I Ranked Every AI Search Strategy S to F, Then Built a Scorecard So You Can Too

Build in Public · #15

A tier list of AI search strategies went round again this week, ranked S to F. I mostly agreed with it, which is rare. So instead of arguing with it, I ranked the strategies myself the way I actually see them, then built a scorecard that grades your site against them and hands you a plan. Here is the thinking behind it.

The spark was a post by borja (@borjafat) on X, laying every AI search strategy out as a tier list, S at the top down to F. Credit where it is due, the framing is theirs and it is a good one. What I want to add is not a rebuttal. It is the version of that ranking I would defend from my own build, plus the honest reasons some things sit lower than the industry likes to admit, and a tool you can use to score yourself instead of just nodding along.

Because that is the problem with a tier list you read and agree with. You close the tab and nothing changes. A ranking is only useful if it tells you what to do next on your own site.

Why I rank by impact, not by effort

Most SEO advice is sorted by how hard something is. That is the wrong axis now. The only axis that matters in AI search is how much a strategy moves whether you get quoted, named and clicked. Some easy things barely move it. Some hard things move it more than everything else combined.

So the ranking is blunt on purpose. S tier is the handful of moves that actually decide whether an answer engine cites you. Everything below it is either a multiplier on those moves or a hygiene factor that quietly loses you citations by its absence but never wins one on its own. Once you see the strategies sorted that way, where to spend your next hour stops being a mystery.

The S tier: the five moves that get you quoted

These are the five I would fight for at the top. Not because they are fashionable, but because each one, done properly, changes whether an AI names you.

Topical authority. AI picks its sources per subject, not per site, which is the single best piece of news for a small operator. You do not need to outrank a giant everywhere. You need to own one narrow category so completely that the model has no better source to pull from. This very hub, everything I have written on answer engine optimization sitting in one interlinked place, is me doing exactly that on purpose.

Original data and research. This is the one I would put at the very top if I had to pick a single move. Publish a number only you have, and the model is forced to name you when it uses it. Princeton’s study on generative engine optimization found that adding original statistics was the single biggest lever on whether content got surfaced in AI answers. Facts nobody else holds are uncopyable, and that is the whole game.

Digital PR. Most AI citations come from earned media, not from your own blog. The uncomfortable truth is that your site is often not the thing that gets cited, it is the thing that makes you worth citing somewhere else. Getting mentioned on sites the models already trust does more than another post on your own domain ever will. My free Press Release Writer is the on-ramp I built for this.

Parasite SEO. The strategy that works before your own domain is worth anything. Reddit, YouTube and LinkedIn are among the most-cited domains in AI answers, and listicles on them are a huge share of what gets quoted. Publishing on a borrowed high-authority domain gets you into answers while your own site is still earning trust. It is the fastest of the five, and it is exactly what my Parasite SEO Powerplay service is built to run.

Buy-intent keywords. The tier that pays for the other four. Comparison and alternative searches convert far better than informational ones, and your pricing page is often the one search you win outright because no competitor can answer it for you. SERP Gap AI finds the buyable gaps where the current results are weak, and Page Sniper tears down what a competitor is actually selling so you can beat their intent page.

The A and B tiers: multipliers, not the engine

Below the S tier sit the strategies everyone already knows. They are not lower because they stopped working. They are lower because they multiply the S tier rather than replacing it.

Backlinks (A). The lever that makes every other tier work. A brand-name link is a mention that also passes equity, and the biggest single jump comes from the first link onto a page that has none. I build toward this with Skyscraper AI, but I will say plainly that links follow good assets, they do not substitute for them. Chase links before you have anything worth linking to and you get nowhere.

Internal linking (A). Free authority you already control and the most underused move on this list. It tells both search and AI which of your pages matter, and it concentrates the strength you already have onto the pages you actually want to rank. I ran a pass across this whole site recently and it cost me nothing but an afternoon. If you do one thing this week, do this one.

E-E-A-T (B). Experience, expertise, authority and trust. It will rarely win you a citation on its own, but its absence quietly loses you ones you should have got. A visible, credible author and clear sourcing make a model more comfortable quoting you. The AEO Readiness Checker flags where those signals are missing on a page.

The bottom of the list: real, but badly oversold

This is the part a tier list is actually for, because it gives you permission to stop over-investing in things the industry treats as gospel.

Schema (C). Useful, worth doing, wildly oversold as a magic bullet. Structured data helps a machine parse your page, but it does not make thin content quotable. Get it right once and move on. My free FAQ Builder handles the one bit of schema that genuinely earns its keep, snippet-friendly question and answer pairs, and then you are done thinking about it.

Domain authority (D). I can speak to this one from direct, slightly embarrassing experience. This site is DR 29 and ranks for almost nothing outside my own name. A borrowed authority score built on syndicated links did not translate into a single ranking, because domain authority is a third-party metric, not something Google or any answer engine actually uses. Do not build a strategy on a number the search engines cannot see.

Page speed (F). It belongs at the bottom. A fast site that says nothing worth quoting gets quoted by nobody. Speed is a tie-breaker between two equally good pages, never a reason one gets cited over a slow one with better answers. Fix it if it is genuinely broken, then never think about it again.

So I built you a scorecard

Agreeing with a ranking is easy. Knowing where your own site actually sits on it is the hard part, and it is the part that decides what you do next. So I turned the whole ranking into an interactive AI Search Strategy Scorecard.

You answer honestly for each strategy, doing it, partly, or not yet. It weights your answers by how much each one actually moves AI visibility, gives you a live score out of 100, and then, and this is the point, hands you a prioritised action plan. Your biggest gaps come first, each with the tool on this site that closes it. It is the difference between reading a tier list and running one against your own work.

It is free, there is no signup, and nothing you enter leaves your browser. It sits at the top of my answer engine optimization hub as the thing I would run first.

My Take: I nearly did not write this one, because the honest version of it is a little uncomfortable. Building the scorecard forced me to lay all my own tools against the ranking, and two things fell out. The good one is that I have a live tool for every strategy in the S, A, B and C tiers, which told me I have been building toward the right target rather than just building fast. The uncomfortable one is that my own site sits at DR 29 and ranks nowhere, which is the exact trap the bottom of this list warns about. I had been quietly proud of that DR number. The ranking made me admit it is worth nothing until the S tier is done. That is what a good framework does. It does not just tell you what to build, it catches you believing your own vanity metric. borja’s tier list did that for me. The scorecard is my attempt to do the same for you, without the ego getting in the way, because it just does the maths.