Until recently, I had misunderstood how AWS credits worked.
I assumed there was a fairly straightforward tier system. You start a company, build your platform on AWS and, as the company grows, you become eligible for higher levels of support.
That seemed logical to me. We were building on AWS. We wanted to work with AWS. I viewed it as the beginning of a long-term business relationship.
That isn’t quite how it works.
How access to the larger credit packages works
AWS does offer a route for bootstrapped companies. Under its current Activate programme, a self-funded startup can apply directly for $1,000 in credits, with selected companies potentially receiving up to $5,000.
The much larger packages—currently advertised as up to $200,000—require an organisation ID from an approved AWS Activate Provider, such as an accelerator, angel investor or venture capital firm. AWS says that eligibility is based partly on a startup’s funding stage and affiliation with one of these providers. (AWS Activate credit tiers)
That doesn’t necessarily mean every company receiving the larger packages has raised investment. But it does mean that, to reach those tiers, somebody outside AWS generally needs to have chosen you first.
Why this matters to bootstrapped founders
That is the part I have a problem with.
We are self-funded, but not necessarily because we sat down one day and decided that bootstrapping was the superior path. We have looked for support. We haven’t received it yet. Like plenty of startups, we are building the product, trying to get traction and attempting to make progress with the resources we have.
Then you discover that because nobody else has backed you, access to meaningful AWS credits is restricted as well.
The companies that have already received help—from investors, accelerators or funding bodies—become eligible for another substantial boost. The companies still trying to prove themselves get a much smaller allowance.
It feels like a second decision has been made about your company, based largely on the first one.
Why AWS designed it this way
I understand AWS’s side of this.
AWS cannot hand hundreds of thousands of dollars in infrastructure credits to anybody who fills in a form. Particularly now, somebody could create a company, launch a massive AI or model-training workload and burn through a serious amount of compute without building a viable business.
There has to be an eligibility process. There has to be some control over it.
AWS is also operating this programme at enormous scale. It says more than 350,000 startups have joined AWS Activate since 2013 and that it has issued more than $8 billion in promotional credits. (AWS Activate application guide) It is easy for me to say that an account manager should get to know every startup when AWS may be dealing with tens of thousands of companies like mine.
Where I think the system falls short
Still, I don’t think the current system is fair.
Rather than AWS making a deeper assessment of which bootstrapped companies are worth supporting, much of that judgement has effectively been outsourced. If an investor, accelerator or recognised startup organisation has backed or accepted you, AWS can use that as proof that you are worth the risk.
If they haven’t, your ability to access the higher tiers is severely limited.
That assumes self-funding is always a positive choice. Sometimes it is. Sometimes it is simply what is left after you have tried to get funding, grant aid or other support and have been told no.
Being told, in effect, “You chose to self-fund, so you don’t need our help” misses the reality of building an early-stage company.
What a better process could look like
I don’t pretend to have the perfect replacement.
I would have expected the process to involve our AWS account manager getting to know the company, understanding what we are building and looking at the evidence that we are making progress. Perhaps the account manager could make a case for the next level of credits, or have authority to approve a limited increase based on traction, technical need and responsible usage.
That would obviously create work for AWS, and I don’t know how easily it could operate at this scale.
But it would also create a genuine relationship.
If AWS helps my company at the beginning, it doesn’t mean I owe AWS anything forever. It does mean that, as our platform develops, we become more embedded in its ecosystem. Moving becomes harder. Trust develops because somebody at AWS took the time to understand what we were trying to build and helped us get there.
That seems commercially valuable to AWS, not just generous to the startup.
What this changes for us
Instead, this discovery has changed how I intend to build our platform.
I had not seriously expected to explore alternatives to AWS. Now I am going to look properly at Google Cloud, Azure and smaller European providers such as Hetzner and Scaleway. AWS is still ahead of many of these companies in important ways, in my view. This isn’t me pretending that every hosting provider offers the same thing.
But there are other reasons to consider a smaller European provider, and there are now good reasons for us to avoid becoming too dependent on AWS.
When I speak to Google Cloud and Azure, one of my questions will be very direct: how does your startup credit system work over time, and what happens if we remain self-funded?
Smaller providers may not have comparable credit programmes at all. That is fine. At least the commercial relationship is clear.
Most importantly, we will prioritise building our platform so that we can migrate away from AWS without having to dismantle the company’s technical foundations. That is sensible architecture anyway, but it has moved much higher up our list because of this.
AWS credits are presented as support for startups from their first prototype through to scale. In reality, the meaningful levels of that support are far easier to reach once another organisation has already validated you.
For a founder still trying to get that first break, that distinction matters.