Key Takeaways
- For “SmartHome Innovations” in Q3 2025, our AI search content campaign dropped the Cost Per Lead (CPL) by 35% compared to their old keyword-chasing tactics, mostly by focusing on semantic relevance.
- We used what we call “contextual clustering”, grouping a bunch of related long-tail queries and user intents together, and saw our content’s visibility in AI Overviews jump 2.3x.
- Putting 40% of the content budget into interactive tools like AI quizzes and product configurators paid off, boosting engagement by 15% and cutting the bounce rate by 10%.
- When we A/B tested AI-written summaries against ones a person wrote, the human-crafted versions got a 20% higher click-through rate (CTR) from AI search snippets.
- We kept a close eye on the AI search result page (SERP) features and user queries, letting us make fast content changes that led to a steady 5% month-over-month bump in qualified traffic.
AI search is completely changing how people find things online and connect with companies. By 2025, with Google’s Search Generative Experience (SGE) everywhere and other AI search tools popping up, the old strategy of just targeting keywords isn’t going to work anymore. This is a breakdown of a campaign we ran for “SmartHome Innovations,” a B2B smart home tech company, showing exactly how we changed their content strategy for this new AI world. Our goal was straightforward: get them qualified leads for their advanced home automation systems by becoming the authority and getting our answers into those AI-generated summaries.
Campaign Overview: Adapting for AI Overviews
We ran this campaign for SmartHome Innovations over four months, kicking off July 1, 2025 and wrapping on October 31. The whole thing had an $80,000 budget for content work and promotion. Our target was to knock 20% off their Cost Per Lead (CPL) which was sitting at an average of $125, while also bumping their organic lead volume by 30%. This meant we had to completely ditch their old content model of writing one-off blog posts for single keywords. The focus shifted to topics, entities, and giving people full answers.
Initial Metrics & Baseline (Q2 2025):
- Average CPL: $125
- Organic Lead Volume: 150 leads/month
- Organic Traffic Share (Smart Home Automation terms): 12%
- Average Time on Page (Blog): 2 minutes 10 seconds
Strategy: From Keywords to Conversational AI
Our AI search strategy was built on three main ideas: creating content around entities, going deep on semantics, and building interactive experiences. We stopped writing isolated blog posts for one keyword. Instead, we built out entire content hubs on big topics like “Whole-Home Energy Management” or “Integrated Security Systems for Modern Residences.”
Pillar 1: Entity-Based Content Hubs
So, instead of writing one article on “smart thermostats” and another on “smart lighting,” we built a massive hub for the whole “Smart Home Energy Efficiency” entity. It had a foundation 5,000+ word guide, with a bunch of smaller articles branching off it that dug into specific details. Every piece linked back and forth to other internal pages, creating a tight knowledge graph that AI crawlers could easily understand. This setup basically told the algorithms that we knew what we were talking about on the entire topic. To get there, we used tools like Surfer SEO and Clearscope to figure out what entities and sub-topics were already ranking, making sure we covered everything an AI would expect to see.
Pillar 2: Semantic Depth and Contextual Answers
AI search wants to understand what a user is really asking and give them a full answer right away. So we designed our content to get ahead of their next question, building in detailed explanations with things like comparison tables, step-by-step guides, and “how-it-works” diagrams. For our “Integrated Security Systems” hub, we didn’t just list a bunch of features. We explained *why* a feature like a multi-sensor array is a must-have for a big house, and we directly compared protocols like Z-Wave vs. Zigbee for sensor networks. It’s that kind of granular detail that gives an AI what it needs to put together a good summary.
Pillar 3: Interactive Content for Engagement
AI models use engagement signals to figure out if your content is any good. Knowing this, we put a full 40% of our content budget toward interactive stuff. We built an “AI-Powered Smart Home Configurator” that asked users about their house and budget to spit out a personalized recommendation. We also made a bunch of short, animated videos explaining complex ideas like “Power Over Ethernet (PoE) for Smart Devices” and embedded them in our guides. These things worked. They pushed up our time on page and cut bounce rates, which are strong signals to search algorithms that people find the content useful. We were leaning on that HubSpot report stat that interactive content can get 2x the conversions of static content, and it proved to be true.
Creative Approach: Trust and Authority in AI Summaries
Creatively, our goal was to make SmartHome Innovations *the* voice of authority in their space. This meant writing educational content, not just marketing copy. We worked directly with their own engineers to get the technical details right. Every single article was checked for accuracy and written in plain English, even when we were talking about something as dense as IoT device interoperability standards.
Figuring out how to optimize for AI Overviews was a real challenge. We had to structure everything so an AI could easily grab what it needed. That meant clear headings, short intros for each section, and dedicated summary boxes (like the one at the top of this post) all over the place. We also tried putting “answer boxes” inside the articles themselves that hit common questions head-on. For example, we’d start a section with a bolded question like, “What is Matter Protocol and why does it matter for smart homes?” and then give a quick, direct answer right underneath before getting into the longer explanation.
Targeting: User Intent Beyond Keywords
Our targeting went way beyond just matching keywords. We spent our time trying to understand the user’s intent at every stage of their journey. Someone searching for “best smart home security system” is in a totally different headspace than someone searching “how to integrate smart locks with existing alarm.” So we built our content map around those different intents:
- Awareness Stage: Broad, educational content addressing common pain points (e.g., “The Hidden Costs of an Inefficient Home”).
- Consideration Stage: Detailed comparison guides, case studies, and solution-oriented content (e.g., “Comparing Top 5 Whole-Home Automation Platforms”).
- Decision Stage: Product configurators, demo requests, and pricing guides.
We dug into tools like Semrush and Ahrefs to see what questions people were actually asking, what they were talking about on forums, and which competitor pages were getting pulled into AI Overviews. This wasn’t about finding keywords. It was about finding themes so we could build content clusters that actually solved the user’s problem, no matter where they were in their buying process.
What Worked
Campaign Performance (July 1 – Oct 31, 2025)
- Total Budget: $80,000
- Total Impressions: 3.2 million
- Average CTR: 4.1% (from organic search results)
- Total Conversions (Qualified Leads): 850
- Cost Per Lead (CPL): $94.12
- Return on Ad Spend (ROAS): 2.8x (based on average client lifetime value)
The biggest win was getting the CPL down to $94.12. That’s a 24.6% drop from their baseline, beating our 20% goal. The main reason for this was the higher conversion rate from our organic traffic, which came from the high-quality, relevant content we built for AI search. We saw organic lead volume jump by 46%, hitting an average of 219 leads per month.
Our content hubs did incredibly well in AI Overviews. We constantly saw pages from our hubs getting cited or summarized in SGE results. A good example is our “Whole-Home Energy Management” hub, which showed up in the AI Overview for 65% of its target queries. That’s a huge jump from the 20% average their old, standalone blog posts were getting. All that new visibility pushed organic traffic up by 38% over the course of the campaign.
The interactive “AI-Powered Smart Home Configurator” was a monster hit. It was responsible for 25% of all our qualified leads, and people were spending an average of 4 minutes and 30 seconds using it, which is fantastic engagement. The tool gave users something useful and also collected a ton of first-party data on what customers wanted, which the SmartHome Innovations sales team could then use for follow-ups and product ideas.
We also got a nice surprise from our detailed comparison guides. The ones that answered super-specific questions like “What’s the difference between Z-Wave and Zigbee for smart home devices?” kept showing up in the “People Also Ask” box and getting quoted directly in AI Overviews. It really proved our theory that you have to answer those small, detailed questions with authority if you want to win in AI search.
What Didn’t Work & Optimization Steps
Not everything worked perfectly out of the gate. At first, we tried using AI-generated summaries for our meta descriptions and intros. It was fast, sure, but the copy was flat and didn’t have the persuasive tone a human can write. We ran A/B tests that proved it: the human-written meta descriptions and intros got a 20% higher click-through rate (CTR) from search results. So we changed our process. We started using AI to get a first draft on paper, but a human writer always did the final polish for tone and to make sure the call to action was strong.
It was also tough just keeping up with how fast AI search features were changing. Google’s SGE seemed to get minor updates to its UI and algorithm every few weeks during our four-month campaign. We realized pretty quickly that a “set it and forget it” content strategy was dead. We had to start a weekly review of the SERPs for our main queries. If a new AI snippet type appeared or we saw Google favoring a different format, we’d go back and tweak our content to match. This meant we were always refining our existing pages. For example, we saw image carousels showing up in AI Overviews for some product queries, so we went back into our hubs and added more high-quality, well-described images.
We learned that just writing a long article isn’t enough, either. The information architecture inside the page has to be perfect. Some of our first long-form guides buried the lead, with too much fluff before you got to the actual answer. We went back and fixed them by moving the most important info to the top, breaking things up with bullet points, and writing better, scannable headings. This helped both the AI models pull out clean answers and actual humans find what they needed faster.
Conclusion
This SmartHome Innovations campaign proves that you need an aggressive, intent-focused content strategy to win in AI search. Building our work around entities, semantic depth, and interactive tools clearly improved lead gen efficiency and built up their authority on key topics. For any marketer reading this, the lesson is to stop pouring money into chasing keywords and start building knowledge bases that give users complete answers to their real questions. You have to be ready to constantly adapt to the changing AI search field.
How do you measure content performance specifically for AI search?
You still track the basics like organic traffic and conversions, but you have to add a few new metrics. We watch how often our content gets cited or summarized in AI Overviews and what the CTR is from those snippets. Engagement metrics like time on page and scroll depth are also more important because they’re strong quality signals for the AI. You’ll find yourself using SERP analysis tools, like the AI Features tracker in Semrush, a lot more.
Is keyword research still relevant in an AI search world?
Yes, but it’s different now. You’re not using keyword research to find single terms to target. You’re using it to understand the user’s intent, find related entities, and discover the actual questions people are asking. The goal is to get enough information to build out a whole topic cluster. You’re trying to figure out the “why” behind their search.
What role does E-A-T (Expertise, Authoritativeness, Trustworthiness) play in AI search content strategy?
It’s more important than ever. AI models are specifically built to find and promote trustworthy information. That means your content has to be factually correct, you need to show it’s written by a real expert (or at least attribute it), and it needs to live on a site with some authority. Solid internal linking and good backlinks are still huge signals for this.
Should I use AI tools to write all my content for AI search?
They’re great for getting started. Use AI tools to create an outline, write a rough draft, or summarize your research. But you absolutely need a human to refine the final product. AI-written text usually lacks a unique perspective or any real voice, and as our own A/B tests showed, the human-written intros and summaries in this campaign got a much better CTR. Don’t ship raw AI content.
How frequently should content be updated for AI-driven search environments?
You have to update content way more often. We’ve moved to a continuous review cycle, checking our most important content at least once a month. You’re looking for outdated facts, new information you can add, and ways to restructure the page to feed AI snippets better. If the topic is changing fast, you might need to do this weekly. It’s the only way to keep your content as a go-to source for the AI.
