13–19 Sep)
Bridging the £21.08 AOV increase
Decomposed into offer take-up, offer basket size, and everything else.
More orders taking the offer added £9.38 and bigger offer baskets
added £13.84, together £23.22. The rest of the business subtracted £2.14. The offer explains
more than 100% of the movement.
Section 04
Correcting the Timeline
Premise correction
The offer is 11 days old, not 2–3 weeks
The first order carrying the seat pad GWP was 9 September.
There were zero between 9 August and 8 September. What was running two to three weeks ago
was the SUMMER 15% code (183 to 207 orders a week), which ended around 30 August with the
"Summer Sale Final Day" email.
So the business swapped a sitewide 15% discount for a
£150 threshold gift-with-purchase. That swap is the story. The discount rate went from
12.9–14.5% during the Summer weeks, down to 8.9% in the gap week, and now sits at
11.5%, cheaper than the sale, with higher AOV.
Daily offer take-up
Share of each day's orders carrying the seat pad GWP, 1–19 September.
Take-up was already running at 23–35% a day from 13 September. The
"Seat Pad GWP" email did not send until 17 September, by which point the step change had
already happened. The email followed the lift; it did not cause it.
Section 05
Not a Campaign or Platform
If a single campaign or platform were responsible, its AOV would rise while the
others stayed flat. Instead AOV rose everywhere at once, the signature of a site-wide change.
AOV change by channel, last 7d vs prior 7d
Platform-reported AOV for paid channels; Shopify and Klaviyo for the rest. All moved together.
Worth noting
The Meta seat pad campaign drives volume, not AOV
It launched 10 September (£0 to roughly £58–82 a day) and took
purchases from 7 to 23. But its own AOV fell from £241.29 to £188.52.
It is bringing more buyers to the offer, not making baskets bigger.
Google spend actually fell 11.5% while conversion value rose 36%. Nothing
was scaled up here. The same traffic simply converted into bigger baskets.
Most offer orders arrived through direct and organic search, not paid.
This is an on-site merchandising effect reaching everyone who lands on the store.
Section 06
Not Email Either
Email AOV did rise, but by roughly the same amount as the site as a whole.
Email orders simply contain the same offer; they are a mirror of the lift, not its source.
Klaviyo campaigns moved +12.0% and flows +13.6%, against a
site-wide +13.8%. Email is tracking the business, not pulling it.
Where email did help
Volume, not basket size
Recipients rose 25,729 to 39,738 (+54%) and attributed orders
went 48 to 71, lifting attributed revenue from £7,995 to £13,289.
Email meaningfully grew revenue this week. It just was not the reason AOV moved.
Section 07
What Customers Bought
The product mix shows the offer working exactly as designed: seat pads and
cushions attaching to rugs and beanbags.
Seat pads moved from a handful of units to dozens, while the larger
anchor items (Sexto rugs, shearling beanbags) held or grew. That is attach, not substitution:
customers are adding the qualifying spend and collecting the free pad.
Section 08
Caveats & What To Do
Caveat 01
The comparison week is soft
6–12 Sep sat in the gap between offers and was the weakest of six weeks
(AOV £152.66, median £92.50). The +13.8% is flattered. Against the four-week pre-gap average of roughly
£178, this week is about level on AOV, but with far better basket
economics and a lower discount rate.
Caveat 02
Watch for cannibalisation
Non-offer AOV fell 1.9%, and anyone already spending over
£150 now receives a free pad they would not have needed. The giveaway cost
£2,928 at retail across 56 orders (roughly £52 each) between 9 and 19 September.
True cost is COGS, which should be checked against margin.
Caveat 03
Attribution overlap
Platform-reported AOV from Meta and Google overlaps with Klaviyo and with each
other. Shopify order data is the single source of truth for the headline AOV figures; platform numbers are
used only to show direction by channel.
High
Check the GWP against margin before scaling
At roughly £52 retail per redemption and 25% attach, the offer is
cheaper than the 15% sitewide it replaced, but only if seat pad COGS supports it. Confirm the landed cost
per pad and set a redemption ceiling.
Protects margin before take-up grows further.
High
Test the threshold upward
The £150 threshold is being cleared easily. Median qualifying spend is already
£177. Testing £175 or £195 would capture more margin from
customers who are clearing it anyway.
Reduces cannibalisation of baskets that never needed the incentive.
Med
Push the offer harder through email and paid
Only one campaign has promoted it, on 17 September. The Meta seat pad campaign is
driving volume efficiently. Both channels have obvious room before take-up saturates.
Take-up is at 25% of orders with minimal promotion behind it.
Med
Re-baseline the reporting comparison
Compare against the four-week pre-gap average rather than the gap
week, so the offer is judged on a fair base rather than a soft one.
Avoids overstating the win in the client update.
Low
Watch non-offer AOV weekly
The −1.9% on non-offer orders is small but worth a trend line. If it keeps sliding
while offer orders grow, the offer is shifting baskets rather than growing them.
Early warning on cannibalisation.
Method. Shopify Admin API order-level data is the source of truth for AOV, basket and product figures,
excluding draft orders, Etsy, pending, cancelled and voided orders. AOV is total order value including tax and
shipping, divided by kept orders. Offer orders are identified by the automatic discount application titled
SHEEPSKINSEATPAD. Meta figures use 1-day view, 7-day click; Google uses last-click conversion value; Klaviyo
uses Placed Order attribution. Platform figures overlap with each other and are indicative of direction only.
All figures recomputed and reconciled independently on 20 September 2026.