I built the automation layer behind a $7M car-electronics brand.
Nine months owning the fulfillment, logistics, and revenue-recovery systems behind Merge Screens — three supplier pipelines, a custom GraphQL shipment tracker, and abandoned-cart recovery built from scratch.
Three Supplier Pipelines, Held Together by Spreadsheets
Merge Screens runs a Shopify store shipping premium car-infotainment hardware — Tesla-style screens, CarPlay modules — sourced through three separate supplier pipelines, each with its own order sheet, its own tracking process, and its own quirks.
When Carlos brought Waqass on, the business was already doing $4M a year — but the operations layer hadn't kept pace with the growth. The revenue was real; the plumbing behind it was improvised.
Order tracking lived scattered across multiple Google Sheets. Fulfillment updates meant manually cross-checking spreadsheets against Shopify, order by order. There was no way to see at a glance which shipments had stalled in transit. Abandoned checkouts — real, high-value revenue walking out the door — were never followed up on at all. And the automation that did exist was brittle: a single teammate inserting a column in a spreadsheet could silently break the whole chain without anyone realising until orders started slipping.
"We are an e-commerce company pushing $4M/year in sales, so there will be often work."
The brief was deliberately open-ended: keep the fulfillment engine running without fail, and build whatever the business needs next. Not a one-off script — an ongoing partnership to own the automation layer as the company scaled.
A System of Automations, Built to Scale With the Business
Rather than patch individual problems, the work grew into a connected automation layer spanning Zapier and Make.com — each piece built with real logic, fail-safes, and error handling so nothing breaks silently. Here is what runs the operation today.
1 · Three-brand fulfillment sync, unified
Independent Zapier pipelines for each supplier. The moment a tracking number lands in a Google Sheet, the automation finds the matching Shopify order, creates or updates its fulfillment, and marks it shipped.
A custom fix handles a case Shopify's native tools miss entirely: in multi-product orders, every line item gets marked fulfilled — not just the first — so customers never receive half-shipped status updates.
A shipment tracker Zapier and Make couldn't build out of the box
No packaged connector existed for checking whether a shipment had actually moved. Rather than settle for a workaround, Waqass went to the API directly:
"Make or Zapier didn't have a node for this, so I read the docs to curate a custom GraphQL query to get tracking info."
The result is a daily automation that queries Shopify's GraphQL API directly across all three supplier pipelines in parallel, flags any order that hasn't moved in three or more days, and automatically clears orders once they're delivered — turning "which shipments are stuck?" from a manual spreadsheet hunt into a standing, self-maintaining report.
Abandoned checkout recovery, built from nothing
There was no cart-recovery system at all — high-value checkouts were simply being abandoned and forgotten. Waqass designed and shipped one on Make.com: a webhook catches every checkout, a staged delay confirms whether it genuinely became abandoned, a filter screens for orders worth chasing, and an AI step enriches the record before it's logged for the sales team. Recovered sales are automatically attributed to the rep who closed them, through a discount-code convention.
Automated refund handling, wired into Klaviyo
A flow that detects returns, verifies delivery status against the carrier, creates the Shopify refund, and triggers the right Klaviyo email sequence — keeping customer communication consistent and on-brand without a human touching it.
5 · A support-to-fulfillment bridge
Customers submit dashboard photos through the support inbox (Reamaze) so the team can confirm the right hardware. Those requests used to get lost between systems. Now they're automatically parsed and attached to an Asana task for the fulfillment team — nothing falls through the cracks.
6 · Order routing across the three brands
Orders flowing in through Asana are parsed by an AI step and routed into the correct supplier's Google Sheet, so each pipeline stays clean and every downstream automation fires on the right data.
Deliverability monitoring across every provider
A dedicated tracker keeps a live check on delivery timing across all suppliers — flagging stalled shipments, clearing delivered ones, and giving the team a single view of logistics health instead of three disconnected spreadsheets.
What Happens When Production Breaks in a Way Nobody Expected
Nine months into an always-on retainer, the real test isn't the build — it's what happens when live production breaks in a way nobody planned for. This is where deep system understanding earns its keep.
The signature incident: twice, a routine spreadsheet edit — a teammate inserting or deleting a column — silently shifted which column Zapier's trigger was reading from. Instead of watching the Tracking Number field, the automation started reading the Carrier field, and fired against the wrong data.
~3,000erroneous automation runs before anyone noticed — found late one night, diagnosed to the exact root cause (Zapier was keying on column position, not column name), and every affected order corrected by hand.
None of these show up in a demo. They show up in production, on live customer orders — and the only way to catch them is to understand the system deeply enough to know something's wrong before the client does.
Still Running, Nine Months Later
- What started as "keep a few zaps running" grew into full ownership of the automation layer — across fulfillment, logistics, support, and revenue recovery.
- The business has grown to a $7M+/year run rate — roughly 3× in the last year — and the automation layer scaled with it, not against it.
- Elevated from outside freelancer to embedded team resource, invited directly into the company's internal channel to work alongside their staff rather than through tickets.
What Carlos Said Along the Way
"I'm liking how fast and proactive you work so far Waqass. Good job so far."
"Good you found this out today Waqass. Thanks a lot man. Very good job and dedication."
"Very happy so far with your speed Waqass."
You really have great capacity to figure out the easiest solution instead of over-architecturing things. Very good job Waqass.
Have a fulfillment or ops process held together by spreadsheets and hope? Let's talk about what an automation layer like this could look like for your business.
Book a call with Botify →