How GetVocal tripled organic demo-intent leads during an SEO and AEO engagement

Discovered Labs checked keyword demand and mapped GetVocal's existing pages to a revised six-pillar structure. Google impressions, click-through rates and rankings informed content refresh criteria. AI measurement tracked GetVocal across buyer scenarios, constraints and comparisons. The method has five steps:

1. Build a model of the category from the product, the customers and real sales conversations.
2. Restructure the pillars so they match the decisions buyers make. For GetVocal that meant going from eight down to six.
3. Map existing pages and keyword demand to each pillar, then build prompts covering buyer scenarios, constraints and comparisons.
4. Track Google rankings, impressions and clicks alongside AI measurement every few days, keeping branded and non-branded results apart.
5. Use indexing issues, underperforming pages and missing topics to prioritize technical fixes, content refreshes and new pages.

Separating branded and non-branded queries and prompts shows whether GetVocal appears when buyers search without naming the company. Google rankings and AI responses reveal different parts of that discovery process.

Discovered Labs' March 16, 2026 audit found that 56% of the GetVocal blog posts audited were not indexed by Google. We prioritize technical SEO work around the pages that have a commercial job:

1. Crawl the whole site and separate the pages that can rank from redirects, duplicates and leftover test pages.
2. Check whether those commercial pages can be reached. On GetVocal’s site two of them had nothing linking to them (orphan pages), and the main conversational AI landing page was missing from the internal link graph entirely.
3. Check which published pages Google has indexed and which still need attention.
4. Select refresh candidates with at least 1,000 impressions in 90 days, either CTR below 1% or average position 5–20, at least three internal links or one referring domain, and relevance to a defined pillar.
5. Implement the fixes ourselves.

Discovered Labs pairs its content technology with an editorial team of 10 writers and editors. The technology supports research, drafting and content scoring, while the editorial team checks facts, sharpens the argument and edits for clarity and tone before publication.

An audit found that three of GetVocal's six topic clusters lacked a main pillar page, even though supporting articles already existed. Discovered Labs uses these gaps to decide what to create or refresh:

1. Choose a buyer question from a priority topic cluster.
2. Decide whether to create a page, refresh one, or leave it alone. Refresh where impressions already exist and the page underperforms. Create where the question lacks a dedicated page.
3. Build the page around the buyer’s question, with a specific answer up front and supporting evidence below.
4. CITABLE is Discovered Labs' proprietary seven-part framework for creating content that AI systems can retrieve, verify and cite. Keep the subject, evidence and any limiting condition in the passage so its meaning stays clear when quoted on its own.
5. Track the page’s Google rankings, impressions and clicks alongside its visibility in AI answers.

Discovered Labs identifies the sources cited in AI answers, then chooses an action based on who controls each source. The method has five steps:

1. Start from the sources assistants cite when they answer the questions that matter.
2. Read what the source says. Absent, wrong, out of date and already well represented are four different problems with four different fixes.
3. Classify the relationship. Discovered Labs uses Own, Claim, Earn or Contest: a page you publish, a profile you can claim, coverage you have to earn, and a source where attention goes elsewhere.
4. Edit a source you own, update a claimable profile, earn independent coverage, or seek credible inclusion. Where inclusion is not possible, build a stronger alternative source.
5. Recheck citations to the source. For Reddit, match cited thread URLs to placement records and distinguish same-month placements, earlier placements and other threads in the same subreddit.

In our Reddit LLM Citations Research, we found ChatGPT could retrieve Reddit content without displaying it as a citation.

GetVocal's April 2026 report recorded more than twice as many citations to Reddit sources as March, with half as many confirmed new placements.