Chapter 13 of 15 · Choosing Where to Compete
1 The bottom line
Attack where the leader's own business model, incentives or customers stop it from copying you. Then it cannot answer without hurting itself, and you get years to grow.
2 Key takeaways
- Pick a move that costs the leader more to copy than to ignore. A rival that must damage its own profits, brand or partners to follow will usually stay put.
- Oracle, IBM and the server makers had gross margins near 80 percent, and
Amazon Web Services1:34:50 targeted 20 to 40 percent, a business the incumbents would not build because it cannibalized them.
Pepsi1:36:40 sold twice as much cola for the same nickel in 1934, and Coke was locked into its small contour bottle.
- Oracle, IBM and the server makers had gross margins near 80 percent, and
- Start on ground the leader does not bother to defend. Neglected markets and customers give you time to learn before anyone reacts.
Sam Walton, Walmart19:35 got a roughly 10-year head start in small towns Kmart skipped.
Alfred Steel at Pepsi1:55:00 took Pepsi's share from the low 20s to 35% by 1955 by going after what Coke ignored.
- Fund the attack from a profit pool the leader does not have. A rival cannot match a price it can only reach by becoming a different company.

Google3:00:03 took Android from about 5% smartphone share in holiday 2009 to 80% by the end of 2013 by paying partners from search revenue.
- Build a system or a speed that rivals cannot copy quickly. Even a rival that wants to follow loses years rebuilding how it works.

Nvidia56:00 shipped a new generation every six months while competitors took 18 to 24 months.
Zara21:00 goes from detecting a trend to a store hanger in under 15 days.
- Expect the giant to answer with its own weapons, scale and distribution. Counterpositioning only protects you where the leader cannot follow, not where it can simply outspend you.
- When

Microsoft41:54 made Internet Explorer free inside Windows 95, Netscape stock fell about a third that day.
Bank of America34:30 absorbed about $2 million of fraud in its Fresno card pilot, a loss no other consumer bank could have survived.
- When
you never want to compete with someone bigger, more established, better capitalized than you on the exact same footing that they're on.
Joe Coulombe, Trader Joe's
2.1 Find what the leader cannot afford to copy
The best ground is where following you would hurt the leader's own revenue, brand or partners.
- Ask what the leader earns from the old way, because it will not fund the end of it. Source:
Thomas Edison49:25
- Edison pitched recording the human voice to Western Union, a telegraph monopoly, and they saw no conceivable use for it.
- The worst buyer for a product that makes another product obsolete is the party earning from the old one.
- James Dyson found vacuum makers sold hundreds of millions of dollars of bags a year and could not see his bagless machine.
- Sell something the leader does not sell, so you are not fighting on its footing. Source:
Joe Coulombe, Trader Joe's35:45
- Pronto Markets was a 7-Eleven clone with
no structural or strategic barriers29:55, financed by a dairy supplier that was then bought by 7-Eleven's owner. - Joe took on more debt for liquor licenses. Liquor was high value, protected by fair trade pricing, and 7-Eleven did not sell it.
- Pronto turned into mostly a liquor store, which bought a couple of years to build the bigger plan.
- Pronto Markets was a 7-Eleven clone with
- Aim at customers and partners the leader is locked into. Source:
Pepsi1:36:40
- Pepsi used recycled beer bottles to give 12 ounces for a nickel. Liquid costs almost nothing, so margins held.
- Coke and its bottlers could not respond without hurting the brand.
- By August 1941, non-Coke colas, mostly Pepsi, held 14% of the US soft drink market.
- Build the model your partner refuses to build, if they are protecting their own customers. Source:
Sam Walton, Walmart1:19:10
- Butler Brothers declined to run the back end of a discount chain, because cutting their markup for Walton would have wrecked relations with other franchisees.
- Walton believed the variety stores were doomed anyway and opened the first Walmart in Rogers, Arkansas, on July 2, 1962.
- That first store did $1 million in sales in its first year.
- Define your goal as the opposite of the incumbent's goal. Source:

Larry Page and Sergey Brin, Google31:54
- Portals earned from banner ads and page views, so a search engine that sent users away faster broke their model.
- Excite was ready to license BackRub for about a million dollars in 1997, but its CEO killed the deal after seeing better results made users leave sooner.
- Pick a rival whose stance stops it from joining you. Source:
Visa2:13:20
- Visa's marketing chief moved its positioning from MasterCard to American Express and told people to take Visa because Amex was not taken there.
- Amex declined the first global Olympic sponsorship at $14 million. Visa paid $17 million for the rights plus $23 million in media.
2.2 Start where the leader is not looking
Neglected places and customers let you learn and mature before the giant notices.
- Build your system in markets the leader thinks are beneath it. Source:
Sam Walton, Walmart19:35
- Kmart chased big metropolitan markets, while Walton learned in small towns where a Kmart store was unthinkable.
- Walton said Kmart let them stay out there while they developed, and that early attention would have been a disaster.
- His own words: "we were protected by our small town market."
- Win the weak fights first, then take on the giants. Source:
Sam Walton, Walmart32:55
- Charlie Munger compared Walton to a prizefighter who built a record against 42 weak opponents so he could reach the finals.
- An efficient system broke small-town merchants, and then Walmart went after the big boys.
- Walmart was 26 times the size of Jamesway yet had profit margins three times as high and sales per square foot of $210, nearly twice Jamesway's.
- List the customers and channels the leader has decided to ignore. Source:
Alfred Steel, Pepsi1:55:00
- Steel defected from Coke and built an all-Black sales team, pitched Pepsi as lighter, and embraced TV to court youth.
- Coke, comfortable and conservative, did none of these.
- He also brought Coke-style operating discipline to Pepsi's bottlers.
- Go local where the leader runs one national machine. Source:
Pepsi Challenge2:25:00
- Scully saw a Dallas bottler's taste-test ads lift share there, then bought camcorders and card tables so local bottlers could film real shoppers.
- Coke's one-site, one-sound national machine could not answer in local markets.
- The result: Pepsi's share rose every year from 1975 until New Coke in 1985, while Coke's fell.
- Keep quiet until you are strong, and prepare in the open where the giant does not read it as a threat. Source:

Eric Schmidt, Google1:57:57
- Schmidt held back a browser in 2001 ("I don't want to moon the giant in 2001"). Google funded Mozilla, paid for default search placement, and hired Firefox engineers into a group led by Sundar Pichai.
- Chrome began in 2006 and launched in September 2008 with Firefox veterans who had waited inside Google for years.

Microsoft2:45:00 urged Compaq and other clones to build 386 machines while publicly following IBM's 286 line, which IBM chose partly to keep PCs away from its mainframe work.
2.3 Fund the attack from a different profit pool
When your money comes from somewhere else, you can charge less than the leader can ever match.
- Give away what the leader charges for, and let another business pay for it. Source:

Google and Android3:08:20
- Microsoft charged a royalty on Windows Mobile, so even five dollars a handset was a huge gap against zero in a low-margin business.
- Google gave Android away as open source and shared search revenue with carriers and handset makers. Bill Gurley called it the

less than free business model3:00:03. - Matching it meant Microsoft becoming an advertising company, which it would not do. Palm and Blackberry could not compete either.
- Skip the feature fight and attack with something the leader's product cannot do. Source:

Google vs Microsoft Office48:09
- Office was protected by its feature wall, file-format lock-in and the enterprise agreement, so Google used the web's ability to let several people edit one document at once.
- Search profits on infrastructure Google already had paid for it.
- Microsoft was forced to bring its crown jewels to the web at great cost. Its productivity segment still made over $120 billion last year, against about $50 billion for all of Google Cloud.
- Price the new model on a margin the incumbent's annuity cannot survive. Source:
Amazon Web Services1:18:40
- Incumbents earned through licenses, audits and paid upgrades. Meter-based pricing and always-current software threatened that annuity.
- James Hamilton, then at Microsoft, needed no proposal, RFP, vendor negotiation or data center space to try S3. His first month's bill was $3.08.
- AWS operating margin was 19.2 percent when first reported in 2015 and about 30 percent later. Dropbox, Instagram, Airbnb and Uber were built on the model.
- Offer a deal structure the leader's roster stops it from matching. Source:

Phil Knight, Nike2:17:59
- With no leverage in 1984, Nike offered Michael Jordan $2.5 million over five years plus a 5% royalty on gross sales of the whole Air Jordan line.
- Converse and Adidas had too many stars and too much to lose to match an upside deal.
- Air Jordan sold $126 million in year one against an internal goal of $3 million over three years.
- Rob Strasser also

signed college coaches1:52:00 so whole teams wore Nike, and had UNLV, Georgetown, Texas, Arkansas and Iona within a month.
- Choose a role that your customers' other suppliers cannot take. Source:
Morris Chang, TSMC1:20:29
- TSMC lets customers like Apple take the credit on stage, because its job is not to market.
- Chipmakers that rented spare capacity before TSMC often demanded the right to market the customer's products under their own brand.
2.4 Build a system rivals cannot copy quickly
Speed, logistics and a different kind of store make you hard to follow even when the leader sees you coming.
- Turn speed into a permanent habit. Source:

Jensen Huang, Nvidia9:05
- Near bankruptcy forced Nvidia into speed, and it made the emergency process permanent starting in 1997 and 1998. It roughly doubled performance at a given price every six months.
- Rivals knew what would be in the next chip but could not ship it as fast.
- The cadence: GeForce 256 in fall 1999, GeForce 2 in spring 2000, GeForce 2 Ultra in fall 2000, GeForce 3 in spring 2001, Ti-500 six months later.
- Build the biggest product at a cost point rivals will not accept. Source:

Jensen Huang, Nvidia6:26
- For Riva 128, Nvidia implemented everything in DirectX it knew of, used the fastest memories, and pushed the bus and frame buffer to what physics allowed.
- The logic: if you build it as large as possible, nothing can be faster.
- Organize the whole company around responding to customers. Source:
Amancio Ortega, Zara21:00
- Ortega built a logistics and distribution system that responds immediately.
- Traditional companies make the whole season's collection at once, while Zara keeps modifying products according to what people ask for.
- An analyst in the book called Zara a groundbreaking textile distribution company at heart.
- Invent the operation the incumbent's incentives keep it from building. Source:

Jeff Bezos, Amazon1:29:42
- Amazon had to invent distribution for unique orders to millions of customers, a different combinatorial problem than shipping trucks to stores.
- Barnes & Noble's incentives pointed away from rebuilding distribution for a small online business.
- Per Brad Stone, the Riggios would not put their best people on an effort that cannibalized profitable stores.
- Leave the price war by changing what the customer can verify. Source:
Ingvar Kamprad, IKEA34:35
- Mail order rivals bought from the same manufacturers and undercut each other on identical items. Prices and quality fell and complaints rose.
Kamprad opened a showroom38:20 in March 1953, advertised it for months and offered free coffee and morning buns, so customers could see quality for themselves.- More than a thousand people queued on opening day, and within a couple of years about half of catalog subscribers had made the trip.
2.5 Plan for the giant's own weapons
Once you matter, the leader will use scale, distribution and patience, so do not fight where those win.
- Do not rely on a distribution channel the giant controls. Source:

Microsoft vs Netscape41:54
- Netscape had a $10 billion market cap, but its revenue was server software sold on the strength of its browser share.
- On December 7, 1995 Microsoft made Internet Explorer free with Windows 95 and signed AOL, CompuServe and Prodigy.
- Internet Explorer went from under 20% share in 1996 to roughly 100% by 2000, and AOL bought Netscape for just over $4 billion in stock in 1998.
- Do not enter a war that the rival's balance sheet can outlast. Source:
Bank of America34:30
- In 1958 it mailed unsolicited cards to 65,000 Fresno customers and chose to expand rather than pull the rip cord. The program was profitable by 1961, and it kept that quiet while rivals read the press and abandoned their plans.
- 22% of credit was delinquent or defaulted, yet the bank had 2 million California cardholders and 20,000 merchants in year one. Only 10 new credit cards launched in the US from 1960 to 1966.
- Read the whole board, not just the price cut in front of you. Source:
Jay Gould27:51
- Vanderbilt cut cattle shipping to a penny a head to hurt Gould's Erie and was laughing as his trains filled with steers.
- Gould had cornered the cattle market and profited from Vanderbilt's shipping losses.
- Gould was in his early 30s, a virtual unknown, against the richest man in America.
- Answer a copied price cut by telling the truth and changing the offer. Source:
Herb Kelleher, Southwest24:54
- Braniff copied Southwest's $13 fare. Southwest ran ads saying Braniff was trying to put it out of business.
- Customers could take the $13 fare or pay $26 and get a free bottle of Scotch, whiskey or vodka, which suited business travelers charging expense accounts.
- 76% of passengers paid the $26 fare and took the bottle. Two years later Braniff dropped the Dallas to Houston Hobby route.
- Assume competitors will always come, and design what they cannot copy. Source:
Joe Coulombe, Trader Joe's1:43:57
- After California repealed fair trade in 1977, Bevmo, Total Wine and discounters slashed wine and liquor prices. Joe doubled down on private label and being one of one, and stopped watching nearby supermarkets.
- His rule: "The answer is to design a store that has no competition." Trader Joe's stayed profitable while many retailers failed.
- In 1970 he had
rejected the race on size50:00, where only one or two chains survive, and chose discontinuous products rivals could not copy.
2.6 Where they differ
A gap that the leader cannot cross can close, and not every win comes from counterpositioning.
- The incumbent may wake up, so the window is limited. Source:

Sundar Pichai, Google2:35:03
- Google shelved a chatbot built on the transformer because direct answers cut ad revenue, and

Noam Shazeer left in 20212:23:26. - After ChatGPT, Pichai reframed AI as a disruptive threat and shipped AI Overviews for some queries and AI Mode for some users, while leaning on video and image models that do not touch search revenue.
- By May 2023 Google was announcing Gemini and AI Overviews at I/O, and revenue stayed at an all-time high.
- Google also

failed to treat the transformer as a platform shift2:03:42, and all eight authors left within a couple of years.
- Google shelved a chatbot built on the transformer because direct answers cut ad revenue, and
- A slow giant can still reorganize and catch up. Source:
Microsoft, Azure1:39:00
- The Windows and Windows Server groups protected license revenue, so cloud ideas were absorbed in internal fights. For about five years new projects drifted to AWS.
- Ballmer split Azure off and moved Satya Nadella in. Azure now has about 21 percent of cloud against AWS at about 39 percent.
- Amazon also fought the last battle with Redshift, copying an Oracle-style warehouse, so
Snowflake2:02:10 grew to about 50 billion dollars storing data on AWS.
- An incumbent can turn its own weight into the moat. Source:
Frank Blake, The Home Depot4:15:00
- Lumber and drywall cannot ride an Amazon van, and a project customer who runs out of grout on Sunday wants it now. Blake built rapid deployment distribution centers in 2009 so store pickup became a large share of e-commerce.
- Today 90% of US homes can get a million SKUs from Home Depot within 2 to 24 hours, and the setup paid off when COVID hit.
- Sometimes the shift does the work, not your tactic. Source:
John D. Rockefeller, Standard Oil1:27:43
- New oil in Texas, California, Russia and the Middle East, plus kerosene losing ground to electricity and cars creating a gasoline market, eroded the monopoly without regulators.
- Standard's refining share was 64% in 1911 against about 90% a decade earlier, with at least 147 other refiners, 21 years after the Sherman Act.
Sources
- Competition, Intensity & Building a $115 Billion Global Bank | Nik Storonsky, Revolut ·
David Senra - The Tycoons: Carnegie, Rockefeller, Gould & Morgan ·
Founders - How Sam Walton Worked ·
Founders - The Home Depot: How America's aging houses built the best stock of the last 45 years ·
Acquired - Inside Bending Spoons: Finding Talent, Leveraging AI & Driving Operational Excellence | Luca Ferrari ·
David Senra - Excellence Is the Capacity To Take Pain | Travis Kalanick, Founder of Uber ·
David Senra - DHH: How to Build a Profitable Company Without Losing Control ·
David Senra - The $44 Billion Company Automating Your Company's Finances ·
David Senra - The Company Apple Couldn't Kill | Spotify Co-CEO Gustav Söderström ·
David Senra - Evan Spiegel, Snapchat: Building a Multi-Billion Dollar Company ·
David Senra - John Mackey: The Dropout Who Changed How America Eats ·
David Senra - Coca-Cola: The Complete History & Strategy ·
Acquired - Trader Joe’s: Hawaiian shirts and counter positioning ·
Acquired - Michael Dell, Dell Technologies | David Senra ·
David Senra - Google Part III: The AI Company. Google is amazingly well-positioned... will they win in AI? ·
Acquired - Google Part II: Alphabet ·
Acquired - Google Part I: Origins of Search. How the Best Business in Human History Happened ·
Acquired - TSMC ·
Acquired - Amancio Ortega: The Genius Behind The Inditex Group ·
Founders - The Founder of IKEA: Ingvar Kamprad ·
Founders - IKEA ·
Acquired - Microsoft Volume II: The Complete History and Strategy of the Ballmer Years ·
Acquired - Microsoft Volume I: The Complete History and Strategy of founding through Windows 95 ·
Acquired - Herb Kelleher: Founder of Southwest Airlines ·
Founders - Visa ·
Acquired - NVIDIA CEO Jensen Huang ·
Acquired - The Biography of Thomas Edison ·
Founders - Nike ·
Acquired - Amazon Web Services ·
Acquired - Amazon.com ·
Acquired - Walmart ·
Acquired - Nvidia: The Machine Learning Company (2006-2022) ·
Acquired - Nvidia: The GPU Company (1993-2006) ·
Acquired - Standard Oil Part II (Extended Cut) ·
Acquired
Part of Choosing Where to Compete in the growth library.