StoreonAutopilot

Running a European home-decor store with AI agents. €1,000 budget. Every number public, weekly.

[StoreonAutopilot @StoreonAuto]

I'm building an online store in public. Run with AI agents. On a €1,000 total budget.

Home decor: cosy stuff + storage. Dropshipping. Ships within the EU. Open since Sept 23.

AI does most of the work: product visuals, code, copy, daily tasks. I check, fix and decide.

You've seen the "AI runs my store, $10k profit" posts.
This is not that.

Here you get the real numbers. Visits, costs, ad spend, sales. Including when sales are zero.

Day 13: 324 visits, 11 add-to-carts, 0 real orders.

Follow if you want to see whether a €1,000 AI-run store can actually work.
Or watch it fail, honestly.

[StoreonAutopilot @StoreonAuto]

@irabukht At $2 an ad, making creative stops being the bottleneck. Every visual in my store is AI-made and that part is easy. Knowing which one converts is the part nobody has automated.

[StoreonAutopilot @StoreonAuto]

@Noahtozero A checkout breaking after a theme change is the scary part of letting AI touch a store. Do you place a test order after every change, or catch it from the numbers?

[StoreonAutopilot @StoreonAuto]

@virthium Fair point on trust. Is a feedback rebate basically paying for reviews? I would worry about breaking platform review rules, especially on a brand-new store with zero reviews.

[StoreonAutopilot @StoreonAuto]

20 days live.
423 visitors.
0 sales.

The store was built almost entirely by AI agents.
Where's the leak?

[StoreonAutopilot @StoreonAuto]

@PanderShirts Classic trailing-data problem: it optimizes for last quarter's winners, not the next six weeks. Do you ignore the suggestions entirely, or still use them as a signal for something?

[StoreonAutopilot @StoreonAuto]

@helenamktg Curious what the agent actually controls in that store: ad budget, product picks, or mostly reporting? In my setup the agents build and write, but I still make the spend calls.

[StoreonAutopilot @StoreonAuto]

@joshuaday @liadyosef Didn't know WebMCP was already live on every Liquid storefront. With almost no traffic yet, is it worth setting up now, or is it a later problem once the basics convert?

[StoreonAutopilot @StoreonAuto]

@liadyosef Thanks, super useful. I'm running a small Shopify store with AI agents and 0 sales so far. If you had to pick one first, MCP server or WebMCP? Trying to work out what's worth the time at my size.

[StoreonAutopilot @StoreonAuto]

Pulled it. About 3% of sessions add to cart, almost all of those reach checkout, and none convert yet (a few checkouts were my own tests). So two leaks: the product page loses most people, and checkout loses everyone left. Fixing checkout first, since that's where intent already exists.

[StoreonAutopilot @StoreonAuto]

The store stays anonymous, that's the one rule of this account. The funnel I can share. Since Sep 23:

430 sessions
14 added to cart
11 reached checkout (a few were my own tests)
0 real orders

Traffic is thin, but the leak that hurts is checkout → order. That's what I'm digging into now.

[StoreonAutopilot @StoreonAuto]

@CoopEcom Thanks! My first 7-day Meta test was about €12/day (€83.50 total, 79 clicks, 0 sales). Starting to think that's too little data to learn from. Would you raise the daily budget first, or fix the creative before spending more?

[StoreonAutopilot @StoreonAuto]

@EricRausch Fair PSA. For a store that's only 3 weeks old with no email list, what would you lock in first for BFCM: the offer, the creative, or the traffic test? Trying to sequence it on a small budget.

[StoreonAutopilot @StoreonAuto]

@ClaudesRH Let’s win this

[StoreonAutopilot @StoreonAuto]

@helenamktg Curious how much of the daily ad decisions the AI makes alone. On my store (AI agents, day 20, 423 visitors, 0 sales) I still approve every budget change. Where do you draw the line?

[StoreonAutopilot @StoreonAuto]

@ecomedyy Profit up while revenue dipped is a good sign. What moved it most, ROAS or product margin? Curious what you'd fix first to get from $6k to $10k a day.

[StoreonAutopilot @StoreonAuto]

@ecommozart CS finding the winning angle makes sense, they hear the real objections daily. Do you feed their tickets into your ad briefs on a fixed schedule or ad hoc?

[StoreonAutopilot @StoreonAuto]

@ecombenni One main spender carrying a 3.67 ROAS is the fragile part. How many new creatives per week are you testing to find the second winner before you push budget?

[StoreonAutopilot @StoreonAuto]

@lukasenECOM $0.36 CPC with a $4 CPM is a cheap click, so I'd suspect the leak is after the click. Did you compare add-to-cart rate to clicks before scaling those 17 ads?

[StoreonAutopilot @StoreonAuto]

@levelsio @nomadscom An API priced at $2.5k to $10k a year, built for agents first. Do you expect the agent or the human behind it to be the one who decides to buy?

[StoreonAutopilot @StoreonAuto]

@AMannhaft Ads that do not feel like ads makes sense in Q4. Does the singing format hold up on cold traffic, or does it mostly work on retargeting? Curious how it plays for low-ticket products.

[StoreonAutopilot @StoreonAuto]

@markdmei Shipping transparency is underrated. On a new store with zero reviews, does an accurate delivery estimate build more trust than a discount before the first purchase, or only after it?