AI Prompts for Writing SEO Title Tags

A fintech startup published 40 blog posts over six months. Every post was well-researched and genuinely useful. Average position in Google: 18. Average click-through rate: 0.8%. The SEO manager finally ran an analysis comparing their title tags to the top-ranking competitors for each keyword. The pattern was immediately obvious: their titles described the articles. The competitors' titles promised specific outcomes. Readers clicking through search results aren't looking for an article about a topic—they're looking for a specific answer to a specific question.

That's the distinction at the heart of every effective ai prompts seo title tags workflow.

Before: The Weak Prompt

Most people approach title tag prompts like this:

"Write an SEO title tag for a post about [topic] using the keyword [keyword]."

The output: "[Keyword]: Everything You Need to Know" or "A Complete Guide to [Keyword]" or "What Is [Keyword] and How Does It Work?"

These titles are keyword-present and search-engine visible. They're also indistinguishable from the other 8 results on the page. They describe a category of content—not a reason to click this specific article.

Why It Fails

Weak title tag prompts fail because they're optimized for search engine bots, not for human searchers making click decisions. Bots need the keyword. Humans need a reason to choose your result over the one above and below it.

The click decision happens in a fraction of a second. In that fraction of a second, the searcher is pattern-matching against their mental model of "what result looks most likely to answer my question." A title that signals specificity ("Why 73% of SaaS Companies Lose Leads in the First Follow-Up") beats a title that signals generality ("Lead Follow-Up Strategies for SaaS") every time—even at lower positions.

⚠️ Common mistake: Treating the title tag and the post headline as the same thing. The title tag is the clickable text in search results. The post headline (H1) is what readers see after clicking. They can be different—and often should be. The title tag is optimized for clicks; the H1 is optimized for confirming the reader landed in the right place. Use separate prompts for each.

After: The Improved Prompt

Here's a title tag prompt teardown that produces click-worthy results:

Write 6 SEO title tag variations for a blog post targeting the keyword "[keyword]". 

Post angle/main insight: [describe in 2 sentences what makes this post's approach specific or different]
Target reader: [who searches this keyword and what they're hoping to find]
Post's primary value: [specific outcome or answer the post delivers]

Title requirements:
- Under 60 characters (count each variation and list the count)
- Include "[keyword]" or a close variation naturally
- Each variation must use a different title structure: (A) number + benefit, (B) question, (C) how-to, (D) contrarian claim, (E) specific outcome with timeframe, (F) comparison angle
- None should start with the keyword—it can appear anywhere in the title
- Avoid "Ultimate Guide," "Complete Guide," "Everything You Need to Know"

After generating all 6, recommend which 2 you'd A/B test first and why.

What this does: The "recommend which 2 to A/B test" instruction forces the AI to make a judgment call—and the reasoning it gives is often genuinely useful for understanding which title structure fits the specific post's angle.

⚡ Pro tip: The "none should start with the keyword" instruction produces significantly more creative titles. When the keyword must appear mid-title, you're forced to lead with something more compelling—a number, a qualifier, a provocative phrase. This constraint alone improves average CTR on titles I've tested by a meaningful amount.

Breaking Down Each Title Structure

Number + benefit: "7 Follow-Up Email Mistakes Killing Your SaaS Demos" — Works because numbers signal a specific, scannable list. The benefit is implied (avoid these mistakes).

Question: "Why Do SaaS Lead Follow-Ups Fail? (And How to Fix Yours)" — Works for informational queries where the searcher is actively diagnosing a problem.

How-to: "How to Write Follow-Up Emails That Get Replies From Cold Leads" — Works for instructional queries. The "that" clause is what makes it specific rather than generic.

Contrarian claim: "Stop Personalizing Your Sales Emails (Do This Instead)" — High CTR potential but requires that the post actually delivers on the contrarian angle. Don't use this structure if the post doesn't have a genuine counter-argument.

Specific outcome with timeframe: "The 3-Email Follow-Up Sequence That Books Meetings in 72 Hours" — Extremely high CTR for commercial queries. Specificity (3 emails, 72 hours) signals that the post delivers something concrete.

Comparison angle: "Cold Email vs. Follow-Up: Which Gets More Replies From [ICP]" — Works for queries where the searcher is evaluating options or approaches.

⚡ Pro tip: For competitive keywords where you're entering a crowded SERP, use this additional prompt: "Here are the title tags of the top 5 results for [keyword]: [paste titles]. Identify (1) what they all have in common structurally, and (2) what title structure is NOT represented that might perform well based on this keyword's search intent. Use the gap to write one title tag that's genuinely different from these 5."

Variations for Specific Page Types

E-commerce product page:

Write a title tag for a product page for [product name]. Primary keyword: [keyword]. This title must: include the product name, include the keyword, communicate the primary differentiator in under 60 characters. The differentiator: [what makes this product different from generic alternatives].

Local service page:

Write title tag variations for a local service page targeting "[service] in [city]." Include: one that leads with the service, one that leads with a benefit, one that includes a trust signal (years in business, number of customers served, certification). Each under 60 characters, character count listed.

Landing page:

Write title tags for a landing page for [offer]. Primary keyword: [keyword]. The page goal: [lead gen / free trial / demo / purchase]. Include the keyword and one signal of the specific value of converting (what they get when they click). Avoid vague CTA language—"Sign Up" is weaker than "Get Your Free [Specific Thing]."

What this does: Landing page title tags have a different optimization goal than blog post title tags. Blog posts need to signal information value. Landing pages need to signal conversion value. The prompts reflect that difference.

⚡ Pro tip: Once you find a title tag structure that consistently performs for your specific audience, save the pattern—not just the individual title. For example: "Number + negative outcome + solution type" (e.g., "7 Onboarding Mistakes That [Negative Consequence] and How to Fix Them"). That pattern becomes a template you apply to multiple posts, not a one-off success.

Save and Reuse This

The fintech startup from the opening scenario rebuilt their title tags using the improved prompt above. Average CTR across their 40 posts improved from 0.8% to 2.4% over three months—without any ranking changes. The same posts in the same positions, just with titles that earned the click.

Title Tag Testing at Scale

Most SEO teams write one title tag per post and never revisit it. But title tags are uniquely testable—you can change them at any point and see CTR results in Search Console within two to four weeks. Unlike ranking changes, which can take months to materialize, CTR changes from title tag optimization are measurable relatively quickly. The teams that treat title tags as ongoing optimization see compounding gains; those that treat them as one-time decisions leave significant traffic on the table.

I want to systematically improve title tags across my underperforming posts. Here are 8 posts with their current title, average position, and CTR from Search Console:
[Post | Current Title | Avg Position | CTR]

For the 4 posts with the lowest CTR relative to their position, write two alternative title tags each. Each alternative must use a different structural approach than the current title and a different approach from each other. After each alternative, write one sentence on what CTR improvement this tests and why.

What this does: Systematic title tag testing turns your existing content into a continuous CTR optimization project. The "what hypothesis this tests" instruction forces strategic thinking rather than random changes—and it builds useful pattern recognition over time about what resonates with your audience.

⚡ Pro tip: One title tag pattern that consistently underperforms despite being extremely common: titles that start with a year ("2026 Guide to..."). These titles age poorly and often lose clicks as the year progresses. Better approach: make the timeliness signal specific and driven by content substance rather than calendar year. "The Approach Replacing Traditional [Topic] Right Now" ages better because the claim isn't tied to a specific year—it signals currency through the word "replacing" rather than through a date that will become a liability twelve months from now.

Save your title tag prompt templates in PromptABCD with notes on which structures work for which keyword types and audience segments. That's a testing library that gets more actionable and more valuable with every quarter of real click-through data behind it.