Fix Kit #5/6: When the Fix Becomes the Solution
We Fixed the System. Now We Had to Agree What That Meant.
The prototype worked. But we still didn’t know if we’d solved anything.
At InHouse Secure, the S/4 demo system finally runs clean. The simulation doesn’t crash. The chaos doesn’t escalate.
Here’s who’s in the room:
- Tom Sayer — me, production manager, edge-of-the-floor realist and one-time pallet smuggler.
- Callum Nair — planning lead, spreadsheet loyalist, fighting to freeze the schedule by lunchtime.
- Joe Mensah — warehouse operations, king of the quiet shrug and guardian of the racks.
- Erin Patel — support, ticket wrangler, professional absorber of Friday panic.
- Priya Desai — master data specialist, logic-seeker and unexpected jokemaker.
- Charlotte Ibrahim — our CFO, calm, precise, allergic to fluff.
- Asha Khan — project manager, air-traffic controller for everything from workflows to personalities.
- And Guy, the consultant who taught us how to breathe through a broken spec.
No fireworks. No applause. No “ta-da” moment. Just a screen. A sandbox. And a test run that didn’t break.
We’d staged a real production kit in S/4. Multiple fixed bins. No errors. No pallet hidden five metres in the air like a squirrel stash.
Here’s the part that stops us: We hadn’t built a single custom enhancement.
It takes time. More than we want to admit. Modelling the real logic. Running it again and again. Tweaking the sequence. Cross-checking it against the way we actually work.
But when it runs clean? The clarity is priceless. Because for the first time, people don’t just say “maybe.” They start to believe: “We can pull this off.”
Third time lucky. But even then — even when the config fits the chaos — something’s still missing.
We Fixed the Flow. But Not the Feeling.
Everyone’s quiet. Not sceptical. Not stunned. Just… waiting.
Because we all know what comes next: The emails. The rollout talk. The assumption that it’ll scale just because it ran once.
We’re not ready. Not yet.
Before anyone can fire off a slide deck, Guy steps in. “Let’s stop. One more session. One question.”
He books another hour.
Same Room. Different Purpose.
This isn’t a post-demo huddle. It’s not classic improv. No riffing. No rapid-fire build. This is something slower. Grounded. More deliberate.
It feels like improv, but what it really is? Group reflection with ensemble thinking.
The kind where everyone speaks. One at a time. Builds gently on what was just said. No corrections. No counters. Just shared momentum.
In the room: Charlotte. Callum. Joe. Erin. Asha. Someone from Finance who survived the audit trail. And me.
Guy opens with: “What would make you say — without hesitation — this worked?”
No smartboards. No KPIs. Just voices. One at a time.
What We Say
Joe: “If I’m not firefighting at 6 a.m. because the sequence changed overnight — that’s a win.”
Callum: “If the published schedule isn’t treated like a placeholder.”
Erin: “If I don’t get six Jira tickets on a Friday tagged URGENT by people who haven’t spoken all week.”
Me: I pause. Then: “If the system tells us something’s missing — and we actually stop.”
They wait. So I go on.
“I know what this sounds like. I’ve spent my whole career keeping the line running. Always. You never stop production unless something’s on fire and nobody’s got the extinguisher.”
They nod. Some smile. Because they know. That’s the rule. Unspoken, enforced by KPIs and caffeine.
“But if we pretend the part is there when it’s not, we feed the same failure again. I’d rather stop for an hour and fix it than steal from the next shift.”
I look at Asha. Then Joe. Then Charlotte.
“This isn’t easy for me to say. But if we want a system that tells the truth, we have to let it. And when it does, that can’t just be my problem. If we want this to work, we stop together. We fix it together.”
Nobody argues. And for once, I don’t feel like I’m carrying it alone.
Then Asha Asks the Quiet Question
“Okay. What would that look like — in real terms?”
The room pauses. Then we build.
The Joke That Sparked the Metrics
Priya (half-joking): “If we go a full month without someone overriding an order in Excel, I’ll buy lunch for the floor.”
We laugh. Then we nod.
And just like that, a metric lands.
Callum: “90% adherence to the frozen schedule. No re-sequencing unless we all agree by 2 p.m. the day before.”
Erin: “Cut ticket volume on plan mismatches by half.”
Me: “If a shortage shows up, we stop. No silent fixes. No sneaky racks. We deal with it in the open.”
Then Guy taps the AI.
How AI Helps Us Think (Not Just Document)
He doesn’t feed it metrics. He feeds it questions:
- “What would make us feel less burned?”
- “What patterns trigger escalations?”
- “What outcomes show we trust the system again?”
It gives us themes. Groups our words. Reflects what we’re all saying — just in different ways.
And it drafts phrasing we can live with. Not legalese. Not fluff. Just clarity.
A one-page “success definition.” No targets yet. Just truth.
What We Leave In (That Might Surprise You)
The core warehouse setup? Stays. Supermarket buffer? Still valid.
But one thing we drop? The five-metre shelf trick.
No more parking pallets out of reach to avoid early grabs. That ends now.
If something’s missing, we stop the line. Deal with it out loud. Fix the cause. Not hide it.
Because stealing from ourselves isn’t clever. It’s how we end up right back here.
And if we say we trust the system? Then we start acting like it.
That Was the Shift
We weren’t removing duct tape. We were choosing which pieces still matter.
What Most Projects Miss
They think go-live is the goal. They assume config running clean means it worked.
But success shows up in what people stop doing:
- The 5 a.m. override.
- The whispered “just fix it, don’t log it.”
- The unspoken rule that says: never stop production.
We didn’t write policy that day. We wrote clarity.
And that made all the difference.
And the Joke?
The one about lunch? Three weeks in. Still no Excel overrides.
We’re hungry. But we’re ready.
P.S.
If your success criteria doesn’t include breathing room… You haven’t finished building yet.
