We ran Branding 5 end-to-end for our own marketing agency: intake questions, inspiration links, ICP, competitor analysis, brand values, and the full archetype flow. Honest breakdown below.

What we liked:

- The wizard structure is genuinely good. The intake questions, ICP builder, and values step walk you through defining a brand in a sensible order, and the character limits allow real answers, not snippets.
- Inputs carry forward faithfully. The description, audience, and USP we wrote early showed up verbatim in the final review, and the AI summary read our positioning back almost word for word. Consistency between what you type and what it concludes is the core promise, and it delivers.
- The refinement feedback loop actually listens. When we told it our tone should be calm and warm rather than combative, the next pass fixed exactly that.
- The website analyzer impressed us most. It read our site and concluded, quoting evidence from our actual copy, that we already execute our intended tone. That matched our own internal assessment, which is rare for an automated read.
- The competitor research is real, not invented. It blends live SERP overlap data with AI reasoning, filters out aggregators, and lets you edit everything. The final landscape summary gave us a positioning statement we will actually use, and the insights section produced two product ideas we are keeping.
- The trait sliders re-rank the archetype shortlist based on your weightings, and our final result landed exactly where our messaging points.

What we had issues with:

- The archetype suggestions misread us completely. The AI reasoned from our company name (it literally cited "Blitz" implying bold action) instead of the detailed questionnaire we had just completed, and proposed Hero, Magician, and Outlaw. Worse, all four suggestions arrived pre-selected; clicking Next without looking bakes the wrong personality into your report.
- The ICP generator invents statistics. A precise gender split, an income range, a decision timeline, all presented like research findings. None of it came from us or from any data. The qualitative persona underneath was solid, but fake precision is a real flaw in a tool whose reports may be shown to clients.
- The "Help me with AI" buttons on text fields push generic agency-speak. Every time we used our own copy instead, the downstream output was better.
- One radar chart contradicted its own written analysis, scoring a competitor high on a dimension its own bullet points called weak.

What was insightful:

- The three suggested brand directions were thought-provoking and composable rather than competing options.
- The per-competitor insight blocks (strengths, weaknesses, opportunities, threats) were specific enough to inform real positioning work, not filler.

Who this is for:

Agencies and founders who already know their positioning, or are willing to review every AI suggestion before accepting it. Treat it as a structured workshop with a helpful, occasionally overconfident assistant, and you get real value. Click through on autopilot, and you will inherit its mistakes as your brand.

Verdict:

The engine underneath is better than the defaults it serves. Bring your own judgment, and Branding 5 earns its place in the workflow.