WillaWave Would the weather have changed your flea and tick order?
Replay a real Omaha season week by week. The normal plan orders to the usual seasonal curve. The weather-adjusted plan scales that curve by the last four weeks of weather: warm spells raise it, more so when they're wet, and the first fall frost cuts it. Move the sliders to see when adjusting for weather pays off, and when it doesn't.
Weather is real (NOAA daily highs, lows and rain, Omaha Eppley Airfield). Sales are hypothetical: a made-up product with a typical flea and tick season.
Your ordering setup
Weeks from placing an order to having it on the shelf.
Any order is at least this many units. 0 means no minimum.
How much weather moves sales
Two numbers. The first is how much weather actually moves sales of this made-up product. The second is how much the buyer believes it does. The closer the guess, the better the weather-adjusted plan does.
Change in weekly sales when the past month runs unusually warm: about 4 °F above normal in midsummer, 6 to 10 °F in spring and fall. Rain makes a warm spell count for more, and the first fall frost pulls sales down. Sales lag the weather by about a week. This setting produces the hypothetical sales.
What the weather-adjusted plan believes. At 0% the buyer ignores the weather and both plans order the same.
2012: how the two plans did
The normal plan ran short in 6 weeks, losing 76 sales; the weather-adjusted plan never ran short. The weather-adjusted plan committed $1,325 more cash and ended the season with 23 units fewer left over.
| Outcome | Normal plan | Weather-adjusted plan |
|---|---|---|
| Units ordered | 2,925 | 2,978 |
| Cash committed | $73,125 | $74,450 |
| Weeks with unmet demand | 6 | 0 |
| Units short (lost sales) | 76 | 0 |
| Left over at season end | 50 units, $1,250 | 27 units, $675 |
| Week of | Demand | Normal: ordered | Normal: cash | Adjusted: ordered | Adjusted: cash |
|---|---|---|---|---|---|
| Mar 5 | 21 | 11 | $275 | 47 | $1,175 |
| Mar 12 | 28 | 58 | $1,450 | 77 | $1,925 |
| Mar 19 | 44 | 74 | $1,850 | 102 | $2,550 |
| Mar 26 | 62 | 87 | $2,175 | 129 | $3,225 |
| Apr 2 | 86 | 111 | $2,775 | 138 | $3,450 |
| Apr 9 | 106 | 109 | $2,725 | 210 | $5,250 |
| Apr 16 | 135 | 131 | $3,275 | 77 | $1,925 |
| Apr 23 | 124 | 124 | $3,100 | 37 | $925 |
| Apr 30 | 98 | 113 | $2,825 | 152 | $3,800 |
| May 7 | 115 | 125 | $3,125 | 94 | $2,350 |
| May 14 | 110 | 120 | $3,000 | 146 | $3,650 |
| May 21 | 124 | 129 | $3,225 | 133 | $3,325 |
| May 28 | 125 | 130 | $3,250 | 65 | $1,625 |
| Jun 4 | 109 | 109 | $2,725 | 124 | $3,100 |
| Jun 11 | 115 | 115 | $2,875 | 80 | $2,000 |
| Jun 18 | 103 | 103 | $2,575 | 78 | $1,950 |
| Jun 25 | 95 | 95 | $2,375 | 186 | $4,650 |
| Jul 2 | 125 | 120 | $3,000 | 141 | $3,525 |
| Jul 9 | 132 | 127 | $3,175 | 161 | $4,025 |
| Jul 16 | 145 | 135 | $3,375 | 159 | $3,975 |
| Jul 23 | 147 | 135 | $3,375 | 139 | $3,475 |
| Jul 30 | 150 | 120 | $3,000 | 103 | $2,575 |
| Aug 6 | 134 | 119 | $2,975 | 105 | $2,625 |
| Aug 13 | 124 | 106 | $2,650 | 63 | $1,575 |
| Aug 20 | 103 | 83 | $2,075 | 39 | $975 |
| Aug 27 | 82 | 57 | $1,425 | 56 | $1,400 |
| Sep 3 | 77 | 52 | $1,300 | 44 | $1,100 |
| Sep 10 | 64 | 44 | $1,100 | 55 | $1,375 |
| Sep 17 | 58 | 38 | $950 | 21 | $525 |
| Sep 24 | 47 | 27 | $675 | 17 | $425 |
| Oct 1 (first frost) | 38 | 18 | $450 | 0 | $0 |
| Oct 8 | 8 | 0 | $0 | 0 | $0 |
| Oct 15 | 6 | 0 | $0 | 0 | $0 |
| Oct 22 | 7 | 0 | $0 | 0 | $0 |
| Oct 29 | 4 | 0 | $0 | 0 | $0 |
How it works: the weather index is the last four weeks' warmth against normal, boosted when it's also wet, and pushed down from the first fall frost. Weekly demand is the curve × (1 + the effect × last week's index), and both plans order up to their forecast for the lead time plus one week. Read the model.
Would this change your order?
If your demand moves with the weather, I build decision tools like this one on your own sales history.