strategy · hard
"Should Amazon enter food delivery?"
Should Amazon enter the food delivery market?
Amazon already answered this question. The candidate who knows what they decided, and can explain why it was the right call, scores higher than the one who invents a strategy from scratch.
What makes this question hard
The prompt sounds like “yes or no, food delivery is big.” The actual test is whether you can reconstruct the strategic choice Amazon already faced, apply a build vs. buy vs. partner frame, and commit to a position with a kill criterion attached.
Feasibility is not the question. In 2026, Amazon can build any logistics product. The questions are viability (can restaurant delivery generate margin at scale against a market where DoorDash holds 67% share and margins are wafer-thin?) and lovability (what does a Prime member actually want at 7pm, and is it a DoorDash clone or something Amazon already has the supply chain to deliver?).
One historical fact you must know before the framework: Amazon launched Amazon Restaurants in 2016 and shut it down in 2019. That attempt already answered “can Amazon build restaurant delivery?” The answer was no. Not because of logistics, but because restaurant supply and customer habit lock-in took DoorDash a decade to build. Any answer that ignores this reads as uninformed.
The build vs. buy vs. partner decision tree
For any market-entry question at Amazon, run these four questions in order:
- Do we have existing customers with an unmet need here? Prime members already use Amazon for grocery and household items. The 7pm meal need is adjacent but not served by the existing product.
- Do we have supply-side advantage? Amazon has Whole Foods stores, a same-day cold chain, and Prime Now infrastructure. It does not have restaurant relationships at scale and would need years to build them.
- Can we win unit economics without a decade of subsidy? Restaurant delivery: no. DoorDash spent years subsidizing orders to build habit. Grocery and prepared meals through owned infrastructure: yes, and the data is already in.
- Does partnership give us optionality at lower cost than building? Yes. And Amazon used it.
Amazon’s actual answer: partner for restaurant, build and own for grocery and prepared meals.
Structure a strong answer
strong
"The market-entry question already has a real answer: Amazon tried building restaurant delivery with Amazon Restaurants, shut it down in 2019, then chose partnership over a rebuild. That history is the frame. My recommendation: don't re-enter restaurant delivery as a DoorDash competitor. DoorDash holds 67% US market share, Uber Eats 24%, Grubhub 9%. The three-sided marketplace (customers, restaurants, drivers) took a decade and billions in subsidies to build. Amazon's right to win is not there.
Where Amazon does have right to win: grocery and prepared meals through Whole Foods and same-day infrastructure. Perishable same-day delivery grew 40x between January 2025 and end of 2025. Fresh groceries now make up 9 of the top 10 most-ordered items in same-day coverage areas. Amazon Now, the 30-minute delivery service, launched in Seattle and Philadelphia in 2025 and expanded to dozens more cities in 2026 at a $3.99 Prime member fee. The lovable wedge is not another three-sided marketplace. It's a Whole Foods prepared meal delivered in 30 minutes to a Prime member who is already in the app buying paper towels.
For restaurant delivery upside, the partnership already covers it. Amazon holds a 2% stake in Grubhub with an option to increase to 15% based on performance metrics including new Prime customers added. Prime members get Grubhub+ (a $120/year membership) at no extra cost. In May 2024, Grubhub ordering was embedded directly in the Amazon app. Over 90% of orders through that integration come from Prime members, which tells you this is a Prime retention instrument, not a food delivery play. The Grubhub partnership covers restaurant optionality at zero logistics build cost.
Kill criterion: if the Grubhub stake does not produce measurable Prime retention lift within two years, and Grubhub's market position continues eroding (which is the consensus forecast, given Wonder Group acquired it from Just Eat), let the 15% option expire and invest entirely in the grocery-to-prepared-meal arc. Success metric: prepared meal attach rate among Prime members who already grocery-shop on Amazon, tracked monthly against same-day coverage expansion."
weak
"Yes, food delivery is a huge market and Amazon has the logistics to compete with DoorDash and Uber Eats." This fails on three counts: it ignores that Amazon already tried and exited restaurant delivery in 2019; it treats logistics as the only moat when the real barrier is restaurant supply and customer habit lock-in built over a decade; and it offers no kill criterion or success metric. Generic market size plus generic asset listing gives the interviewer no signal about your decision-making.
What the interviewer is checking
Amazon strategy rounds are graded against Leadership Principles. Two apply directly here.
Customer Obsession means starting from what the Prime member actually wants at 7pm, not from “food delivery is a large market.” The strong answer identifies the prepared-meal need within the existing grocery context, not a restaurant app customers would have to download separately.
Invent and Simplify applies to the build vs. buy vs. partner choice. Amazon simplified by partnering on restaurant delivery and inventing on grocery (Amazon Now, perishable same-day at 2,300+ US cities and towns). Copying a competitor’s model at high cost would have been the opposite of this principle. Interviewers reward candidates who can name the principle and connect it to the actual decision.
One more thing worth naming: Grubhub (owned by Wonder Group after the Just Eat sale) is widely forecast to exit, merge, or be acquired within the next few years. That makes Amazon’s 15% stake option strategically significant independent of whether Grubhub remains a going concern. Amazon gets optionality on the most distressed player in a concentrated market without committing logistics infrastructure to a category with thin margins.
The viability lens
In 2026, “can we build it” is not a differentiating answer. Any well-resourced company can build a delivery app. The real question is whether restaurant delivery is a paid problem Amazon can win given the competitive structure, and whether grocery and prepared meals are the lovable wedge that extends what Prime members already do. Lead with that lens.
See proving viability for how to pressure-test market entry from the customer need outward, and the working backwards framework for the PR/FAQ discipline that would make this a real Amazon-style product decision.