VERIXA / Trust Center
How we work, written down.
This page exists so you can judge the practice before you judge the pitch. It states how we work, how your material is handled, where a person signs off, what software is and is not allowed to decide, and the things we will not claim.
Standards, not marketing / Eight principles / One explicit list of what we will never say
Operating principles
Eight rules the work runs on.
01Evidence before opinion
No recommendation leaves here without the observation that produced it attached. Findings are tied to specific data from your account, never a generic checklist applied from outside.
02Human accountability
Decision support may be delegated. Accountability for the final decision never is. A person reviews every deliverable before it is sent, and on a first engagement that review is elevated to an independent approver.
03What AI does, and does not
AI accelerates research, analysis and drafting, and is directed to challenge weak assumptions rather than agree by default. AI has no authority to approve a deliverable for delivery, to commit VERIXA to pricing or scope, or to decide what we offer. Every tool is judged on one test: does it extend human judgment, or try to replace it.
04How we access your account
Access is granted through the platform’s own permission and invite system, never a shared password, at the minimum scope the engagement needs rather than blanket administrator rights. It is formally revoked and confirmed within five business days of completion. Confirmed, not assumed.
05Platform rules are not negotiable
VERIXA never recommends, performs or assists with anything that breaks a platform’s terms of service. That includes review manipulation, prohibited ranking or bid tactics, and account sharing. Where a client asks for it we decline and record the decision. This is never traded for revenue or retention.
06Who owns what
You own your account data and the findings we produce for you. VERIXA owns its underlying methodology and templates. Using your results as a public case study requires separate written permission, captured at the time and revocable by you on request. It is never assumed from general contract language.
07Conflicts of interest
Before accepting a new client we check for direct competitive overlap with an existing one. Where it exists we either decline the engagement or disclose it to both parties and proceed only with informed consent from both.
08Quality, disputes and refunds
Deliverables are checked against a written quality standard before delivery. If you dispute the quality of one, it is reviewed against that same standard rather than reflexively conceded or refused, and a refund is possible where it genuinely falls short. We do not escalate a dispute publicly under any circumstances.
09Saying we do not know
Unexplained is a permitted result. Where the evidence does not settle a question, the report says so and states what would be needed to settle it, instead of reaching for a plausible story.
10Growth is earned, never appearance
VERIXA does not represent itself as larger, more established or more proven than it actually is. That is why this site has no client logos, no testimonials and no case studies. There are none yet, and inventing them would contradict the only thing we are selling.
The line we hold
What VERIXA will never claim
Most of what damages trust in this industry is not a bad service. It is a claim that cannot be checked. Here is what you will not find on this site or in a VERIXA document.
- Guaranteed rankings, sales, or revenue outcomes
- Client names or logos we do not have written permission to use
- Testimonials or reviews we did not receive from a real client
- Case studies built from illustrative or borrowed numbers
- Certifications, badges or partnerships we do not hold
- A special relationship with Amazon, because there is not one
- Influence over Amazon decisions, rankings or enforcement
- Recovery of a suspended account as a certainty
- Headcount, offices or history beyond what is true today
- A finding stated with more confidence than its evidence carries
What we ask for, and what we do with it
What we ask for first
Reports you can export yourself: business reports, advertising reports, inventory and account health extracts. In most engagements this is enough to reach a finding, and it means no credential ever changes hands.
When access is needed
Some questions cannot be answered from exports. When that happens we say which question requires it, what level of access it needs, and how long we need it for. You grant it through Amazon’s own user permissions, at the lowest level that answers the question, and you remove it when the work closes.
What happens to your material
It is used for your engagement. It is not sold, not shared with other clients, and not used as training material for anything we build. On request we delete the working copies once a report is delivered and confirm in writing that it is done.
What software does, and what it is not allowed to do
We use software, including AI tools, to gather, sort and cross reference material. That is where its authority ends. A model output is treated exactly like an unverified observation: it enters the record, it is checked against a source, and it is either confirmed or marked unverified. No finding reaches you on the strength of a model output alone, and nothing a tool could not verify is presented as though it were established.
The practical test we apply: if a client asked where a sentence in the report came from, could we point at a dated source rather than at a tool? If the answer is no, the sentence does not ship.
Where a person signs off
Three points, and all three are human. A named person approves the observation set as complete enough to reason from. A named person approves each finding and its stated confidence. A named person approves the verification note, including the ones that report a failure. Those approvals are recorded with the work.
What you should expect from us
A scope agreed in writing before work starts. A report where every claim points to a dated source. The findings we decided against acting on, with reasons. An honest statement of what we could not establish. A return visit to check our own advice. And no claim on this website or in any document that we cannot show you the basis for.
What we tell you when something goes wrong
If we get a finding wrong, you hear it from us first, with what we missed and why. A practice built on evidence has to report against itself the same way it reports against an account.
A claim you cannot check is not evidence. It is just a confident sentence.