September held flat, carried by your returning customers and Labor Day
September 2026 at a glance
vs. September 2025| Metric | September 2026 | vs. Sep 2025 |
|---|---|---|
| Net sales | $386,529 | +1.1% |
| Orders | 6,010 | −8% |
| Average order value | $67.31 | +14% |
| Ad spend Meta, Google, TikTok | $64,391 | +49% |
| ACoS 20% opened for the push | 16.7% | inside 20% |
| Returning customer sales | $315,816 | +2% |
Net sales by month
2025 vs 2026 · thousands of USDShopify net sales Jan–Sep: $4.12M in 2026 vs $4.39M in 2025; the gap sits in June (Anniversary) and August. Ad spend: Meta, Google Ads and TikTok via AgencyAnalytics.
The ads that sold showed a real person or a real problem
Top sellers in September
Meta · sales and ROAS
What didn't sell
polished, generic images
The pattern
Problem-specific angles and real people convert; aesthetic ones get clicks but not sales, for the second month in a row. The three organic-style UGC videos drove ~37% of prospecting purchases, two months running.
What we committed to in the August recap: done
- DoneResolve the brand search cost increase. Stop paying up for brand searches and let organic carry them. Organic picked the traffic up.
- DoneRework the Fall creative set. Warm Spices came back, from 0.73× to 3.03×, and the new set leads with problem-specific angles.
- DoneBuild out Good Pet, plus In Bloom. The Pets creative scaled 5× and held 4.2×. The In Bloom sampler is live since Sept 30.
- DoneScale TikTok while the ratio holds. Spend +27% vs August, and the return held: 4.32× → 4.52×.
Google: moving off brand worked
Brand searches
share of brand searches where our ad showedWith less brand spend, Google organic sales roughly doubled. People looking for Edens Garden still find you.
Where the budget went
Google · September| Campaign | Spend | Cost / conv. |
|---|---|---|
| Performance Max (3) | $9,027 | $22–29 |
| Search · single products | $6,601 | $48 |
| Search · catch-all | $3,014 | $21 |
| Shopping · high priority | $2,315 | $27 |
Single-product search is held to a cost-per-new-customer target. Brand stays small and capped, for misspellings and variants.
Prospecting kept showing the same ideas to the same people
How many times each person saw our ads
healthy on cold: 3 or lessReaching 1,000 new people cost $59 in June and $143 in September. Same budget, fewer new people.
The concepts are tired, not the audience
With broad targeting, the creative does the targeting. Twenty versions of one idea find the same people twenty times. The click got cheaper, because people who know us click again, but getting someone to add to cart now costs 80% more ($1.99 → $3.59).
Good news: the site converts like last year
Sessions that added to cart
Sep 2025 → Sep 2026What this changes
Once Meta visitors add to cart, they check out like everyone else. The gap is before the cart: people who've already seen us many times. The site isn't what's holding back new customers. The lever is what we put in front of new people.
A concept is one reason to buy, for one person, in one moment
Different arts · same concept
Meta treats them as one ad- Stress Less roll-on on a lavender background "Calm, in your pocket."
- The same roll-on on white, new photo "Your pocket-size calm."
- The same message as a 10-second video Bottle turning, headline on screen.
Useful to refresh a winner. It doesn't find new customers.
Different concepts · same product
a new audience for each- The 3pm reset A woman between meetings: "The thing in my bag my coworkers keep asking about."
- The car-line mom Ten minutes before school pickup: "My two-minute reset before the chaos."
- The gift for a friend who's stretched thin "For the friend who's holding everyone together."
Each one reaches people the others don't. That's how frequency comes down.
Ten concepts, from 13,303 of your reviews, your site searches and what already worked
Several build on ideas you've run before. What's new is the person and the moment.
One sketch per concept, in your look
Sketches to show the idea, not final ads. Scenes are AI mock-ups built from your product shots; live ads use your photo bank, UGC or a shoot.
How we test them: six weeks, inside the campaign we already have
Stress · Sleep · The swap · Fall
The biggest review themes, plus the season we're in.
First order · Body · Mom kit · Start with six
New-customer entry points: low-risk first purchases.
Gifting · 17 years + the winners
Ready before Black Friday, with what we learned.
How we'll report October
four numbers| Number | September | Goal |
|---|---|---|
| Prospecting frequency times each person saw us | 9.4 | under 3 |
| New people reached main prospecting campaign | 166K | ↑ |
| Meta visitors who add to cart Shopify | 3.6% | ↑ |
| New customers vs last year | −9.5% | ↑ |
Three things for the season ahead
25 Days of Wonder: the plan and the budget
Your offers, dates and budget for November, so we can build the media plan and the creative around them in time.
A little more lead time on promo creative
Assets and copy about a week before launch give us room to review them, suggest improvements and set everything up well.
A brainstorm on the concepts, together
The ten concepts are sketches, not finished ads. An hour with your team to shape them: what feels like Edens Garden, what you've already tried, and what's missing.
Sources: Shopify Analytics, Meta Ads, Google Ads, TikTok Ads and AgencyAnalytics, September 1–30, 2026 against the same period in 2025. Meta's attribution model changed in March 2026, so its year-over-year figures are not a clean comparison.
