Every Friday at 4 p.m. WIB, for two years, Rizky Pratama's team in Jakarta did the same dance. Someone opened Q3_Forecast_FINAL_v14.xlsx. Someone else asked whose laptop had the macro that actually worked. Outside, the sky over Kuningan did its usual afternoon collapse into rain, ojek drivers pulling ponchos over their jackets at the lights, horns stacking up in the macet below the office window. Inside, eight sales reps sat around a table and told a story about the quarter that none of them quite believed.
"Gapapa, itu masih on track" — it's fine, that one's still on track — became the unofficial motto of the call. A deal that hadn't moved in three weeks stayed at 80% because moving it to 40% meant explaining why, out loud, to Rizky's boss, who was dialed in from Singapore. Nobody wanted to be the one who dragged the number down. So the spreadsheet grew a second, invisible layer of optimism on top of the real pipeline — a layer made entirely of people not wanting an awkward Friday.
"We weren't forecasting the business. We were forecasting how each rep felt about the conversation they were about to have with me."— Rizky Pratama, Revenue Lead
The Friday Ritual, Deconstructed
Rizky had inherited the spreadsheet from his predecessor, who had inherited it from someone before that. Fourteen tabs. A macro that recalculated weighted pipeline based on stage, except the macro broke every time someone inserted a row, which was often, because deals kept sliding between stages that didn't map cleanly onto the tabs. By the time the number reached the regional review, it had passed through six sets of hands, each one nudging it a little closer to the target everyone secretly knew they were supposed to hit.
None of this was dishonesty, exactly. It was self-preservation dressed up as diligence. A rep who marked a deal as "at risk" got a coaching call. A rep who kept it green got left alone. The spreadsheet rewarded confidence, not accuracy — and confidence is cheap to fake at 3:45 p.m. on a Friday when everyone just wants to go home before the flooding starts on Jalan Gatot Subroto.
The Day We Deleted It
What changed wasn't a new discipline or a stricter manager. It was removing the thing that had to be filled in by hand at all. Rizky's team moved their pipeline onto piRevenue mid-quarter — deal capture built from the notes, calls and messages reps were already producing, not a form bolted on top of their day. The forecast stopped being a document somebody assembled once a week and became a number that simply existed, continuously, built off actual stage movement and logged activity rather than a rep's mood on a Friday afternoon.
The first live number it produced was 22% lower than the spreadsheet's number for the same week. Rizky remembers staring at it, half-convinced something was broken. Nothing was broken. The spreadsheet had just been wrong — quietly, consistently, for two years.
Why the Number Got More Honest, Not Less
This is the part that surprised Rizky's team most. A smaller number felt, at first, like a worse one. It wasn't. It was simply a number nobody had touched. An agent logs a deal from the call notes and the WhatsApp thread the rep forwards after a client meeting; it doesn't have an opinion about whether Friday's regional call will go well. It doesn't get coached into optimism. A deal that's gone quiet for three weeks shows up as quiet, in weighted stage and probability, whether or not anyone wants to explain it out loud.
That single change — removing the human incentive to round a number up — did more for forecast accuracy than any amount of "forecast hygiene" training ever had. Sandbagging disappeared too, for the same reason: there was no longer a manual step where a rep could quietly under-report a strong deal to protect next quarter's number. The pipeline simply reflected the pipeline.
Rizky's boss in Singapore noticed the drop before Rizky had a chance to explain it, and the first reaction, understandably, was suspicion — had something gone wrong with the migration, had deals gotten lost in the switch? It took two more forecast cycles to prove the opposite: the new number moved less week to week, and when it moved, it moved because a real deal had genuinely progressed or stalled, not because someone had reworked a formula the night before the call. Predictability, it turned out, was worth more to the regional review than a bigger number ever had been. A forecast that's wrong in a consistent, explainable direction is manageable. A forecast that swings ±20% depending on who updated it last is not.
What the Team Does With Friday Now
The forecast call still happens. It's fifteen minutes now instead of ninety, and nobody opens a spreadsheet during it. The conversation is about three specific deals that need a decision — a discount to approve, a stakeholder to loop in — not a line-by-line audit of whether every number in a 14-tab file adds up correctly. Rizky's reps spend the other seventy-five minutes doing what they were hired to do: talking to buyers in Jakarta, Surabaya and Bandung, closing the deals the agent flagged as going quiet before they went cold.
See a live forecast on your own pipeline
The Number You Can Actually Defend
Rizky still walks into the regional review every month. The difference is what he's carrying in. Not a printout massaged by six people on a Friday afternoon, but a number built the same way, every day, from what actually happened in the pipeline — no rounding, no rehearsal, no macro that breaks when someone inserts a row. It's a smaller number, most weeks. It's also the only one he's ever fully believed.
There's a second, quieter change Rizky didn't expect. Reps stopped treating the forecast as something to manage and started treating it as something to trust — which meant they stopped hiding weak deals from him too. When the number in the system isn't going to be used against you in a Friday cross-examination, there's no reason to keep a stalled deal looking healthier than it is. Rizky's team now flags at-risk deals earlier than they ever did under the spreadsheet, simply because doing so no longer costs them anything socially. Honesty turned out to be a side effect of removing the audience.
That's the whole bet behind piRevenue: agents do the unglamorous, continuous work of logging and weighting the pipeline so the forecast is simply true by default. The close — the actual conversation with the actual buyer — stays entirely human. Rizky's team didn't get a smarter spreadsheet. They got to stop needing one.
Stop feeding the CRM. Watch it feed you.
