- How to choose a bug bounty platform in six steps
- Open bounty and researcher community platforms
- On-demand pentesting and Web3 bounty platforms
- Open programmes, private invitations and pentesting on demand
- Triage: who reads the reports first
- Smart contracts and Web3 programmes
- Getting ready for the first report
- What a programme costs to run
- Where I'd start
Bug bounty platforms connect you with outside security researchers who look for vulnerabilities in your systems and report them for a reward. Choose between an open crowdsourced programme, vetted on-demand pentesting and a platform built for smart contracts, then decide whether you want the platform's own team to triage reports before your engineers see them.
What happens on the morning your programme goes live and forty reports arrive, half of them duplicates and a few of them describing the same missing header in slightly different words? That's the question to answer before you pick a platform, because the platform decides who reads those reports first, which researchers are allowed in and what kind of code they're looking at.
Here's the order I'd work through it in.
How to choose a bug bounty platform in six steps
- Write down what's in scope: the domains, apps, APIs and contracts you're happy to have tested, plus anything that's off limits, such as a payments partner's systems.
- Decide on the model: an open programme where researchers pick their targets, a private programme for invited researchers, or pentesting on demand with a start and finish date.
- Work out who'll read incoming reports, and whether you need the platform's triage team to filter and validate them before they reach your engineers.
- Agree how you'll rate severity and what each level is worth to you, starting from a published rating scheme if you don't already have your own.
- Match the platform to the code you run: general web, mobile and APIs, or smart contracts and on-chain code, which need researchers with different skills.
- Line up the fix side before launch: who patches, how quickly, and who tells the researcher the issue is closed.
Steps one and two narrow the shortlist on their own. Steps three to six are where programmes quietly fail, so don't skip them because the platform demo looked tidy.
Open bounty and researcher community platforms
HackerOne
HackerOne runs bug bounty programmes alongside other exposure management work, and calls itself a leader in continuous threat exposure management.
Bugcrowd
A crowdsourced cybersecurity platform that publishes its own Vulnerability Rating Taxonomy and talks of moving towards preemptive security.
Intigriti
For a bounty with someone else filtering the noise, Intigriti is a bug bounty platform offering managed triage options.
YesWeHack
An offensive security and exposure management platform with bug bounty inside it, built on the idea of turning testing into evidence.
Standoff 365
Standoff 365 is a bug bounty platform built around researchers earning rewards once a vulnerability they found is confirmed.
Zerocopter
Billed as a one-stop shop for security needs, Zerocopter is a security services platform powered by its community of researchers.
On-demand pentesting and Web3 bounty platforms
Synack
Penetration testing that pairs AI with on-demand researchers, along with vulnerability management, integration and reporting, rather than an open bounty.
Cobalt
Cobalt is an offensive security platform built around continuous pentesting, which places it on the pentest-as-a-service side of the field.
Immunefi
Crypto and smart contract projects are the focus: Immunefi runs bug bounties and audit competitions, with managed triage on offer.
HackenProof
Pay-for-results bounty programmes, penetration testing and triage, offered to Web2 and Web3 teams alike.
Open programmes, private invitations and pentesting on demand
An open programme is the classic bounty: you publish a scope and a reward table, and researchers choose what to look at. You get breadth. You also get volume, and you don't choose who turns up. That's fine for a public website with a mature security team. It's a lot to absorb for a small company with one engineer who also runs the deploys.
A private programme narrows the door to researchers you've invited, which keeps the volume manageable while you learn how your team handles reports. I'd start here and open up later. It's far easier to widen a programme than to explain why you've suddenly closed one.
Pentesting on demand sits at the other end. You get a defined test, a set of researchers and a report, rather than an open-ended stream. Synack, for example, describes AI and human penetration testing with access to on-demand researchers rather than an open bounty. If what you need is a test before a launch or an audit, this model fits that job better than a bounty does.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
Triage: who reads the reports first
This is the part that's easiest to underrate. A report isn't a finding until someone has reproduced it, checked it isn't a duplicate, confirmed it's in scope and given it a severity. On a busy programme that's a job in itself, and if it lands on your developers it'll eat the time they should be spending on fixes.
Some platforms sell triage as part of the service, so your team reads validated reports rather than raw submissions. Intigriti, for instance, describes managed triage that turns higher report volume into validated findings. Ask any platform you're considering exactly what its triage covers: reproduction, de-duplication, severity scoring, or all three. Then ask what happens when you disagree with a severity call, because you will.
A shared severity scale saves arguments. Bugcrowd publishes its Vulnerability Rating Taxonomy, and even if you don't use that platform, reading a published scheme will help you write your own reward table in terms researchers already understand.
Smart contracts and Web3 programmes
Code that holds or moves money on a chain is a different target from a login page. Once it's deployed, a flaw can be exploited in public, and I'd want researchers who spend their time on contracts rather than web apps. If your product is a protocol or a set of contracts, look at platforms built for that work first.
Two models show up here. A bug bounty runs against live code over time, while an audit competition puts a fixed piece of code in front of researchers for a set period. Immunefi lists audit competitions and bug bounties as separate options, which is a useful way to think about the choice: audit before launch, bounty after.
If you run both a web app and contracts, check whether one platform covers both or whether you'd be better running two programmes with two scopes.
Getting ready for the first report
A bounty finds problems. It doesn't fix them. Before you launch, make sure the rest of your security basics can stand up to people looking hard. Leaked keys and tokens are exactly what an outsider can stumble on, so it's worth getting a handle on secrets management first, and if your product is mostly an API, a pass with dedicated API security tools will clear the obvious issues before researchers are rewarded for finding them.
Write your scope page as if the reader is in a hurry and slightly sceptical. List what's in, what's out, how to report, and what researchers can and can't do while testing. Then name the person who'll answer questions. Don't leave a researcher's question sitting unanswered for a week.
What a programme costs to run
Pricing differs by vendor and by scope, so ask each one for a quote and read its pricing page closely. There are two budgets to plan for, and it's worth asking how each platform handles both: what you pay the platform, and what you pay researchers in rewards. HackenProof, for one, describes its bounty programmes as pay-for-results.
Your reward table is the part you control. Set it by severity, decide how much you'll put aside in a quarter, and agree in advance what happens if that pot runs out early. A programme that pauses because the budget's gone does more damage to your standing with researchers than a modest reward table would have.
Do you make bug bounty or pentesting platforms? Tool makers can ask to be considered for a featured place on this list.
Where I'd start
A young company with a small team: a private programme on a platform with managed triage, a narrow scope and a short reward table. A larger company with its own security staff: an open programme makes sense once the private one runs smoothly. A launch or compliance deadline: pentesting on demand. A protocol or token project: a platform built for smart contracts, with an audit competition before launch and a bounty once the code is live.
Whichever route you take, judge the platform on the boring parts: triage, reporting and how it handles disputes. The researchers will find things either way. It's the handling that decides whether the programme is worth running.
Want your tool on this list? Featured spots and reviews here are paid partnerships. Here's how to partner with me.


