Five stores. Five automation pipelines. Real numbers.
Not mockups, not projections. Here's exactly what we built, how we built it, and what changed, across WooCommerce, Shopify, social media DM commerce, and BigCommerce.
The situation
Mike runs a clothing store, a WooCommerce store selling apparels to customers across Abuja and Lagos. Orders had grown steadily for a year, but so had the WhatsApp and email pile of "where is my order" messages. She was manually checking WooCommerce order statuses against her courier's tracking page every morning, then copying updates into individual replies. Some customers waited two full days for a response. Returns were worse: each one meant a back and forth of five or six messages before a refund was even approved.
What we built
Audited the manual process
Shadowed Mike for two days to log every WISMO message, return request, and how long each took to resolve by hand.
Connected WooCommerce to n8n
Set up a webhook so every order status change in WooCommerce triggers an automation instantly, no manual checking required.
Built the WISMO auto-reply flow
Order status pulls from WooCommerce and the courier API, merges into a message template, and sends automatically by WhatsApp and email.
Automated return approvals
Return requests are checked against a rules engine (item condition, days since delivery); simple cases auto-approve, edge cases route to Mike on Slack.
Tested with real order data
Ran the flow against two weeks of historical orders before going live, to catch edge cases like split shipments and address changes.
Launched and monitored
Went live gradually, watching the first fifty automated replies closely before letting it run fully unattended.
The result
"Running three platforms used to mean three separate messes. airmirate put everything on one line and it just works."
The situation
Layer & Line sells streetwear on Shopify, with a spike in traffic every time Tariq posts a new drop on Instagram. The problem: most of that traffic added items to cart and never checked out. Tariq could see the abandoned checkout count in Shopify but had no system to follow up, so every drop leaked sales he never got back. Manually messaging even the biggest carts took hours he didn't have between sourcing and packing orders.
What we built
Mapped the drop cycle
Studied three previous drops in Shopify analytics to see exactly when and why carts were abandoned during high traffic spikes.
Connected Shopify checkout events to n8n
Set an automation trigger on abandoned checkout, capturing customer, cart contents, and time since abandonment.
Built a three-touch recovery sequence
SMS at 1 hour, email at 6 hours with a soft nudge, and a final WhatsApp message at 24 hours with a small time-limited discount.
Added stock-aware logic
If the abandoned item sold out in the meantime, the flow automatically swaps in the closest in-stock alternative instead of a dead link.
Load-tested for drop-day traffic
Simulated a full drop-day spike to confirm the automation held up under sudden volume, not just steady daily traffic.
Launched on the next live drop
Went live during an actual product drop, watching recovery rates in real time and adjusting message timing on the fly.
The result
"I genuinely didn't believe the number until I saw it in Shopify myself. Carts that would have just disappeared are now actual orders, and I didn't touch a single one of them."
The situation
Voltage Gadgets sells phone accessories and small electronics on Shopify, a category with a specific problem: customers ask detailed pre-purchase questions (compatibility, warranty, charging specs) before they'll buy, and post-purchase questions (defects, warranty claims) after. Funmi's small team of two was fielding both, all day, across Shopify's inbox, Instagram DMs, and email, with no way to prioritize which question actually needed a human.
What we built
Catalogued common questions
Pulled six months of support history to identify the 20 questions that made up 80% of all inbound messages.
Built a product-aware AI agent
Connected an AI agent in n8n to the live Shopify product catalog, so answers about specs and compatibility are always accurate, not scripted.
Automated warranty claim intake
Defect reports now auto-collect order number, photo, and issue description in one structured flow instead of a scattered back and forth.
Set a confidence threshold
The agent answers confidently on catalog facts, but escalates ambiguous or emotional messages straight to Funmi's team on Slack.
Unified all three channels
Shopify inbox, Instagram DMs, and email now feed into the same automation instead of three separate manual queues.
Reviewed and refined weekly
First month included a weekly review of escalated conversations to keep improving the agent's answers.
The result
"We were basically two people trying to be a call center. Now the agent handles the repetitive spec questions, and my team only sees the conversations that actually need a human."
The situation
Pressed & Poured has no website. Blessing sells entirely through Instagram posts and WhatsApp orders, which is common for food and drink brands starting out. The problem was volume without structure: orders came in as DMs, voice notes, and comments, with no consistent way to capture an address, confirm a payment, or know what was actually in stock that day. She was manually copying every order into a notebook before her evening delivery run, and mistakes were common.
What we built
Mapped the real order journey
Tracked exactly how orders arrived across Instagram comments, DMs, and WhatsApp to find where the process broke down.
Built a WhatsApp ordering flow
Set up an automated WhatsApp menu so customers select products and quantities in a guided flow instead of free-form messages.
Connected Instagram DMs into the same system
Comments and DMs on Instagram now route straight into the same WhatsApp ordering flow, so nothing gets missed on a busy post.
Automated daily stock updates
Blessing updates one simple stock sheet each morning, and the flow automatically stops offering items that have sold out that day.
Structured every order automatically
Each completed order auto-populates a delivery sheet with address, items, and payment status, ready for the evening run.
Piloted on a single week of orders
Ran the system in parallel with her notebook for one week to confirm accuracy before fully switching over.
The result
"I didn't think I needed a website to need automation. Turns out the chaos was never the website, it was DMs and voice notes with no system behind them. That part is fixed now."
The situation
Bramwell Home runs on BigCommerce, selling furniture and home decor with higher order values and longer delivery windows than a typical store. That combination meant every order generated multiple check-in messages from anxious customers over the two to three week delivery window, on top of the usual support and cart recovery load. Emeka's ops team of three couldn't keep pace with both proactive updates and reactive questions at the same time.
What we built
Audited the full order lifecycle
Mapped every stage from order placed to delivered, including the manufacturing and freight steps unique to furniture.
Connected BigCommerce to n8n
Built a webhook layer so every order and fulfillment status change fires an automation instantly, across the full multi-week timeline.
Built proactive milestone updates
Customers now get automatic updates at order confirmed, in production, shipped, and out for delivery, before they have to ask.
Layered in the after-sales system
Added the same WISMO and returns automation used in other case studies, tuned for BigCommerce's order data structure.
Added cart recovery for big-ticket items
High-value abandoned carts trigger a personal, lower-pressure follow up rather than a generic discount blast.
Rolled out in phases
Launched proactive updates first, then support automation two weeks later, then cart recovery, so the ops team could adjust gradually.
The result
"With a three week delivery window, silence is what kills customer trust. Now the customer hears from us before they even think to ask, and my team's time goes to the orders that actually need a person."
Ready to see your own pipeline built?
Every case here started with the same first step: a short call to map what's currently manual. Let's do the same for your store.
Book a call