5 Questions to Ask Before Hiring an AI Search Company
Introduction
The AI search optimization market is growing fast. That means there are more vendors every month, more claims, more packages, and more confusion about who actually delivers results versus who just has a good sales pitch.
This article gives you five specific questions to ask any vendor before you sign a contract. These aren't gotcha questions. They're diagnostic questions designed to reveal whether the vendor understands AI search, whether they do the actual work, and whether their approach will produce AI recommendations or just produce invoices.
At Yazeo, we welcome these questions. We built our process around answering them specifically. But this article isn't about us. It's about helping you evaluate anyone in this space, including us.
Question 1: "what specific work do you do that's different from traditional SEO?"
Why this question matters. The most common fake in the AI optimization market is the rebrand: SEO agencies relabeling their existing work as "AI optimization." If the vendor can't clearly articulate the difference, they're selling you SEO with a trendy label.
What a good answer sounds like:
"Our AI optimization work includes: building 15+ citations per month on authoritative sources beyond typical SEO directories, managing entity data consistency across 30+ web sources (not just Google Business Profile), creating content specifically structured for AI extraction (question-based, answer-first format with FAQ sections), implementing comprehensive structured data beyond basic SEO schema, diversifying your review presence across multiple platforms, and monitoring your AI recommendation status across ChatGPT, Gemini, Perplexity, and Google AI Overviews. Our SEO work focuses on keyword rankings and Google traffic. Our AI work focuses on entity authority across the broader web."
What a red flag answer sounds like:
"Our SEO work is designed to improve your overall online visibility, which includes AI." This is the rebrand. They're doing SEO and calling it AI optimization.
"We use AI tools to optimize your content for better search performance." This is AI-assisted SEO, not AI search optimization. They're using AI as a production tool, not optimizing for AI as a discovery channel.
Question 2: "Can you show me a before-and-after example of a client who went from not recommended by AI to recommended?"
Why this question matters. Results are the proof. A vendor that has successfully moved a client from AI invisibility to AI recommendation can show you the evidence: screenshots of AI responses before and after, timeline data, and specific metrics.
What a good answer sounds like:
"Here's a client in [industry] in [city]. When they started, ChatGPT didn't mention them at all for '[service] in [city]' queries. After 4 months of work, ChatGPT names them in approximately 70% of relevant queries. Here's the before screenshot. Here's the after screenshot. Here's the monthly progression of their citation count and AI mention rate."
What a red flag answer sounds like:
"We're a newer company and don't have case studies yet." This isn't automatically disqualifying (everyone starts somewhere), but it means you're taking more risk. Proceed only if their process answers and pricing are strong.
"We can't share client results due to confidentiality." Every legitimate marketing firm can share anonymized case studies without naming the client. Complete inability to show any results is concerning.
"Our clients see improved online visibility." This is vague. "Online visibility" could mean anything. You need to hear "AI recommendation" specifically.
Question 3: "how do you measure success, and what metrics will I see in monthly reports?"
Why this question matters. If the vendor measures success with Google metrics (rankings, organic traffic, click-through rates), they're measuring SEO, not AI performance. AI-specific metrics are different and require different monitoring.
What a good answer sounds like:
"Our monthly reports include: AI recommendation status across ChatGPT, Gemini, Perplexity, and Google AI Overviews (are you being named in recommendation queries?), AI description accuracy (is the description correct?), citation count growth (how many new authoritative citations were built this month?), entity consistency score (how consistent is your data across web sources?), competitive AI position (how do you compare to competitors?), and (when available) AI-attributed lead tracking."
What a red flag answer sounds like:
"We track keyword rankings, organic traffic, and domain authority." These are SEO metrics. They tell you nothing about AI visibility.
"We provide access to our dashboard where you can see your AI visibility score." Dashboards that show a score without explaining what drives it and what actions are being taken to change it are monitoring tools, not optimization services.
Question 4: "how many citations do you build per month, and on what types of sources?"
Why this question matters. Citation building is the #1 factor in AI recommendations. The volume, quality, and specificity of citations directly determines whether AI recommends your business. A vendor's citation strategy tells you whether their work will produce results.
What a good answer sounds like:
"We build 12 to 20 citations per month on authoritative sources including: industry-specific directories, local chambers of commerce and business associations, professional association directories, local and regional publications, BBB, and high-authority review platforms relevant to your industry. Each citation is individually researched, manually placed, and verified. We use your standardized entity data for every placement to ensure consistency."
What a red flag answer sounds like:
"We submit your business to hundreds of directories through our automated system." Automated blasts to low-authority directories don't produce AI visibility. AI weights source quality heavily. 200 low-authority citations are worth less than 20 high-authority ones.
"We focus on building backlinks to your website." Backlinks ≠ citations. A backlink from a guest post that doesn't mention your business name in context doesn't help AI visibility. Citations require named mentions of your business with consistent entity data.
"Our content strategy naturally builds citations over time." Content-driven citation growth is too slow (1 to 3 per month organically). AI needs 25 to 30 citations as a minimum threshold, which requires active, deliberate placement.
Question 5: "what's a realistic timeline for results, and what happens if we don't see them?"
Why this question matters. The honest timeline for AI search optimization is: first signals in 60 to 120 days, meaningful recommendations in 4 to 6 months, compounding returns after 6 months. Any vendor promising faster results is either misleading you or defining "results" as something other than actual AI recommendations.
What a good answer sounds like:
"You'll see citation count growing and entity data improving by Month 1 (process metrics). You should see first AI mentions on at least one platform by Month 3 (early outcome). Consistent AI recommendations across major platforms typically develop by Month 4 to 6 (results). If we don't see any AI recognition by Month 3, we'll evaluate and adjust the strategy. We recommend a 6-month minimum engagement to reach the compounding phase, but we provide monthly reporting from Day 1 so you can track progress continuously."
What a red flag answer sounds like:
"You'll be on ChatGPT within 30 days." This is almost certainly false for any business starting from a low visibility baseline. AI recommendation is not something you can buy with a fast timeline guarantee.
"Results vary and we can't guarantee any specific outcome." While technically honest (no ethical vendor guarantees specific AI responses), this answer combined with no case studies and no measurable milestones suggests the vendor can't point to any track record of producing results.
"We require a 12-month commitment upfront." A vendor confident in their results should be willing to demonstrate progress within 3 to 6 months. Requiring 12 months upfront with no intermediate milestones is a risk transfer, not a service agreement.
How to use these questions in a vendor evaluation
Run through all five questions with any vendor you're considering. score each answer:
- Clear, specific, confident answer with examples: 3 points
- Adequate answer but vague on specifics: 2 points
- Red flag answer (matches the warning patterns above): 0 points
| Score | Assessment |
|---|---|
| 13-15 | Strong candidate. They understand AI search and can articulate their approach. |
| 9-12 | Moderate. Some capability gaps. Push for more specifics before committing. |
| 5-8 | Weak. Significant knowledge or capability gaps. Consider alternatives. |
| Under 5 | Walk away. This vendor is selling something other than AI search optimization. |
The Meta-Question: Why Would a Legitimate Vendor Share This Framework?
You might wonder why an AI search optimization company (that's us, Yazeo) would publish a buyer's guide that could be used to evaluate us.
The answer is straightforward: we want informed buyers. Informed buyers ask better questions, set realistic expectations, and make better partners. They don't churn because of misaligned expectations. They don't blame the vendor for outcomes the vendor never promised.
We'd rather lose a prospect to a competitor who genuinely outscores us on these five questions than win a client who didn't ask the right questions, had unrealistic expectations, and churned in 3 months.
Ask us these five questions. Ask our competitors these five questions. And choose the vendor whose answers are the most specific, the most honest, and the most verifiable.
Start with data. Run your free AI visibility audit at yazeo.com to understand your current position before you talk to any vendor. Knowing your baseline makes every vendor conversation more productive, because you can evaluate their promises against your specific reality.
Key findings
- Five diagnostic questions reliably distinguish legitimate AI optimization vendors from rebranded SEO agencies and monitoring-tool sellers.
- The most revealing question is #1 (what specific work differs from SEO), because it exposes rebrands immediately.
- Case studies with before/after AI screenshots are the strongest proof of capability. Inability to show any results is a significant red flag.
- Citation strategy specifics (volume, source types, manual vs. automated) are the most predictive indicator of whether the vendor's work will produce AI recommendations.
- Realistic timelines (60 to 120 days for first signals, 4 to 6 months for recommendations) distinguish honest vendors from those overselling fast results.
Frequently asked questions
Ask the hard questions before you write the check
The AI search optimization market is full of vendors making big claims. Some of them deliver. Many don't. The five questions above separate the two groups before you spend a dollar.
Don't sign based on a sales pitch. Sign based on answers. Specific, verifiable, realistic answers to specific, diagnostic questions.
Run your free AI visibility audit at yazeo.com and come to every vendor conversation with data. When you know your baseline, you can evaluate every promise against reality. And that's how you make a decision you won't regret.