Business result
Revenue, cost or margin, supported by a method, a baseline and a stated period.
Results
Evidence with its context.
An internal system, technical delivery, an illustrative example and entrepreneurial practice — with the evidence and limits of each made explicit.
Internal project · checked 7 October 2026
Applying the method to this website.
The requirement was practical: when someone makes an enquiry, keep a reliable record and give the person responsible a clear next step. We checked the public form, stored request, CRM handoff and internal alert separately. The handoff follows defined rules; the commercial response remains a human responsibility.
On 7 October 2026, a clearly labelled synthetic enquiry submitted through the live form was found in Supabase. The automatic schedule then delivered the internal alert and created one matching Brevo opportunity with the original request details. The existing test contact was reused without adding it to a marketing list. Brevo recorded the alert as delivered; the recipient’s inbox was not inspected.
Two subsequent worker calls found no pending task. The health endpoint reported two delivered tasks, none pending and none requiring review. This verifies that sequence on the live system, not every possible failure. Separately, consented visit measurement was checked in English and French; accepting analytics is not required to send an enquiry.
Follow-up on 10 October 2026: the production-test alert was located in the recipient’s inbox, with successful SPF, DKIM and DMARC checks. It was still marked unread. This confirms mailbox receipt; it does not establish that a person acted on the enquiry.
This is our own system, tested in production on 7 October 2026 with a synthetic request, not a client case study or a commercial lead. Independent uptime monitoring still needs to be connected. Uncertain delivery outcomes require human review. No revenue or time saving is claimed.
What the checks establish
Keep receipt, follow-up and measurement distinct.
| Part of the system | Evidence at this checkpoint | Limit |
|---|---|---|
| Contact receipt | One labelled synthetic request submitted through the live form and found in Supabase. | A test request, not a commercial lead. |
| CRM and internal alert | Exactly one matching opportunity; existing test contact reused without marketing list membership. The provider recorded one delivered alert. | Provider delivery does not establish that a person read the message. |
| Scheduled processing | The automatic schedule processed both tasks. Two subsequent worker calls were idle; the health endpoint showed no pending or review task. | A dated production check. Independent uptime monitoring remains to be connected; uncertain outcomes require human review. |
| Visit measurement | Consented page views observed in EN/FR; refusal checked separately. | Analytics does not count every operational enquiry. |
| Homepage performance | One mobile lab comparison on 7 October 2026: performance 79 to 86; LCP 5.2 to 3.8 seconds. | A laboratory observation, not a field-performance or business-result guarantee. |
Khalid set priorities and authorised releases. AI coding assistants helped prepare implementation, tests and documentation, including independent code review. Khalid remains responsible for publication decisions and commercial follow-up.
Proof types
Different claims require different evidence.
Operational improvement
Time, errors or delays measured before and after a change.
Technical delivery
A system designed and built. This alone does not establish a business outcome.
Teaching example
A fictional scenario explaining an approach. No client engagement or result is claimed.
Evidence
Proof, clearly labelled.
A worked example
A fictional onboarding scenario that shows the decisions and measurements I would consider. No client result is claimed.
Read the exampleKlicky portfolio
Websites and applications from the studio I founded. Technical delivery is distinct from a measured business outcome.
View the portfolioAn operator’s perspective
My role in the businesses I run, presented separately from client engagements.
Explore the venturesClient case studies are published only with client approval, a measurement method and a stated period.
How to assess a result
Ask what changed, and how it was checked.
A baseline you can compare
Ask what was observed before the change, over which period, and whether the work after launch was comparable. Count review and exception-handling effort.
A defined role and scope
Distinguish the consultant’s decision work, the studio’s delivery and the client team’s contribution. A screenshot can show an interface; it cannot establish revenue or time saved.
A testable workflow
The fictional onboarding walkthrough makes proposed rules and acceptance criteria inspectable. It is educational material, with no measured client outcome.
Read the illustrative onboarding exampleYour next step
Use the evidence to choose the conversation.
A decision to make
Bring the alternatives, constraints and current cost of the problem. We can examine what deserves investment.
Discuss AI consultingA workflow to deliver
Bring the process, current tools and the change you need. We can check dependencies and define acceptance.
Discuss AI implementationLet’s find the right next step
What would change if this workflow worked?
Start with a free 20-minute conversation about the problem, the people and the practical constraints.
Discuss your project