Fundamentals — the new search reality · 10 min read
Choosing your priority surfaces
You can't optimize equally for five search surfaces on one founder's schedule. A scoring framework to pick the two or three that move your niche — and ignore the rest without guilt.
Lesson 1 gave you the map: five search surfaces, each with its own crawler, signals, and citation pattern. The trap is what most people do next — try to win all five at once.
You have one founder's worth of hours. Spread them evenly across five surfaces and you end up mediocre on every one and dominant on none. This lesson is how to choose the two or three that matter for your product, and how to feel fine about deprioritizing the rest.
Why you're allowed to ignore surfaces
Recall the stat from Lesson 1: only 11% of domains cited by ChatGPT are also cited by Perplexity for the same query. That number cuts two ways.
- The bad news: optimizing for one surface buys you almost no free coverage on the others. There's no "rank once, win everywhere."
- The good news: because the surfaces are nearly independent, winning one does not require the others. You have permission to go deep on two surfaces and skip two entirely — the skipped ones were never going to come along for free anyway.
Prioritization isn't cutting corners. It's the only strategy the math actually supports.
The three questions that rank a surface
A surface earns your surface-specific effort only if it scores on all three axes. Rate each one 0, 1, or 2:
- Audience fit — do the people who would actually buy from you search here? (Not "does anyone use it" — does your buyer.)
- Winnability — can a new, low-authority domain realistically earn visibility here in months, not years?
- Conversion path — does visibility here lead to a click, signup, or sale — or just a name-drop with no path back to you?
Total the three for a score out of 6. Prioritize the surfaces scoring 4–6. Consciously drop anything at 2 or below.
The five surfaces, scored for a typical indie dev product
| Surface | Audience reach | Winnability (new domain) | Conversion | Indie verdict |
|---|---|---|---|---|
| Google blue links | Massive | Low — Domain Authority still rules here; 12–18 mo to traction | High (transactional intent) | Long game — don't lead with it |
| Google AI Overviews / AI Mode | Massive (~30% of queries) | Med–high — 47% of AIO citations come from pages ranking below #5 | Medium (zero-click risk, but brand exposure + selective clicks) | Schema-driven bonus |
| ChatGPT Search (Bing index) | ~800M weekly | Medium — Bing's index is easier to enter than Google's; rewards extractable structure | Med–high (AI-answer clicks convert 4–7×) | Core for most |
| Perplexity | Smaller, high-intent | Med–high — rewards data-rich pages, recency, real backlinks | High (research-then-buy intent) | Core for technical niches |
| Claude / Gemini grounded | Smaller, developer-heavy | Medium — rewards clean topic clusters and citation-friendly structure | High (your buyers literally live here) | Core for dev tools |
The pattern to notice: the surface with the most reach (Google blue links) is also the hardest for a new domain to win and the slowest to pay off. The AI-citation surfaces invert that — smaller audiences, but new domains can get cited the same week they publish, because Domain Authority lost most of its correlation (r=0.18 in 2026). That inversion is the entire opening for indie sites.
A worked example: a dev-audience product
Score the surfaces for something like a tool, library, or course aimed at indie developers:
- Perplexity — fit 2, winnability 2, conversion 2 = 6/6 → primary
- Claude / Gemini — fit 2, winnability 1, conversion 2 = 5/6 → primary
- ChatGPT Search — fit 2, winnability 1, conversion 2 = 5/6 → primary
- Google AI Overviews — fit 1, winnability 2, conversion 1 = 4/6 → secondary
- Google blue links — fit 1, winnability 0, conversion 2 = 3/6 → background investment
The conclusion flips the 2022 playbook on its head. The old advice was "start with Google blue links, build backlinks for a year." For a new dev-audience domain in 2026, you lead with the AI-citation surfaces — Perplexity, ChatGPT, Claude — where your buyers already are and where a clean new domain can win citations fast. You treat AI Overviews as a schema-and-extractability bonus that comes mostly for free once you're engineering pages for citation. And you treat blue links as a 12-month compounding investment running quietly in the background — never the thing you stake month one on.
This is why citation-first thinking matters: your priority surfaces should skew toward wherever a new domain can earn citations cheaply, not wherever the raw traffic looks biggest.
When your audience flips the answer
The framework transfers; the answer doesn't. Re-score for a different buyer and the priorities invert:
- Local service business (plumber, dentist): Google blue links + AI Overviews + Google Business Profile dominate. Perplexity and Claude score near zero on audience fit — skip them without a second thought.
- B2B SaaS selling to marketers: ChatGPT and Google AIO lead (that's where the buyer researches tools), Perplexity secondary, Claude a bonus.
- Consumer / DTC product: Google AIO + blue links + ChatGPT; the research-heavy engines barely register.
If you copied the dev-audience answer above onto a plumber's site, you'd pour months into Perplexity for an audience that isn't there. Always re-run the three questions for your buyer.
"Priority" is not "neglect the fundamentals"
Deprioritizing a surface does not mean breaking the basics on it. Technical hygiene — fast pages, valid HTML, sitemap.xml, clean robots.txt, baseline schema — lifts every surface at once and costs you nothing extra per surface. That floor stays in place everywhere.
What prioritization governs is where you spend the extra, surface-specific effort:
- ChatGPT → Bing Webmaster Tools + IndexNow submission (Module 2)
- Perplexity → original data tables, methodology, aggressive recency (Modules 3, 5)
- Claude / Gemini → tight topic clusters and entity disambiguation (Module 5)
- AI Overviews → FAQPage / HowTo schema tuned for extraction (Module 3)
You keep the shared foundation universal and aim the expensive, hand-crafted work at your two or three chosen surfaces.
Your action checklist
Copy this into your notes and do it before you write a single optimized page.
- Score all five surfaces 0–2 on audience fit, winnability, and conversion (total /6)
- Circle your 2 primary surfaces and 1 secondary
- Write one sentence: "We win on X and Y first, because our buyers are Z."
- For each primary surface, name the one surface-specific tactic you'll invest in (see the module map above)
- Set a measurement baseline per primary surface now, so you can tell later whether the bet paid off (Module 7)
- Calendar a re-score in 90 days
Sidebar — re-score quarterly. Surface share is still moving fast: AI Overviews appear on ~30% of queries and climbing, AI Mode is expanding, and ChatGPT Search keeps adding users. A priority list that was right in Q1 can be stale by Q3. This is a living decision, not a one-time setup.
You now have the mental model (five surfaces), the goal (citation-first), and a way to choose where to aim. The last fundamentals question is the one most courses skip entirely: is the domain itself clean enough to build all this on?
→ Next: Audit a domain before you buy it
This is one of the paid lessons. Unlock every module and every paywalled article for $199 one-time.