Case study · Devotional candles
Alumenza makes devotional candles. Its AI workspace builds and fixes listings on all four channels and runs the Amazon ads.
One message put the candle line on Walmart in a day, and the same catalog went to Faire that afternoon. A write only counts once the channel itself shows it.
Done
58listings live on Walmart the same day
Counts are the agent’s reports that day; Walmart’s own count on Sep 27 was 62 items. Typos fixed, file path removed.
Listing fixes on the candles: new main images on the scented packs, and the candles moved to Amazon fulfillment on Oct 10. Each previewed, accepted and read back. The candle listings were made before the AI had an Amazon connection; the Amazon ads are below.
Sage, Cinnamon and Camphor on Oct 7, built from their Amazon listings: ASIN, UPC, main image and price, each as a 3-pack and a 6-pack. Free-shipping minimum raised to $75, checked with test carts.
58 listings live through Walmart’s API on Sep 17, with stock and prices, and the plain candles merged into one variation family. 62 items on Walmart’s own count by Sep 27. 16 price changes on Oct 9, read back.
23 wholesale listings live on Sep 17, with stock set and wholesale prices changed three times that evening.
Run through Auren since April 2026. Changes wait for a yes, and Auren’s limits block the ones that go too far.
Connected Oct 7, 2026. Each listing’s offer is read back, since a withdrawn listing can look live for minutes.
Connected Sep 2, 2026: 599 bank transactions posted and checked in Xero, 0 failed, 0 duplicates. It never pays and never clicks Reconcile.
Auren, the Amazon ads system our founder built, has run Alumenza’s ads since April. Each change is checked first, and Auren’s own limits block the ones that go too far.
Gross, before refunds and fees. Auren has run the ads since April.
April to June from Amazon’s seller data; July and August from Sellerboard; September is its first 28 days, as Auren published them.
Not all of the growth is the ads. New products launched in May and June, stock was restocked in August, and small TikTok and Google ad tests ran. Weekly sales have been flat since mid-September; the Oct 12 run on the Auren page is the response.
Requests in the owner’s words, typos fixed. Times are Eastern. Ads runs with no request on file are above.
create new campaigns, structure them and run them… i’d rather test the keywords, know they won’t work and negate, rather than just sit here and do nothing
3 campaigns, 12 ad groups and 12 product ads. A validate run checked the campaigns first; Amazon turned down 1 keyword as too long.
i want to add these into faire is what i am saying, the same items
Built from the Walmart feed and its SKUs: first 3 family listings with 64 variants, then a page for each color, saint and scent. Stock set high on the owner’s word.
…execute all of these, such high acos is definitely supposed to be lowered
A validate run first: 4 clean. All 4 went live and were read back, like “virgin mary candle” from $0.38 to $0.29. The top-of-search raises also asked for stayed on hold until Sep 21, after a Sep 7 cut.
create a faire listing for this too. it’s also on amazon, it’s the cross catholic candle 7 day candle as well, only in 6 pack… show me the listing title before you create it. make sure you put 6 pack in the front
Matched the product from the owner’s photo and files, showed the title, and waited. The owner changed the price and wording; then it created the listing and read it back as for sale.
i want to push the items from other places to ebay
Every listing passed eBay’s preview first, and the publish waited for the owner’s pricing. All 25 read back as live.
We start with a free assessment: the systems you use, the first workflow worth building, and what useful progress looks like.
More case studies: Once Legos →Auren →