AWS does not segment customers by size. It segments by business model, by technical maturity and by how the company creates value with technology — which is why a mid-sized fintech and a large manufacturer can sit in different segments despite similar revenue, and why the same product gets sold to each in a completely different language. For anyone selling alongside AWS, knowing which segment a customer sits in determines the pitch, the programme funding available, and who at AWS will engage.
Why size-based segmentation misleads
The instinct is to sort by revenue or headcount, because that is how most vendor territories are drawn. It produces the wrong conversation. A digital-native company with two hundred employees may run more infrastructure, spend more, and move faster than a traditional enterprise with five thousand — and the buying process in each is structurally different.
The segmentation that AWS actually uses reflects how the customer relates to technology: whether it is the product, the delivery mechanism, or the back office. That relationship determines the buying cycle, who signs, what the objections are, and which incentives apply.
The segments in practice
Startups. Technology is the product. The decision maker is technical and often the founder, the cycle is short, and cost sensitivity is acute but temporary — the constraint is runway, not policy. Programme support exists specifically for this group, and it is substantial enough to change the economics of an early deployment.
Digital natives. The business was built on cloud and has scaled past the startup stage. Technically sophisticated, price-aware in a sophisticated way — they optimise rather than economise. The conversation here is about architecture and unit economics, not about whether cloud is appropriate.
Independent software vendors. They build software that other companies buy, so their cloud costs are inside their own margin. Every conversation with this segment is ultimately about their gross margin and their customers’ procurement requirements — including where data can sit.
Enterprise. Large organisations, often with existing on-premise estates and a governance process to match. Long cycles, multiple stakeholders, and a strong preference for evidence over demonstration. The decisive factors are risk, compliance and the migration path, not feature comparison.
Small and medium business. Cloud is infrastructure, not strategy. Nobody wants to discuss architecture. The buying trigger is a concrete operational problem, and the solution has to arrive as a package rather than a project.
Public sector. Procurement rules dominate everything, timelines follow budget cycles rather than business need, and compliance requirements are non-negotiable. Selling here is largely about navigating process correctly.
Migration and modernisation. Not a customer type so much as a motion: organisations moving existing workloads. The conversation is about the current cost of running things as they are, and the risk of the transition.
What changes when you know the segment
The same service gets described three different ways. To a digital native, an object store is durability and cost per terabyte at scale. To an enterprise, it is survival of a data-centre failure and an answer to the audit question. To an SMB, it is not having to think about backups again.
Programme funding differs by segment as well. Support that transforms the economics for a startup does not exist in the same form for an established enterprise, and enterprise migration incentives have no equivalent for a small business. Pitching a programme that does not apply to the segment costs credibility immediately.
The practical test: before the first meeting, decide which segment the customer belongs to based on how they create value — not on their revenue. If the answer is unclear, that ambiguity itself is useful information about their maturity.
Where partners fit
Segments differ in how much they need a partner rather than the vendor directly. Digital natives and startups often prefer to build themselves and want a partner only for specific gaps. Enterprises and public sector buyers need a partner for the process as much as the technology. SMBs need someone to own the outcome entirely.
Reading this correctly determines whether the proposal should offer expertise, execution, or a managed result — and offering the wrong one is the fastest way to lose a deal that was otherwise winnable.
Key points
- Segmentation follows business model and technical maturity, not revenue or headcount.
- Each segment has a different decision maker, cycle length and dominant objection.
- The same service must be described differently per segment — durability, risk, or simplicity.
- Programme funding is segment-specific; citing the wrong one damages credibility.
- Segments differ in whether they want expertise, execution, or a fully managed outcome.
Can a company sit in two segments?
Yes, and it happens most often with software vendors that also run substantial internal infrastructure, or enterprises with a digital-native subsidiary. When it does, the buying process usually follows one segment and the technical conversation the other — worth separating explicitly.
How do you identify the segment before the first meeting?
Ask how the company makes money and what breaks if the technology stops. If the product itself disappears, it is a startup or digital native. If delivery stops but the business continues, it is enterprise or SMB. The answer takes one question and reframes the entire pitch.
Does segment affect pricing?
Not list pricing, but it affects which programmes, credits and commercial constructs are available, which changes effective cost significantly. It also affects what the customer considers expensive, which matters as much as the number.
Which segment is hardest for a partner to serve?
Traditional enterprise, because the barrier is trust and process rather than technology, and both take time that cannot be compressed. It is also where partner involvement adds the most value, which is why the cycle length is worth accepting.
Материал обновлён 5 сентября 2026