CallForge Start a conversation
Menu
Start a conversation

How CallForge Works

Clarity first. Useful work next.

We connect business context, technical delivery and independent challenge. Work starts with a clear problem and agreed scope—not an automatic build.

Start a conversation

Working together

A clear path from context to handover.

A typical engagement is scoped around these four steps. The detail depends on the problem, deliverables and support you agree with CallForge.

01

Understand the context.

Discuss the business problem, users, existing systems and constraints. Establish where CallForge can usefully contribute.

02

Agree scope and direction.

Define deliverables, assumptions, responsibilities, commercial terms and review points before substantive work begins.

03

Build and review.

Carry out the agreed assessment or implementation in reviewable stages. Keep decisions visible and use appropriate technical, factual and human review.

04

Handover and defined support.

Agree the handover, access and documentation needed for the work. Any ongoing care is separately scoped, with support boundaries and responsibilities made clear.

Prospect research

Before outreach: our evidence method

Before contacting a business, we check its public evidence and current customer experience. Where there is a credible opportunity, we prepare a focused recommendation and appropriate concept work. Research does not automatically lead to a proposal or outreach.

This evidence method applies when CallForge independently identifies prospective clients. It is not a required prospect-research workflow for inbound client projects, which start with the agreed engagement process above.

The method is deliberately selective. Research does not guarantee qualification; qualification does not guarantee a prototype; neither guarantees outreach.

  1. 01

    Define

    Set the market, offer and standard.

    Agree the buyer, qualification criteria and disqualifiers before discovery begins.
    Brief
  2. 02

    Discover and research

    Build verified business context.

    Find candidates and research them across multiple sources, separating observed facts from working inference.
    Evidence
  3. 03

    Test

    Challenge the apparent opportunity.

    Check the thesis against the organisation’s own public evidence and retain material contradiction.
    Scrutiny
  4. 04

    Qualify or reject

    Decide before anything is generated.

    Reject cases that lack support, commercial significance or a proportionate reason to continue.
    Decision
  5. 05

    Develop useful work

    Make the intervention concrete when justified.

    Prepare the written recommendation and appropriate proof. Website proposals use desktop and mobile design previews; a working implementation requires a separate authorised next step.
    Conditional
  6. 06

    Independent challenge

    Assign reviewers to try to break the case.

    Challenge the evidence, commercial reasoning and proposed intervention for unsupported inference or weak fit.
    Challenge
  7. 07

    QA and human approval

    Inspect every recipient-visible element.

    Complete factual QA, visual QA and human review before any next action is approved.
    Authority
  8. 08

    Prepare

    Create the outreach-ready evidence pack.

    Package the approved case and next action. CallForge does not send outreach autonomously.
    Ready

Rejection is productive

A weak case should disappear before it becomes outreach.

CallForge is not optimised to preserve every lead. It is designed to protect attention by stopping work that cannot support a credible commercial recommendation.

R–01Insufficient source support
R–02Weak commercial significance
R–03Material contradictory evidence
R–04Prototype would add theatre, not evidence
R–05Factual, visual or human review not satisfied

Controls inside the work

Trust comes from inspectable decisions.

The operating details remain private. The control principles do not: evidence is traceable, uncertainty is retained, proof is appropriate and humans keep authority.

01

Evidence and inference stay distinct.

A reviewer should be able to see what was observed and what was concluded.

02

Contradiction is preserved.

Counter-signals and unresolved questions remain part of the recommendation.

03

Proof fits the decision.

Use a design preview, annotated journey or focused brief when it makes the recommendation clearer. Do not confuse a concept with a working system.

04

Factual and visual QA are explicit.

Claims, sources and recipient-visible work are reviewed before release.

05

Human authority is final.

The system can prepare a case; a person decides what happens next.

What leaves the system

A recommendation that can be inspected, challenged and declined.

Every recommended lead must carry verified context, a specific problem, source support, contradiction, commercial significance, an intervention decision, challenge, QA and a human-approved next action.

Explore the delivery domains

Start a conversation

Start with the problem that matters.

Tell us what needs to improve. We will agree a clear scope, evidence standard and proportionate next step before substantive work begins.

Start a conversation