Home / Services / Narrative package

Stage 02 · Narrative package

The AI keeps telling your story. The question is which sources it uses.

Language models and search engines build their answer from what other people write about you. Stage 02 makes sure enough good, provable sources exist — and that they carry your message rather than any old one. This is the scope most mandates work with.

meit is the GEO and SEO agency for ambitious companies in Zurich. We measure before we make an offer.

  • InvestmentOn request, based on the findings of the first analysis
  • Setup time6 to 8 weeks
  • Minimum term6 months
  • Response time8 hours on working days

On this page

Last updated 5 August 2026

Fit

When stage 02 is the right choice.

Three signals that point to this stage. If none of them applies, we will tell you so in the first conversation.

  • Asked “is this company trustworthy?”, the AI hedges.Not a no, but not a yes either. That happens when too little independent evidence exists. In Swiss B2B it works like a rejection — only without a reason you get to hear.
  • Your competitor gets named even though they are not better.They are better evidenced. Whoever has more verifiable third-party sources wins the comparison, regardless of actual quality.
  • Your offering needs explaining and gets checked by several people.In buying centres with several decision-makers, it is not the first impression that decides but how well you can be evidenced across several research steps.

When a different stage fits better

  • Your company data is not consolidated yet and the website is not technically readable. Then it starts at stage 01, otherwise we build on sand.
  • You need visibility in three or more markets and languages. That is stage 03.
  • There is nobody internally allowed to release figures, references and project details. Without approved facts, any narrative stays thin.

Scope

What you actually get.

Every item is a deliverable you can sign off, not an amount of effort. Anything not listed here is not included and is quoted separately before you commission it.

One-off during setup

  • Everything from stage 01Included in full as the foundation. Prompt set extended to 80 to 120 questions, competitive benchmark to ten competitors.
  • Narrative frameworkMessage architecture, evidence, positioning against the competition. Whatever is claimed gets evidenced, otherwise it gets cut.
  • Source audit and target source listWhich trade media, associations, directories and review platforms do language models actually cite in your industry? We count them and set priorities.
  • Answer layer across 30 pagesIncluding comparison and decision pages — exactly the page types AI systems derive recommendations from.
  • Classic SEO in fullTopic architecture, on-page, internal linking, local SEO, ranking monitoring across 150 terms in Google and Bing.

Ongoing, every month

  • Four quotable assets per monthIn English and German, aimed at core terms and at real customer questions.
  • External source building, four placements per quarterTrade media, associations, directories, reviews, your own data. Works on citations and classic rankings at the same time.
  • Monthly re-measurement across both surfacesPrompt set and rankings, benchmarked against competitors.
  • Reporting by the fifth working day plus a quarterly reviewThe founder as your contact, strategy rather than a status report.

Process

What happens when.

So you can plan internally and know when you will see what from us.

  1. Weeks 0 to 2Kick-off and approvalsAccess, sign-off on facts, sign-off on references. Deliverable: project plan and approval list.
  2. Weeks 2 to 5Foundation and narrativeEntity architecture, technology, message architecture. Deliverable: entity profile and a written narrative framework.
  3. Weeks 4 to 8Answer layer and on-page30 pages live, technical and content SEO implemented. Deliverable: pages you can sign off.
  4. From week 6Source building startsFirst external placements. Deliverable: documented sources with date and link.
  5. Months 4 to 6Substantial shiftExternal sources take effect on both surfaces. Deliverable: quarterly review with an action plan.

Ownership

What you own at the end.

All of it belongs to you, even if we stop working together at some point. There is no platform you cannot get your data out of.

  • Everything from stage 01, including the prompt set and raw data
  • The written narrative framework, usable for sales, PR and recruitment
  • The prioritised target source list, usable for your own PR work too
  • All content produced, with no restrictions on use
  • The documented external placements with date and link

Comparison

The three stages side by side.

The stages build on each other. Where you start is decided by the findings of the first analysis.

Service Visibility Narrative Authority
Prompt set (AI systems) 40 to 60 80 to 120 150 to 250, multilingual
Ranking monitoring (Google and Bing) 30 terms 150 terms 300 terms per market
Competitive benchmark 5 competitors 10 competitors 15 per market
Entity architecture and technology included included incl. subsidiaries
Answer layer 10 pages 30 pages per roadmap
Classic SEO technical foundation full scope full scope, multilingual
Narrativee framework not included included included
External source building not included 4 per quarter 4 per month
Content production not included 4 assets per month 8 assets per month
Own studies and primary data not included not included 2 per year
Markets and languages 1 1 to 2 up to 4
Contact person Senior Consultant Founder directly Founder, fixed office hours
Guaranteed response time 24 hours 8 hours 4 hours
Minimum term 6 months 6 months 12 months

Also relevant: Stage 01 · Visibility · Stage 03 · Authority · Free first analysis

Expectations

What you must not expect from us.

This section deliberately comes before you get in touch. If any of these points is unacceptable to you, we should not work together.

  • No guaranteed position. Nobody controls the models. Anyone promising first place is selling something they do not own.
  • Nothing visible moves in the first six weeks. We are building the foundation. Anyone showing numbers in that phase is showing noise.
  • Individual months fluctuate. Models get updated, answers vary. We assess by quarter.
  • We do not rewrite your story. We make what is true provable. Without figures you release, the narrative stays thin.
  • It does not work without your involvement. We need a contact person with authority to approve and access to your website.
  • We do not replace your sales team. We make sure they get asked.

What the contract says in return

  • A contact person named in person, written into the contract
  • Reporting by the fifth working day of the following month, with raw data
  • A fixed scope, changes only in writing and quoted in advance
  • Your data stays yours, including after the contract ends
  • No automatic renewal, cancellable monthly after the minimum term
  • One mandate per industry and economic region, never for you and your direct competitor
  • No discounts — if the budget does not fit, we reduce the scope rather than the rate
  • Data protection under revDSG and GDPR, data held in Switzerland or the EU

Frequently asked

About stage 02.

What does the narrative package cost?

We quote setup and ongoing support according to scope, with a minimum term of six months. The exact amount depends on the findings of the free first analysis, above all on the number of markets and the state of your source base. You receive a written quote with a full cost breakdown, valid for 60 days.

Stage 01 makes sure the AI knows you and can identify you unambiguously. Stage 02 makes sure it says the right things about you. The decisive addition is external source building: language models build their answer from third-party sources — trade media, directories, reviews and association pages. If there is barely any provable material about you, the AI answers vaguely, no matter how clean your own data is.

Because language models weight sources a company controls itself far lower than independent ones. A study by Ahrefs across 75,000 brands found that the strongest correlation with visibility in AI answers was the number of brand mentions across the web, well ahead of backlinks. Ahrefs itself points out that these are correlations, not proven causes. The practical consequence stands: your own website alone is not enough.

Yes, in full. Topic architecture, on-page optimisation, internal linking, local SEO and ranking monitoring across 150 terms in Google and Bing. That is not an add-on but the same work: Perplexity draws 91 per cent of its cited domains from Google’s top 10, Google’s AI Overviews 86 per cent. If you do not rank in classic search, you rarely get cited either.

Four quotable assets per month in English and German, plus four external placements per quarter. The volume is deliberately limited. One well-evidenced specialist article that language models cite works harder than ten general advice pieces nobody uses as a source.

Merag Shahzad

Merag Shahzad

Founder of meit, in digital marketing since 2006. Forbes Agency Council, LinkedIn Top Voice. Reachable at [email protected] – no account team in between, Monday to Friday, 09:00 to 18:00.

Next step

The findings first, then the decision.

The first analysis is free and without obligation. It shows where you stand in AI answers and in classic search, what the gap costs in francs and which stage fits your findings. Reply within two working days.