Trajectory, not snapshot: how to read a new entrant before it eats your lunch

The wrong axis is the snapshot

The dismissal in every case has the same shape: compare the attacker's current feature set to the incumbent's current scale, conclude the attacker is small, move on.

Three things this comparison misses.

It misses incentives

Amazon's logistics network is not a standalone profit center. It lowers Amazon retail and marketplace cost. It increases the value of Prime. It supports ads and data revenue elsewhere. The same dollar of logistics investment shows up in multiple Amazon P&Ls. FedEx's logistics network has one P&L to feed. That asymmetry is not visible in a feature comparison; it is visible in a margin tolerance comparison. Amazon can price more aggressively, invest longer, and accept lower per-shipment margins because the payoff is downstream and elsewhere.

It misses trajectory

A feature checklist captures where the attacker is today. It does not capture where the curve points. Netflix in 2003 was a DVD-by-mail service. It was a real business with a clear ceiling and an unimpressive feature set against Blockbuster's stores. By 2007 it had started streaming. By 2013 it had original content. None of that was in the 2003 snapshot. The incumbents who looked at the 2003 snapshot and concluded "small, niche, not us" were technically right about the snapshot and structurally wrong about the curve.

It misses option value

The attacker does not need to win the incumbent's whole business. The attacker needs to peel off the customer segments where the incumbent's scale and end-to-end-ness do not matter. Mid-market DTC brands shipping a few thousand orders a month, sellers already in the Amazon orbit, companies whose CFOs want one logistics invoice instead of five: these are the easy first wins. By the time they add up to a number that matters in the incumbent's P&L, the relationship is already established.

The right axis is three layers

This is the framework we use with clients when the question on the table is "is this new thing a real threat or just hype." Three layers, in order.

1. Incentives and economics

Start with one question: what is the attacker actually optimized for?

Most competitive analysis treats the attacker's product as the thing to score. Treat the attacker's incentives as the thing to score. Where does their profit really come from? What does their service do for adjacent revenue lines they own? Where do they have cross-subsidies you do not? An attacker whose product is also a customer-acquisition flywheel for a larger business has different margin tolerance than a standalone competitor. That asymmetry is the most reliable predictor of how aggressive they can be on price and how patient they can be on time.

This is where the FedEx read gets fragile. Amazon's logistics product is not its main business. FedEx's logistics product is. The asymmetry is structural, not strategic.

2. Trajectory, not snapshot

The question is not "can they do X today." The question is "if they keep going, where does the curve point."

Map three time horizons honestly:

• 1 to 3 years. What happens if they stay in their current lane and execute. Conservative case, easy to model, often dismissed because it looks small.
• 3 to 5 years. What happens if they add one adjacent capability each year. This is the lane most attackers actually walk; one capability per year compounds faster than the incumbent expects.
• 5 to 10 years. What happens if they reach parity or exceed you on the dimensions that actually matter to your customers (not your features, your customers' job-to-be-done).

You do not need to predict any of these. You need to know which scenario triggers which response inside your business. Without trigger points, the trajectory work is just speculation. With trigger points, it is decision support.

3. Options: defend, partner, leapfrog

The output of a feasibility study is not a slide; it is a set of concrete strategic options. Three shapes recur.

1. Defend. Tighten the value proposition where you are structurally advantaged. Make it harder for the attacker to peel off your best customers. Invest in efficiency on your side of the line so the price gap narrows. Defense is not "do nothing"; defense is "make the place where you are strongest harder to attack."
2. Partner. Treat the attacker as infrastructure instead of competition. Build a layer above it. Vertical specialization, white-glove service, regulated-industry compliance, integration with workflows the attacker will not build. The attacker becomes a vendor, not a rival. This is what plenty of companies did with AWS once they realized the question was not "how do we beat AWS" but "what do we build on top of AWS."
3. Leapfrog. Use the moment to make a bigger move. New product, new business model, new geography, new customer segment. The attacker's arrival creates cover for changes that would have been unthinkable a year earlier; the board is suddenly open to options it would have killed in calmer water. Operators who wait until the threat is undeniable miss the window to leapfrog.

What this article is not

• Not a prediction about Amazon Supply Chain Services. We do not know whether ASCS will reshape global logistics, plateau at a useful 3PL, or become quietly internal again. None of that is the point.
• Not a critique of Raj Subramaniam personally. His read on his own business might be right. The "completely different" framing is doing what incumbent communication is supposed to do: settle the market. That is a different job than the strategic-feasibility job, which has to actually answer the question for the board.
• Not a sales deck for a specific SDS engagement. The framework above is how we think, not what a feasibility study looks like in your specific market. The numbers, the trigger points, the options analysis: those are scoped to your business, not to an example in someone else's industry.
• Not a fit assessment. We do not take every engagement, and the fit conversation is the first thing we have.

One-sentence takeaway

When a new entrant lands in your market and the incumbent's instinct is to say "that's not what we do," the work to do is not feature comparison; it is incentives, trajectory, and options, with concrete trigger points the board can act on.