By Andy Siemer, CEO
When a new client gets on a call with us, how do they usually want to start? Almost never with the thing that's actually costing them money. They want to start with the thing that's annoying them.
A client of ours runs a custom apparel and promo shop. We'll call the owner Marcus in this telling, no company name needed. Marcus got on a call with our team and walked us through his stack, and about ten minutes in he stopped and said it himself: "we are — our current systems are literally just piecemeal together." A sourcing tool for finding product. A Google Doc for tracking jobs, because somebody made it up one afternoon and it stuck. And underneath all of it, an old ERP system running quotes, invoicing, and purchase orders for the whole back office.
Marcus already knew which piece he hated most. It was the ERP. Antiquated, clunky, the kind of system that works but makes you wince every time you touch it. He told us flat out he'd love to replace it. "I would love to replace that system. Like, absolutely love to. It's not high on my priority list, though." That last sentence is the one that matters. He knew what he wanted to rip out. He wanted it gone. That didn't make it the right place to spend the first dollar.
That's usually how it goes. Whatever's ugliest gets the attention, because it's the thing everybody on the team complains about, the thing that makes people roll their eyes in meetings. And the instinct that follows is a big-bang instinct: rebuild the backend, replace the whole system, fix the thing everyone hates in one swing. I get why. It's the visible wound. But the ERP, bad as it was, wasn't where Marcus's business was actually bleeding.
So when the conversation drifted toward the backend rebuild, I said it plainly: "definitely wouldn't start there."
So where was the time actually going? Marcus's team wasn't losing it to backend plumbing that only saves pain later — it was losing it to work that touches revenue today. Client sends over a Canva file for a t-shirt design, and somebody on the design team manually converts it into a vector file for printing. A customer wants ideas for a job, and a CSR builds a sourcing presentation from scratch, digging through the same tools, assembling the same kind of slide deck they've assembled a hundred times. Design proofs bounce back and forth as Dropbox links and attachments, with all the friction that comes with hunting down the latest one. None of that shows up on a system diagram anywhere. It just shows up as hours nobody has.
So I told him plainly: "I [can get] you more profit fixing those front-of-house issues, and then swapping over that system could be easier." That's the sequencing, and it's not a consolation prize for delaying the ugly fix. It's the faster path to money. Template the sourcing presentations so a CSR reviews and sends instead of building from a blank page, put the proof revisions somewhere the latest version is just the version. Automating the Canva-to-vector conversion is totally doable — first thing we'd check is what format the print pipeline actually accepts. Do that first, and the CSR team gets its time back right away, freed up to spend more of it on the bigger, more custom orders instead of the repeatable ones a system can carry.
And then, once that pressure comes off the front of house, the ERP swap gets easier too. You're not trying to replace a system while the same team is drowning in manual sourcing work. You've got room to do it right.
I assume your list looks similar — a system everybody hates up top, and the real leak three lines down. Correct me if I'm wrong. The ugly system is rarely the expensive one. The expensive one is usually three people redoing the same manual work every day.
The part that told me this was more than my own bias toward sequencing: nobody on our side had to argue for it. One of our engineers said the same thing on the call, unprompted: fix the customer-facing stuff first — it saves the most money and costs the least to execute. Two people independently landing on the same order isn't proof, exactly. But it's a fair sign the sequencing is right.
More often than people expect, the first deliverable usually isn't a build — it's a data-flow map, plus the list of things we're deliberately not touching yet. The biggest item on the list is rarely the one worth fixing first, and saying that to the client on paper is most of the value of the first conversation. Everything after that is just execution.
Marcus's ERP is still sitting there, antiquated as ever, the system he'd love to be rid of. It's just not first. When the front of house stops leaking, swapping it gets a lot easier — and cheaper.
First pass is a map of where your money actually leaks — you correct it. Start the conversation.