Project Kansas
Pepsi is winning the sip. We can win it back.
Competitive and market analysis · Atlanta · Late 1984 · Confidential
vs
01 / The question
We outspend Pepsi by $100 million. Why are we losing?
- Ad spend vs Pepsi
- +$100M
- Vending machines
- Twice as many
- Coke soft-drink growth, 1976 to 1978
- 13% to 2%
This analysis answers one question
Are we losing on the brand, or on what's in the can?
If it's the can We change the formula. First change in 99 years.
02 / The market
Our lead fell from 6.8 points to 2.9 in nine years.
Our lead, 1984
2.9pts
Lead cut by more than half. And in supermarkets:
-1.7 Pepsi now outsells us in the grocery store.
Share of all US soft drinks. Mid-1950s: Coke outsold Pepsi better than 2 to 1.
03 / Who competes, and on what
We win every lever we pay for. We lose the sip.
| Where we compete | Coke | Pepsi |
|---|---|---|
| Advertising | $100M more | Less |
| Vending machines | Twice as many | Half |
| Fountain and restaurants | Dominant | Trailing |
| Shelf space | More | Less |
| Image with the young | Heritage | The Pepsi Generation |
| Blind taste test | Loses | Wins, on TV, since 1975 |
04 / What customers value, and how we measured it
Customers told us what they want: a sweeter sip.
- Tasters
- ~200,000
- Research budget
- $4 million
- Time
- 2 years
- Format
- Blind, one sip
Take the test: one sip, then pick a cup.
55% of 191,000 tasters picked the sweeter cup over Coke and Pepsi.
05 / Strengths and weaknesses, theirs and ours
The SWOT says change the formula. One box is empty.
S Strengths
The machine
Fountain, twice the vending machines, $100M more advertising.
W Weaknesses
Loyalty is leaking
Coke-only drinkers fell from 18% (1972) to 12% (1982).
O Opportunities
A sweeter formula
Beats Pepsi and our own Coke with 55% of 191,000 tasters.
T Threats
The Pepsi Challenge
Pepsi-only drinkers rose from 4% to 11%. Pepsi leads in grocery.
F What customers feel
Not measured
10 to 12%
of focus-group tasters were angry at the very idea of a change, and said they might quit Coke. The survey had no box for them.
06 / What the data can't tell us
We tested a sip. People drink a can.
- A sip is not a canThe sweeter drink wins a single sip. That edge fades as you keep drinking.
- A vote is not a lossTasters were never told that picking the new one meant losing the old one.
- 200,000 is not 200,000Only 30,000 to 40,000 of them tasted the final formula.
Observation room · One-way glass
Winner
The one you grew up with
Illustrative. Never measured.
07 / Implications
If the sip decides, the formula is the lever.
- SoMoney won't close a taste gapWe already outspend Pepsi by $100 million.
- SoThe new taste beats theirsAnd ours: 55% of 191,000 in blind tests.
- SoReplace, don't addThe new taste becomes Coke. The old one retires.
99
08 / Recommendation, and how we'll test it
Launch the new taste April 23, 1985. The hotline is our test.
- Launch
- April 23, 1985
- Watch: hotline
- 400 calls a day
- Watch: share
- 21.7%
Hotline calls a day1,500
Day79
Test failed
- Apr 23New Coke launches
- May 28Old Cola Drinkers of America forms
- JuneHotline at 1,500 calls a day
- Jul 11Coca-Cola Classic returns
09 / What the analysis missed
Test the whole can. Then ask what losing it would feel like.
"The deep and abiding emotional attachment"
What the research missed, per Coca-Cola president Donald Keough, July 1985
- 79 days
- From launch to Classic's return
- 3%
- New Coke's share once Classic was back