From Zero to Recommended: A Practical GEO Roadmap for Swiss B2B

Starting From Zero Is More Common Than You Think

If you have just tested your company's AI visibility and discovered that ChatGPT, Claude, and Perplexity have no idea who you are — welcome to the majority. Most Swiss B2B companies, even well-established ones with strong Google rankings, are effectively invisible to AI assistants.

The good news: zero is not a permanent state. With systematic effort, most companies can achieve measurable AI visibility within 30 days and meaningful recommendation presence within 90 days. This roadmap gives you a concrete, week-by-week plan to get there.

Before You Start: Establish Your Baseline

You cannot measure progress without a starting point. Before implementing anything, spend an hour documenting your current AI visibility:

1. Test five to ten prompts that your prospects would realistically use. Ask ChatGPT, Claude, Perplexity, and Google's AI features. Examples: "Best [your category] providers in Switzerland", "Who offers [your service] in the DACH region?", "Compare [your category] solutions for Swiss [your target industry]".
2. Document every response. Note whether you are mentioned, whether the information is accurate, how competitors are described, and which sources are cited.
3. Score your baseline. A simple framework: 0 = not mentioned anywhere, 1 = mentioned in one platform, 2 = mentioned in multiple platforms, 3 = recommended with accurate information. Most companies starting out score 0 across the board.

per4mx automates this baseline process across all major AI platforms, running dozens of relevant prompts and scoring your visibility against competitors. But even a manual baseline is better than no baseline.

Week 1: The Technical Audit

Before creating any content, make sure AI models can actually find and access your existing information. This week is about removing blockers.

Day 1-2: Crawler Access

• Check your robots.txt file. Ensure GPTBot, ClaudeBot, PerplexityBot, Applebot, and Google-Extended are explicitly allowed.
• Review your server logs for the past month. Are any AI crawlers receiving 403 or 429 errors?
• If you use a CDN or WAF (Cloudflare, Akamai, etc.), check that AI user agents are not being blocked at the edge.
• Test your key pages with a user-agent switcher to confirm they render content without JavaScript.

Day 3-4: Index Verification

• Log into Google Search Console and verify your key pages are indexed.
• Register at Bing Webmaster Tools (most Swiss companies skip this). Submit your sitemap. Verify your site. This single action makes you discoverable to ChatGPT's web search.
• Check your presence in Common Crawl by searching commoncrawl.org.

Day 5: Baseline Monitoring

• Set up regular AI visibility monitoring. Whether you use per4mx or do it manually, establish a weekly cadence for testing the same prompts across all major AI platforms.
• Document any existing mentions (even incorrect ones) — they are your starting point.

Week 2: Build the Foundation

With technical access confirmed, this week you build the structured information layer that AI models need.

Day 6-7: Create Your llms.txt File

• Write a clear, factual llms.txt file following the standard format. Include: company description, products and services, key facts (founding year, location, employee count, markets served), differentiators, and key links.
• Deploy it at yourdomain.ch/llms.txt.
• Reference it in your robots.txt with a comment: # See also: /llms.txt

Day 8-9: Implement Schema Markup

• Add Organization schema to your homepage with complete company information.
• Add Product or Service schema to your main offering pages.
• Add FAQ schema to any FAQ sections.
• Validate all schema with Google's Rich Results Test.

Day 10: Content Audit

• Review your top ten website pages. Are they written in clear, factual language that an AI could quote? Or are they full of marketing jargon?
• Identify pages that need rewriting — typically the homepage, About page, and main product/service pages.
• Check that your heading hierarchy (H1 to H3) is clean and logical on every key page.

Week 3: Create and Publish

Foundation in place, now you generate the content that AI models will find and cite.

Day 11-13: Publish Expert Content

• Write and publish two to three authoritative articles on your website's blog. Topics should directly address questions your prospects ask: "How to evaluate [your category] providers in Switzerland", "What does [your service] cost for Swiss SMBs?", "[Your industry] trends in the DACH region for 2025".
• Each article should be 800 to 1200 words, fact-rich, and free of promotional fluff. Include specific numbers, comparisons, and actionable advice.
• Optimise each article for both traditional SEO and AI discoverability — use clear headings, structured data, and natural language.

Day 14-15: Update Key Landing Pages

• Rewrite your homepage and About page to be AI-friendly. Lead with factual descriptions, not slogans. State clearly: who you are, what you do, who you serve, and what makes you different.
• Update your product and service pages with specific details: pricing models, technical specifications, integration capabilities, and client success metrics.

Day 15: Issue a Press Release

• Write and distribute a press release through Presseportal.ch or a similar service. The news hook can be a product update, a partnership, a milestone, or even a market analysis.
• Include a detailed company boilerplate with all key facts.
• The goal is multi-platform indexing — getting your company information onto authoritative news sources that AI models trust.

Week 4: Amplify and Measure

Your foundation is built and content is published. Now amplify reach and take your first measurements.

Day 16-18: Build External Mentions

• Update or create your company profile on key directories: zefix.ch, local.ch, LinkedIn company page, Crunchbase, and any industry-specific directories for your vertical.
• Ensure consistent company information across all profiles — same name, same description, same facts.
• Request listing on relevant industry association member directories if you are a member.

Day 19-20: Community Engagement

• Contribute one to two thoughtful posts in industry-relevant LinkedIn groups or forums.
• Answer category-level questions on platforms where your prospects research (Reddit, Quora, industry forums).
• Do not self-promote — contribute genuine expertise. The mentions accumulate naturally.

Day 21: First Measurement

• Re-run all baseline prompts across all AI platforms.
• Compare results to your Week 0 baseline.
• Realistic expectations for Week 4: Perplexity is most likely to show improvement first (it always searches in real time). Google AI Overviews may also show changes. ChatGPT and Claude changes will be slower unless they trigger web searches for your prompts.

What to Expect: 30 Days vs 90 Days

Setting realistic expectations is essential. Here is what most Swiss B2B companies experience:

After 30 Days

• Perplexity: Your company likely appears in relevant queries, especially if you published quality content and issued a press release. Citations from your website and press release sources.
• Google AI: Improvement in AI Overviews for queries where your pages now rank well. Faster than other platforms because it uses Google's real-time index.
• ChatGPT: Marginal improvement. If ChatGPT triggers web search for your category, you may appear. Training-data-based answers will not change yet.
• Claude: Minimal change in 30 days. Claude's training data updates are infrequent, and it is conservative about web search.

After 90 Days

• Perplexity: Consistent mentions across most relevant prompts. Your citation count should have grown as more content and mentions accumulate.
• Google AI: Strong presence in AI Overviews for your key terms, assuming your SEO is solid.
• ChatGPT: Noticeable improvement. Your Bing presence is established, your press releases have been indexed, and there is a growing body of content for ChatGPT to find when it searches. Training data improvements depend on when the next model update occurs.
• Claude: Gradual improvement. Your web content is more established, and if Claude performs web searches for your category, it can now find you.

The Most Common Mistakes in the First 30 Days

Companies that fail to see improvement typically make one of these errors:

• Skipping Bing. Without Bing Webmaster Tools setup, you are invisible to ChatGPT's search. This is the single most common oversight.
• Blocking AI crawlers. They fix their robots.txt on Day 1 but forget about the WAF, CDN, or WordPress security plugin that is still blocking bots.
• Writing marketing content instead of expert content. "We are the leading provider of..." gets filtered out by AI models. Factual, useful content gets cited.
• Expecting instant results from training data. Training data updates happen on each AI provider's schedule, not yours. Focus on real-time search visibility first.
• Not measuring. Without regular monitoring, you cannot tell whether your efforts are working or where to adjust. Track weekly, without exception.

The Day-by-Day Accountability Tracker

Use this tracker to hold yourself accountable during the 30-day sprint. Print it, pin it to your wall, and check off each item as you complete it.

Week 1 Tracker

• Day 1: Downloaded and reviewed robots.txt. AI crawlers allowed: Yes / No
• Day 1: Checked server logs for GPTBot, ClaudeBot, PerplexityBot activity
• Day 2: Tested 10 buyer prompts across ChatGPT, Claude, Perplexity, Google AI. Results documented
• Day 3: Registered at Bing Webmaster Tools. Sitemap submitted
• Day 3: Verified key pages in Google Search Console
• Day 4: Checked Common Crawl presence
• Day 5: Baseline AI visibility score calculated: ___ / 100
• Day 5: Competitor visibility documented for top 3 competitors

Week 2 Tracker

• Day 6: llms.txt file drafted with all company facts
• Day 7: llms.txt deployed and verified accessible at domain root
• Day 7: robots.txt updated with llms.txt reference
• Day 8: Organization schema added to homepage
• Day 9: Product/Service schema added to key offering pages
• Day 9: FAQ schema added to existing FAQ sections
• Day 10: Top 10 pages reviewed for content quality. Pages needing rewrite identified: ___

Week 3 Tracker

• Day 11: Expert article 1 drafted and published
• Day 12: Expert article 2 drafted and published
• Day 13: Expert article 3 drafted and published (optional)
• Day 14: Homepage rewritten with factual, specific descriptions
• Day 14: About page rewritten with quotable facts and clear positioning
• Day 15: Press release written with AI-optimised boilerplate
• Day 15: Press release distributed through Presseportal.ch

Week 4 Tracker

• Day 16: zefix.ch listing verified and updated
• Day 16: local.ch listing verified and updated
• Day 17: LinkedIn company page updated with consistent description
• Day 17: Industry association directory listing updated
• Day 18: Crunchbase profile updated (if applicable)
• Day 19: First community contribution posted (LinkedIn group or forum)
• Day 20: Second community contribution posted
• Day 21: All 10 baseline prompts re-run across all AI platforms
• Day 21: Week 4 visibility score calculated: ___ / 100
• Day 21: Improvement documented: +___ points from baseline

Budget Planning for the 30-Day Sprint

Here is a realistic budget breakdown for a Swiss B2B SMB executing this roadmap:

Activity	Direct Cost (CHF)	Time Investment
Technical audit (robots.txt, Bing, logs)	0	4-6 hours
llms.txt creation and deployment	0	1-2 hours
Schema markup implementation	0-500 (developer time)	3-5 hours
Content rewriting (homepage, About)	0	4-6 hours
Expert articles (2-3)	0-2,000 (if outsourced)	8-15 hours
Press release (writing + distribution)	300-1,500	3-4 hours
Directory updates	0	2-3 hours
AI visibility monitoring (per4mx)	79/month	1 hour/week
Total	399-4,099	30-45 hours

For most Swiss B2B companies, the 30-day sprint can be executed for under CHF 1,000 in direct costs and approximately one working week of marketing time. This is a fraction of what a single trade show booth costs — and the impact is measurable and compounding.

Role Assignment: Who Does What

For companies with more than one person involved, here is how to distribute the work:

• Marketing lead: Owns the overall roadmap, writes content, manages directory listings, coordinates press release. Time commitment: 20-25 hours over 30 days.
• Web developer or IT: Implements schema markup, deploys llms.txt, updates robots.txt, reviews server logs. Time commitment: 5-8 hours over 30 days.
• Founder or CEO: Reviews and approves llms.txt content, provides quotes for press release, validates competitive positioning. Time commitment: 2-3 hours over 30 days.
• Sales team: Provides input on what questions prospects ask (shapes content topics and prompt testing). Time commitment: 1-2 hours total.

For solo founders or very small teams, one person can execute the entire roadmap. The time investment is front-loaded in weeks 2-3 (content creation) and lighter in weeks 1 and 4.

Beyond Day 30: Sustaining Momentum

The 30-day roadmap gets you started, but AI visibility is an ongoing discipline. After the initial sprint:

• Publish one to two expert articles per month
• Issue a press release every quarter
• Update your llms.txt file whenever your offerings change
• Monitor AI visibility weekly across all platforms with per4mx
• Update directory listings annually
• Track competitor AI visibility and adjust when they gain ground

The 90-Day Extension: Building on Your Foundation

After the initial 30-day sprint, extend your programme with these monthly milestones:

• Month 2 (Days 31-60): Publish two additional expert articles targeting specific buyer scenarios identified during Week 4 measurement. Contribute one guest article to a Swiss industry publication. Add FAQ sections to your three most-visited product pages. Begin active participation in one industry forum or LinkedIn group.
• Month 3 (Days 61-90): Issue your second press release. Create or update detailed case studies with specific, quotable metrics. Expand content to address German-language AI queries if you have not already. Conduct a full competitive AI visibility analysis — where do competitors outperform you, and what are they doing differently?

By day 90, companies that follow this extended roadmap typically achieve consistent mention rates of 40-60% across relevant prompts on Perplexity and Google AI, with growing visibility on ChatGPT and Claude.

Troubleshooting: When Progress Stalls

If you have followed the roadmap but are not seeing improvement, check these common blockers:

• CDN or WAF blocking AI crawlers. Your robots.txt may allow AI crawlers, but your Cloudflare, Akamai, or server-level security may be blocking them at a higher level. Check your server access logs for 403 responses to GPTBot, ClaudeBot, and PerplexityBot.
• JavaScript-rendered content. If your key pages rely on JavaScript to display content, AI crawlers see an empty page. Test by disabling JavaScript in your browser — if the content disappears, it is invisible to AI crawlers.
• Content is too generic. If your expert articles address broad topics ("The Future of Digital Transformation") rather than specific buyer questions ("How Swiss Manufacturers with 50-200 Employees Evaluate MES Software"), AI models have no reason to cite you for specific buyer queries.
• Bing indexation incomplete. Check Bing Webmaster Tools for crawl errors. Some Swiss websites have technical issues that prevent Bing from indexing them fully, even when Google has no problems.
• Inconsistent company information. If your directory listings, LinkedIn, and website describe your company differently, AI models have low confidence and may avoid citing you. Run a consistency audit across all sources.

The companies that treat AI visibility as a continuous programme — not a one-time project — are the ones that build lasting recommendation presence. Start with 30 focused days, then maintain the rhythm. For proof that this approach works, read our case study of a Swiss B2B company that reached top 3 on every AI model. And to choose the right tools for ongoing monitoring, see our comparison of AI visibility tools for B2B.

Frequently Asked Questions

Can I complete this roadmap without any technical skills?

Most of the roadmap is non-technical: writing content, updating directory listings, drafting press releases, and running AI visibility tests. The technical steps — updating robots.txt, deploying llms.txt, implementing schema markup — require basic web server access. If you use a managed CMS like WordPress, Squarespace, or Webflow, most of these tasks can be done through the platform's interface. If you have a developer on your team (even part-time), allocate 5-8 hours of their time for the technical steps. If not, a freelance web developer can handle all technical tasks in a half-day engagement.

What if I do not have time to write expert articles?

Content creation is the most time-intensive part of the roadmap, but there are efficient alternatives. Consider: (1) repurposing existing content — if you have sales presentations, client proposals, or internal knowledge base articles, these can be adapted into blog content with relatively little effort. (2) Recording voice memos about topics you know well and using AI transcription and editing to create articles. (3) Using per4mx's content generation feature, which produces AI-optimised articles based on your company information and identified visibility gaps. (4) Hiring a freelance writer who specialises in B2B content for the Swiss market — budget CHF 500-800 per article.

Is this roadmap relevant for companies that already have some AI visibility?

Yes, but you should adapt it. If you already appear in some AI responses, your baseline is not zero — it is whatever your current citation rate and accuracy scores are. Skip the steps you have already completed (e.g., if you are already registered with Bing, skip Day 3-4). Focus your energy on the areas where you are weakest: if ChatGPT mentions you but Claude does not, prioritise training data strategies. If your information is cited but inaccurate, prioritise consistency auditing. The roadmap framework applies regardless of starting point — the specific actions just shift based on your gaps.

How do I convince my management team to invest 30 days in this?

Three persuasion approaches that work with Swiss B2B management: (1) The demo approach — in a management meeting, ask ChatGPT and Perplexity to recommend companies in your category. If competitors appear and you do not, the problem sells itself. (2) The cost argument — the total investment is under CHF 1,000 and approximately one week of marketing time. Compare this to a single trade show (CHF 10,000-30,000) or a Google Ads campaign (CHF 2,000-5,000/month). (3) The competitive urgency argument — AI visibility is a first-mover game. Once a competitor establishes itself as the AI-recommended provider in your category, displacing them requires significantly more effort than establishing the position first. Every week of delay is a week your competitors may be getting ahead.