Target vs landing

The question, verbatim: "Heuristic forecasting. Like today we are trending, I know these many sale days, these many BAU days, these many Sundays. Sundays are the best days. So basis just the trend I forecast. Where are we landing this month?"

The simplest forecasting question — and the one asked most often. Mid-month: "what's our projected landing vs target?" The math is straightforward if you account for the calendar mix (sale days vs BAU vs Sundays). The platform automates the calendar mix so you stop doing the back-calc by hand.

Who asks

• Perf-marketing leader, daily after the 15th of the month
• Finance, for the Friday RCA
• Founder, weekly during sale months

Frequency

Daily during the back half of the month. Weekly mid-month.

Data you need

• orders — daily revenue and units
• targets — monthly target per product and overall
• Calendar metadata — which dates are sale events (BAU, RR, Diwali sale, BBD, etc.), which are public holidays
• Historical day-of-week multipliers (Sunday lift, Monday dip, etc.) per product

How to ask it

• "Where are we landing this month vs target? Decompose by product."
• "At current run-rate adjusted for remaining sale days and weekend mix, what's the projected month-end revenue?"
• "How much faster do I need to run for the rest of the month to hit target, by product?"

What you'll get back

A landing-projection panel:

Overall: ₹X.XX Cr projected vs ₹Y.YY Cr target — landing at -8% / +2% / etc.

The projection uses:

• Day-of-week multipliers from the trailing 6 months (Sundays typically +20-30% vs BAU)
• Sale-day uplifts from prior comparable sales (e.g. "Diwali sale +180%")
• Decay-adjusted trend for the trailing 7 days (so today's drop isn't blindly extrapolated)

The logic

For each product:
├── Compute MTD daily run-rate (last 7 days, calendar-adjusted)
├── For each remaining day in month:
│   ├── Tag day-type (BAU / Sunday / Sale / Holiday)
│   ├── Apply historical multiplier for that tag
│   └── Sum projected revenue
├── Add MTD actuals → projected landing
├── Compare to target → Δ
└── If Δ < threshold → compute required uplift to recover

How to make it recurring

Save as a tracker with:

• Cadence: daily, runs at 10 AM IST
• Threshold: alert if projected landing < 95% of target on any product with >10 days remaining
• Output: a one-line summary per product, ranked by gap

Pitfalls

• Wrong calendar tagging. If a sale day isn't in the calendar table, the projection misses the lift. Maintain the calendar table monthly.
• Treating thin samples as truth. Day 3 of a month is not a reliable basis for landing projection — wait until day 10-12 before trusting it heavily.
• Ignoring the recovery-pace number. The Δ tells you the gap; the recovery-pace number tells you what's actually required day-over-day. The recovery pace matters more — that's the actionable figure.
• Confusing projection with commitment. Projection is the model's best guess given current trend; it isn't a forecast you can promise to anyone. Use it to decide whether to spend more.

Related

• Channel budget scenarios — when the answer is "spend more"
• Marketing efficiency
• Product-lag RCA — when projection says you're behind