I had a wake-up call about three months ago. I was reviewing analytics for a B2C client and noticed something odd. Their organic Google traffic was down 18% quarter over quarter, but their revenue was up. The leads were coming from somewhere, but not from the channel I'd been spending 90% of my time on.
When I dug into it, the growth was coming from YouTube embeds driving email signups, Reddit threads sending high-intent visitors, and ChatGPT citations generating a steady trickle of visitors who converted at nearly 5x the rate of Google organic traffic.
That's when I stopped treating SEO as a Google-only discipline. Here's what I've learned since.
The Data That Changed My Mind
Semrush published research on daily time spent across search surfaces in the US, and the ranking caught me off guard (Semrush, 2026):
- TikTok: 52 minutes/day
- YouTube: 46 minutes/day
- Instagram: 33 minutes/day
- X (Twitter): 32 minutes/day
- Facebook: 31 minutes/day
- Reddit: 30 minutes/day
- Snapchat: 30 minutes/day
- Google Search: 30 minutes/day
- ChatGPT: ~26 minutes/day
- Pinterest: 14 minutes/day
Google Search is eighth. People spend more time searching on TikTok, YouTube, Instagram, X, Facebook, Reddit, and Snapchat than on Google. I'm not saying Google is irrelevant. It still drives the most referral traffic to websites. But the idea that it's the dominant search behaviour is objectively wrong when you look at where people actually spend their time.
Add the other numbers and the picture gets clearer:
- 31% of Gen Z start searches on AI platforms instead of Google
- Google Search pageviews fell 34% year-over-year (Press Gazette)
- 68% of Google searches end without a click
- AI search traffic converts at 14.2% versus Google's 2.8%
- ChatGPT handles over 1 billion queries per day
Putting all your SEO effort into Google is like putting all your money into one stock. It might still be a good stock, but the risk concentration is irresponsible.
Platform by Platform: What I've Actually Changed
ChatGPT and Perplexity
ChatGPT drives 77% of all AI-driven website referral traffic. Perplexity holds about 15%. Together, they represent a new search channel that barely existed two years ago and now processes over a billion queries daily.
What works here is different from traditional SEO. AI models don't scan title tags and meta descriptions. They parse your content for direct, clear answers to questions. They look for structured data, specific numbers, and authoritative statements they can cite.
What I actually do:
- Structure content with clear question-and-answer patterns in the first 200-500 words
- Include specific data points with sources (AI models prefer citable facts)
- Build comprehensive topical coverage so the model recognises the site as authoritative on a subject
- Check robots.txt to confirm AI crawlers aren't blocked
- Monitor brand mentions in ChatGPT and Perplexity weekly using manual queries
The content that gets cited most often by AI isn't the most "optimised." It's the most directly useful. Pages that answer questions clearly, with numbers and evidence, without burying the answer under 500 words of preamble.
YouTube
YouTube is the second-largest search engine by query volume and people spend 46 minutes a day there. But most SEO strategies treat video as an afterthought, if they consider it at all.
Here's what I've found works:
What I actually do:
- Create video versions of top-performing blog content (the research is already done, just present it)
- Focus on search-friendly titles with specific queries people type into YouTube
- Write detailed descriptions with timestamps, key points, and links back to the website
- Use YouTube chapters to match specific search intents within a single video
- Embed videos in corresponding blog posts (this creates a two-way traffic loop)
The client I mentioned earlier? Their YouTube videos weren't even that polished. Phone recordings with screen shares. But they were answering specific questions that nobody else on YouTube was covering in their niche. That specificity matters more than production value.
Reddit content now ranks prominently in Google search results. Google's partnership with Reddit means that Reddit threads often appear in the top 5 for "best X" and "X vs Y" queries. And Reddit itself has 30 minutes of daily search surface time.
But you can't just show up and promote your stuff. Reddit communities will destroy you for that.
What I actually do:
- Identify the 3-5 subreddits where the target audience genuinely hangs out
- Participate authentically for weeks before ever mentioning a brand or product
- Answer questions with genuine expertise, linking to helpful resources (sometimes mine, mostly not)
- Create original, data-backed posts that provide standalone value
- Monitor brand mentions and respond when relevant
The traffic from Reddit is lower volume but remarkably high intent. People who click through from a Reddit discussion where someone genuinely recommended your resource are much more qualified than random Google visitors.
TikTok
52 minutes of daily attention. For certain demographics, TikTok IS the search engine. "TikTok SEO" sounds silly until you look at the numbers.
I'll be transparent: this is the platform where I have the least personal experience. But I've watched clients succeed with a few specific approaches:
What works:
- Short, direct answers to specific questions (same pattern that works for AI search)
- On-screen text that matches common search queries
- Series formats that build topical authority over time
- Behind-the-scenes content that demonstrates expertise rather than just claiming it
TikTok rewards authenticity and specificity over polish. A solo expert sharing real knowledge in a 60-second video can outperform a brand's polished production. If you have genuine expertise, TikTok is one of the few platforms where that expertise alone can build an audience.
How I'd Split My Time (Honestly)
If I were starting a content strategy from scratch today for a B2B or B2C brand, here's roughly how I'd allocate effort:
- Google SEO: 40% (still the largest referral source, but no longer 90%)
- AI search optimization (ChatGPT/Perplexity): 20% (highest conversion rate, growing fast)
- YouTube: 20% (massive daily attention, compounds over time)
- Reddit: 10% (high intent traffic, builds genuine authority)
- TikTok/Social: 10% (audience dependent, test and measure)
For existing sites that are already heavily invested in Google, I wouldn't abandon that foundation. I'd reallocate incrementally. Drop Google effort from 90% to 60%, put that 30% into the highest-potential alternative channels, and measure what moves.
What I'd Stop Doing
- Stop treating Google as the only metric that matters. Track AI citation rates, YouTube search impressions, Reddit referral traffic, and branded search volume alongside organic traffic.
- Stop creating content only for Google's format. A blog post that ranks well on Google but can't be cited by AI, adapted into a video, or discussed on Reddit is only serving one-fifth of your potential audience.
- Stop ignoring branded search. According to the Semrush research, branded search volume is the best leading indicator of overall search visibility. If people are searching for your name, you're winning on every platform. If they're not, you're invisible on most.
- Stop waiting for perfect before publishing on new platforms. My client's phone-recorded YouTube videos outperformed competitors with professional production. Start rough, improve over time.
The 80/20 of Search Everywhere
If this all feels overwhelming, here's the minimum effective version:
- Keep doing Google SEO, but structure content to answer questions directly (this helps AI search too)
- Check where your audience actually spends time by interviewing 5-10 recent customers about their research process
- Pick ONE additional platform to invest in based on those interviews
- Track branded search volume monthly as your north star metric
- Repurpose what works by turning blog content into video, turning video into Reddit discussions, turning all of it into AI-citable resources
The era of Google-only SEO worked because Google was where everyone searched. That's simply not the case anymore. The data is clear. The question is whether you adjust your strategy before the traffic decline forces you to.
I waited longer than I should have. Don't make the same mistake.
