AI tools directory: discover, compare, and choose the best AI products

An AI tools directory helps you find, compare, and shortlist AI products for a specific job without bouncing between dozens of reviews, product pages, and stale top-10 lists. The best AI tools directory doesn't just collect names. It gives you enough context to judge fit fast: use case, pricing, category, and whether a product was built around AI or just had it bolted on later.

If that sounds useful, it's because most people aren't looking for a directory in the abstract. They're trying to solve a buying problem under time pressure. Maybe you need a coding assistant, a customer support agent, a design tool, or a research product this week. Instead, you open five listicles, see the same near-duplicate products, and still can't tell which options are credible, current, or meaningfully different.

That tool overload is real, and static blog posts are struggling to keep up. In fast-moving categories, directories matter more because they work as a reference index instead of a frozen roundup. A strong index filters for signal. It cuts down tool bloat, clarifies category boundaries, and helps you compare products the way buyers actually shop: by job to be done, role, price, free tier, and product scope.

This guide follows that decision process. We'll look at what makes an AI tools directory trustworthy, which filters matter, how to tell when a listing is mostly AI marketing, and when a directory is more useful than scattered recommendations from forums or videos. That matters because a directory is only as good as its editorial method.

At AgentsIndex, the working assumption is simple: AI is the noun, not a feature. Listings are checked against source pages, pricing, docs, and product positioning so the index reflects a fast-changing market as closely as possible. If you already know the job you need done, browse the index by category and go straight to the shortlist.

What an AI tools directory should actually help you do

An AI tools directory should help you do three things quickly: find tools that fit your job, compare them on the details that affect a buying decision, and rule out products that only mention AI without being built around it. In plain language, an AI tools directory is a structured catalog of AI-native tools and AI-first products, organized so you can browse by category, use case, role, pricing, or product type instead of opening 20 separate tabs and piecing the market together yourself.

That sounds obvious, but the category gets muddy fast. Some pages call themselves directories when they are really affiliate roundups. Some act like marketplaces even though they don't handle transactions. Some recommendation threads are useful, but they're not built for side-by-side comparison. If you're trying to choose between writing tools, coding tools, customer support agents, or workflow builders, those differences matter because each format solves a different part of the problem.

A directory is for structured discovery. It should let you scan a reference index, narrow the field, and understand what each product does, who it is for, and how it is priced. A review post or roundup is different. That format usually reflects one author's shortlist, with more opinion and less coverage. It can help when you already know the category and want a curated take, but it's weaker when you need breadth, clean filtering, and a way to compare many options on the same frame.

A marketplace is different again. Its job is to facilitate a transaction, installation, or in-product distribution. Marketplaces are often optimized around listings, rankings, and conversion paths inside an ecosystem. They can help, but they usually show only what is sold or distributed there. A community recommendation thread solves another problem entirely. It gives you nuance, edge cases, and real-world tradeoffs from people who have actually tried the tools. That's valuable when you want to hear where a product breaks, what support feels like, or whether a free tier is enough. But threads are messy by design. They're hard to search, inconsistent, and rarely updated into a clean comparison set.

This is why good buyers use all four formats differently. Use a directory to get the map. Use community discussion to pressure-test your shortlist. Use the vendor site to verify claims, see docs, and understand the latest pricing. Use review content for a sharper editorial take when you're down to a few finalists. The problem starts when a page tries to be all of these at once and does none of them well.

Why the category boundary matters

The most useful filter in this space is whether AI is the noun, not a feature. In other words, is the product fundamentally an AI product, or is AI just one add-on inside a broader SaaS tool? That distinction matters because buyers are often searching for AI-native tools with workflows, pricing, interfaces, and product roadmaps centered on AI. If a directory mixes those products with every SaaS platform that added a chatbot, the index gets noisy. You stop comparing like with like.

People also shop in more specific ways than most generic listicles allow. They want filters for task, role, pricing, free tier, and whether the product is truly AI-first. A designer browsing image generators is making a different decision from an engineering team comparing coding agents or a support lead looking at AI chat tools. Directories win on speed when they mirror that buying behavior. Communities win on nuance. The best research process combines both.

What makes a directory trustworthy

Trust starts with source-backed verification. A listing should be checked against the vendor's own homepage, pricing, docs, or changelog, not copied from another blog. That's how you avoid stale descriptions and made-up comparisons. Update cadence matters too. AI categories change too quickly for a directory to sit untouched for six months, especially when pricing, free plans, and product scope change constantly.

Category clarity is another signal. If a site can't explain why a tool belongs in one category instead of another, the taxonomy is probably loose. Pricing visibility matters for the same reason. Even a simple note like free, freemium, usage-based pricing, or plans start at a certain point gives the reader a better first-pass filter. Editorial independence matters most when sponsored placement exists. Paid visibility isn't the problem by itself. The problem is when sponsorship changes what the listing says, hides tradeoffs, or turns the page into ad inventory in disguise.

Picture a product lead opening a search result for "best AI tools," hoping to narrow a shortlist before a team meeting. Instead, they land on a page crammed with banners, recycled descriptions, and rankings that feel bought. After ten minutes, they still don't know which tools have a free tier or which category each product actually belongs to. That's the ad-farm listicle problem in one scene. A real directory should lower decision effort, not increase it.

A simple way to visualize the difference is as a map with three paths. One path is browse catalog: fast, structured, filterable. Another is ask community: slower, messier, but rich in tradeoffs. The third is vendor site: narrow, self-reported, and best for final verification. A trustworthy AI tools directory sits at the center of that map. It isn't the only input, but it's the fastest clean starting point when you need to discover, compare, and choose with less noise.

That is also the standard a directory like AgentsIndex should meet. The point isn't to be the loudest list on the page. It's to maintain an edited catalog of AI-native tools, make category boundaries clearer, surface pricing and use-case information quickly, and give founders a visible path to submit a product for review when the index misses something worth including.

Which directory features matter most when you compare AI tools

The fastest way to compare AI tools directories is to ignore the homepage hype and score the directory on seven things: filter quality, category depth, pricing visibility, source verification, update cadence, editorial scope, and sponsorship labeling. A directory with 5,000 listings can still waste your time if you can't narrow by task, role, free tier, or fit.

That matters because most people are not shopping for "more AI tools." They are trying to solve a job: find a coding assistant for a small dev team, shortlist AI meeting tools with a free plan, or compare customer support agents that fit a specific budget. The better directory is the one that gets you from broad curiosity to a credible shortlist in minutes, not the one with the loudest claim about being the biggest.

Use these criteria as a quick filter before you spend time trialing any tool. If a directory helps you cut your list from 200 options to five realistic candidates, it's doing its job.

Start with filters, not listing count

Raw listing count is a vanity metric. It sounds useful, but it rarely matches how buyers choose software. In practice, the filters matter more than the size of the catalog because they determine whether you can separate signal from noise. Social research around AI tool discovery keeps circling back to the same point: people want filters that match how they shop, not how directories market themselves.

The most useful filters usually map to five buying questions. First, what task are you trying to complete: coding, research, design, support, sales, writing, automation, or education? Second, who is the tool for: founder, marketer, developer, support lead, recruiter, or operations team? Third, what is the pricing model: free, freemium, usage-based, or seat-based? Fourth, is there a free tier or trial? Fifth, does the tool belong in this category at all?

Those filters save time because they remove false positives early. If you need a free AI transcription tool for a solo creator, a giant directory without free-tier filtering is less useful than a smaller edited catalog that shows pricing upfront. If you want AI-native support agents, a directory that mixes them with legacy help desk products will blur the field and make comparison harder.

This is also where category labels, use-case tags, and best-for guidance matter. Category labels tell you where a tool sits in the market. Use-case tags tell you what it can actually help with. Best-for guidance adds the buyer context most directories skip, such as whether a product fits solo users, startups, enterprise teams, or technical operators. Without that third layer, readers still have to infer fit from scraps.

A directory doesn't need to win every category to be useful. But if it fails on filters, pricing visibility, and update cadence, it will probably cost you more time than it saves.

Treat stale data as a trust problem

One of the most consistent user concerns is that directories go stale fast. That's not a minor maintenance issue. It's a trust factor. AI products change names, pricing, feature sets, category fit, and target users fast enough to break static list pages surprisingly quickly. A listing that was accurate 60 days ago can now point you to the wrong plan, use case, or category.

Stale data creates two problems. First, it slows down comparison because you have to re-verify everything on the vendor site. Second, it undermines confidence in the rest of the directory. If the first three listings you open have wrong pricing or outdated positioning, you stop trusting the catalog as a research tool.

Look for visible signs of update cadence. Does the directory show when a listing was reviewed? Does it explain how listings are built? Does it check source pages, docs, pricing, or changelogs? Does it revisit products when they change meaningfully? These signals matter more than glossy design because they tell you whether the catalog is an ongoing resource or a one-time content dump.

For a reference index like AgentsIndex, this distinction is central. The point isn't just to publish a giant list of names. It's to maintain an edited catalog of AI-native tools, checked against primary sources, so readers can compare what the index contains today rather than what it contained last quarter.

Editorial scope keeps the catalog coherent

Another feature that matters more than people expect is editorial scope. In a fast-moving category, a directory needs rules for what gets indexed and what does not. Otherwise the catalog turns into a loose pile of AI chatbots, browser extensions, agencies, prompt packs, courses, and legacy SaaS products with a thin AI layer.

A clear scope helps readers compare similar products on fair terms. If you are evaluating AI-native tools, you want a directory that distinguishes them from software where AI is just an added feature. That boundary isn't always perfect, but the effort to define it is already a trust signal. It shows the editor is making a category judgment instead of chasing search volume with vague "top tools" pages.

This is also why best-for guidance is useful. Two tools can belong to the same category and still fit very different buyers. One may be best for solo builders who want a generous free plan. Another may be best for mid-market teams that need approvals, security controls, and integrations. Category labels alone do not explain that difference. Good directories do.

What should you ignore? Start with vanity counts, vague claims about "the best AI tools," and any sponsored placement that isn't labeled. A serious directory can sell sponsorship and still remain useful, but readers should never have to guess whether a ranking exists because of editorial judgment or payment. Clear labels protect trust. Hidden promotion destroys it.

The same skepticism applies to generic top-50 listicles. If every tool sounds interchangeable and every description reads like vendor copy, you're not looking at a comparison resource. You're looking at recycled packaging.

Use directories and community feedback together

Readers often frame the choice as browse a catalog or ask a community. In practice, the best workflow is both. A directory is better for structured discovery: narrowing by category, pricing, role, and product type. Community feedback is better for edge cases: hidden limitations, onboarding friction, support quality, and how a tool holds up after the demo.

A practical approach looks like this. Use a directory first to build a clean shortlist of tools that match your task, budget, and team context. Then use community feedback to pressure-test the shortlist. That second step helps you catch issues the listing may not cover, such as weak exports, poor reliability, aggressive upsells, or a steep learning curve.

That balance is especially helpful when shopping in a noisy category. Directories give you structure. Communities give you texture. You don't need to pick one source only; you need to know what each source is good at.

Video roundups can help here, not because one creator can settle the market, but because they reveal how people compare tools in the wild. They often surface the same buyer questions found in communities: what is worth paying for, what still feels immature, and which tools hold up after the first week of use.

The bottom line is simple. When you compare AI tools directories, judge the research system before you judge the catalog size. Better filters, clearer pricing, source-backed verification, visible update cadence, and honest editorial scope will help you find high-signal products faster than any giant unlabeled list. Use those criteria as your quick screen, then spend your trial time only on the tools that survive it.

How to compare AI tools by category, pricing, and real fit

The fastest way to use an AI tools directory well is to compare tools for the job, not by hype or how often a name shows up in listicles. Start with one workflow, group the options by category, cut near-duplicates, then compare pricing, integrations, and team fit. That sounds obvious, but it solves the complaint that comes up over and over when people browse AI products: too much noise, not enough help narrowing the field fast.

A good AI tools directory should make that filtering easier, not harder. If you have to open 20 tabs just to learn which products are writing tools, which are research tools, and which are really automation products dressed up with AI copy, the directory is failing its main job. The point is to move from browsing to a shortlist you can test in one sitting.

Start with category, but do not stop there

Category is your first filter because it removes obvious mismatches. Most buyers want a place where they can quickly compare tools for writing, research, automation, design, coding, or agents. That is useful, but category alone is still too broad. A writing tool for drafting blog posts is different from a writing tool for sales emails. A coding tool for autocomplete is different from an agent that can plan and execute multi-step work. If you stop at category, you still end up with tool bloat.

A better approach is to pair each category with the core job you need done. Ask one blunt question: what do I need this tool to do this week? Not in theory, not someday, not across the whole company. Just this week. That framing strips out a lot of noise.

This table isn't a buyer's guide by itself. It's a sorting frame. It helps you ask better questions inside a directory: which category matches the work, what pricing model am I signing up for, what does the free tier allow, and who is this built for? Those filters matter more than generic claims about productivity.

Use a four-step shortlist method

Once you have the right category, use a simple shortlist method.

1. Define the job in one sentence. Example: "We need a tool that turns interview transcripts into publishable research notes." If you can't define the job clearly, every tool will look worth trying, which is exactly how teams end up overbuying.
2. Remove near-duplicates. If three products do almost the same thing with the same pricing model and workflow, keep one. Don't waste evaluation time comparing clones.
3. Compare pricing with actual usage in mind. A cheap monthly plan can get expensive if the useful features sit behind hard usage caps, seat minimums, or paid integrations.
4. Test workflow fit. Run one real task through each finalist. Not a demo prompt. Not the homepage example. Your actual workflow.

That last step matters most. Directories help you discover and compare, but they can't tell you whether a tool fits the way your team already works. Only a live test can do that.

What pricing tells you, and what it does not

People often say they want to narrow by price and category fast, and they're right. Pricing is one of the best early filters in an AI tools directory because it exposes who the product is designed for. A usage-based product may suit occasional, high-value tasks. A seat-based plan may make more sense for a team that needs steady access. A credit system can be fine, but it often hides the true cost of repeated use.

Free versus paid is even trickier. A free tier tells you the product wants low-friction adoption. It does not tell you the product is affordable at scale, complete enough for production work, or serious about team use. In practice, a free tier tells you four things: how quickly you can test the core workflow, whether the onboarding is confident, where the paywall begins, and which capabilities the company considers premium.

That is why "has a free tier" should never be the deciding filter by itself. Some free plans are generous enough to prove real value. Others are just a gated demo. If a free version blocks exports, integrations, collaboration, or the main model quality, you're not testing the product you would actually buy.

This is also where an edited index is more useful than a generic roundup. On the home page of a curated directory, the useful question isn't only whether a tool is listed. It's whether the listing makes pricing, scope, and category boundaries clear enough to compare without guesswork.

Mini-story: one workflow, three tools, two easy rejections

Take a small content and operations team trying to solve one workflow: turn long customer interview transcripts into a short internal brief, three sales takeaways, and one publishable insight. They shortlist three tools from the same directory view: one writing tool, one research tool, and one agent product.

At first glance, all three seem worth trying. The writing tool is polished and inexpensive, but it mostly rewrites text after the team has already organized the material. It helps at the end of the workflow, not at the messy beginning. The agent product is impressive in demos, but it needs more setup, more oversight, and more integration work than the team can justify for a weekly task.

The research tool wins because it matches the job more closely. It can handle source material, synthesize themes, and produce a first-pass brief the team can refine elsewhere. The team rejects the other two, not because they are bad products, but because they are the wrong fit. That is the mindset most buyers need. You're not picking the most impressive tool in the abstract. You're picking the one that removes the most friction from one real workflow.

That story also explains why category filters and pricing filters should work together. If that same team had filtered only by lowest price, they might have picked the writing tool. If they had filtered only by novelty, they might have picked the agent. The better choice sat in the middle: not the cheapest, not the flashiest, just the best workflow fit.

How to judge real fit before you commit

Real fit usually comes down to a few practical checks.

• Does it reduce steps, or add orchestration overhead?
• Does it work with the tools your team already uses?
• Can a non-expert get value in the first session?
• Is the output good enough to keep, or do you rewrite everything anyway?
• Will the pricing still make sense after the trial period ends?

If the answer breaks down on two or three of those checks, the tool is probably browse-worthy but not shortlist-worthy. That distinction matters. A directory should help you discover options, but the real value is helping you reject bad fits faster.

That is also why trust matters. Many readers are rightly skeptical of AI tool roundups that are really ad inventory in disguise. A useful directory needs clear scope, current listings, and an editorial standard for what gets included. If you're comparing options in a space this crowded, you need confidence that the category labels, pricing notes, and product descriptions are current enough to support a real decision. If you build or run an AI-native product and want it reviewed under that standard, you can submit it for consideration rather than trying to force-fit it into a stale list.

The practical takeaway is simple: shortlist three tools max for the same job, then test them on one real workflow this week. That keeps the comparison honest, limits tool bloat, and gives you a much better read on value than another hour of browsing.

Why trusted curation matters more than having the biggest AI tools list

A directory with 10,000 tools is not automatically more useful than one with 1,000. In AI, raw volume often creates the opposite problem: more noise, more duplicates, more stale pricing, and more products that barely qualify as AI-first. If your job is to discover and compare real options quickly, a smaller index with tighter standards usually beats a giant list that accepts everything.

That difference matters because most buyers are not browsing for entertainment. They are trying to answer practical questions: Which tool fits this workflow? Does the free tier still exist? Has pricing changed? A broad catalog can help with discovery, but an edited reference index is better for decision-making because it filters before it floods.

This is why trusted curation matters. The strongest AI directories are starting to look less like endless link dumps and more like reference indexes. They give you category context, pricing clues, use-case framing, and enough editorial judgment to separate serious products from noise. That is closer to how people compare software now. They don't want 500 tabs. They want a shortlist they can trust.

Big lists create false confidence

A huge AI tools list can signal coverage, but it can also create false confidence. Readers assume that if a tool is listed, someone checked it. Often, no one did. On many directory pages, products are pulled in from old submissions, scraped descriptions, launch platforms, or recycled "best AI tools" roundups. The result looks comprehensive, yet the data may be thin, outdated, or copied from marketing pages without verification.

That is where skepticism about affiliate pages and ad farms comes from. Users have learned to spot the pattern: inflated rankings, vague descriptions, no clear inclusion rules, and suspiciously positive summaries that read the same for every product. If a page makes money from attention but tells you nothing about how listings are checked, treat every recommendation as provisional. Size alone does not create trust. Method does.

Edited indexes do the slow work buyers can't

The real value of an ongoing, edited catalog is not that it contains every tool. It is that it does the slow verification work a buyer doesn't have time to repeat for every vendor. A reliable directory should check a product's homepage to confirm what it actually does, pricing pages to see whether plans start where the listing claims, docs to verify integrations or technical scope, and changelogs to catch meaningful product shifts. That process sounds basic, but it's exactly what keeps an index useful after the first click.

In a fast-moving category, this matters more than in older software markets. AI products change positioning quickly. A writing assistant becomes a workflow agent. A chatbot platform adds coding features. A "free" product moves to usage-based pricing. Without re-checking source pages, a directory goes stale fast. Then the category labels drift, the comparisons break, and buyers waste time evaluating tools that no longer match the description they found.

That is also why scope discipline matters. A broad catalog may lump together AI-native tools, legacy SaaS with one generative feature, consulting services, prompt libraries, and random browser extensions. An edited index sets a tighter boundary. At AgentsIndex, the framing is closer to a reference index of AI-native tools and the tools that power them, not a catch-all software warehouse. That narrower scope makes comparison sharper because the category itself is cleaner.

Editorial independence is part of the product

There is another trust issue worth stating directly: people assume directories are monetized first and curated second. Sometimes that is fair. Sponsored visibility, affiliate incentives, and paid placements can distort what gets seen. The fix is not pretending monetization does not exist. The fix is separating commercial inventory from editorial copy and labeling paid placements clearly.

That separation is a quality signal in itself. An index can sell ads or sponsored placements and still remain useful if the labels are obvious and the written listing is not for sale. Readers are usually not offended by monetization. They're offended by hidden incentives. Editorial independence is not branding fluff here. It is part of what makes a directory usable as research instead of just distribution.

The same principle applies to submissions. A founder submitting a tool should be the start of review, not the end of it. A trustworthy directory homepage should make it easy to understand what gets indexed, and a submit flow should signal that inclusion depends on fit and verification rather than whoever asked first. That is how an ongoing, edited catalog stays credible over time.

So when you evaluate any AI tools directory, ask a simple question before you trust the rankings: can this publisher explain how listings are checked and updated? If the answer is vague, the list is just a list. If the answer includes source checking, category discipline, labeled paid placements, and editorial review, you're looking at something more valuable: a working reference index that helps you choose with less guesswork.

What the data says about AI adoption and why directories keep getting more useful

The market shift is the easy part to see. The hard part is keeping up with what it creates. AI adoption moved from 20% of organizations in 2017 to 47% in 2022, then 55% in 2023, according to McKinsey. That's not a slow, linear category. It's a compressed expansion curve, and every jump in adoption brings another wave of products, wrappers, copilots, model layers, and workflow tools into the market. For buyers, the result is simple: more options, less clarity.

That is why an AI tools directory matters more now than it did even a year ago. When a category is small, a handful of blog posts and a few bookmarked reviews can do the job. When the category starts changing every quarter, static content turns into historical context. Useful, sometimes, but not enough for active evaluation. People trying to choose an AI product usually don't want to read 15 separate listicles, open 20 tabs, and reverse-engineer pricing, use case, and product scope on their own. They want a faster filtering layer.

The chart matters because it explains the downstream buyer problem. Adoption growth does not just mean more companies using AI. It also means more vendors positioning themselves as essential, more overlap between categories, and more products that look similar from the outside but differ in pricing model, depth, integrations, or whether AI is actually the core product. In a crowded market, discovery is no longer a nice extra. It becomes part of the buying infrastructure.

Why directories become more useful in fast-moving categories

A fast-moving category needs an updated reference index, not just annual blog posts. An annual roundup can tell you what looked important at the time it was published. It can't reliably tell you what changed three months later, which products added pricing, which tools shifted upmarket, which categories got crowded, or which listings no longer belong in the same bucket. That gap matters more in AI because the category boundary itself is still unstable. Some tools are AI-native. Others are older SaaS products with AI features added on top. For a buyer, that distinction affects what should even make the shortlist.

This is also why many discovery meta-projects date quickly. Plenty of directories, roundups, and community-maintained lists were still framed around 2024 assumptions, categories, and pricing snapshots. That's not a criticism of the format so much as a reminder of the environment. In AI, the discovery layer ages fast. If the index is not maintained, the user pays the cost in extra research time and bad comparisons.

The practical need is straightforward. Buyers want to filter by use case, role, price, free tier, and whether a product is truly AI-first. They want to compare without guessing. They want to reduce trial-and-error cost before signing up, moving data, or onboarding a team. Static blog posts rarely solve that on their own because they are built to publish once. A reference index is useful when it is built to be checked, updated, and reorganized as the market changes.

That is the real advantage of a current AI tools directory. It does not replace deeper product research, demos, or community feedback. It compresses the first half of the process. Instead of asking, "What exists?" and "Which of these are actually comparable?" across dozens of pages, you start with a cleaner map of the market. In a category where product pages, pricing, and positioning change constantly, that map keeps getting more useful, not less.

Frequently Asked Questions

How to choose the best AI product without getting lost in the list

If you've spent an hour opening 12 tabs and still can't tell which product is worth testing, the fix is usually not more research. It is a better selection method. The best AI product is rarely the one with the loudest launch or the biggest directory footprint. It is the one that fits one real job, sits in your price range, and is listed in a directory that keeps scope tight and entries current. Readers want signal, not noise, and that means starting with a smaller, cleaner set of options instead of trying to scan the whole market at once.

This week, keep it simple. Pick one job you want help with, like prospect research, support triage, note-taking, or code generation. Then narrow the field by category and price. Shortlist three products that are clearly AI-native, not older software with an AI tab added on top. From there, test one workflow end to end. Don't ask whether the tool looks impressive in a demo. Ask whether it saves time, improves output quality, or removes a repetitive step you already hate doing.

That is why the best directory is not the largest one. A giant list with stale pricing, vague categories, and loose definitions creates more work for the buyer. A useful index does the opposite. It helps you filter fast, compare real options, and trust that the listing still reflects the product today. Trust comes from freshness, clear scope, and source-checked coverage. In a category this crowded, those details matter more than raw listing count.

A practical decision sequence looks like this: define the job, set a rough budget, filter to the right category, compare three realistic candidates, then run a live test on one workflow. If it works, keep going. If it doesn't, move to the next option without restarting the whole search. That turns an AI tools directory from passive reading into a working shortlist.

If you want a lower-friction place to start, begin on AgentsIndex and browse the category you care about first. You don't need to evaluate everything, and you don't need to commit on day one. Start with a few source-checked listings, compare what is actually relevant, and keep the index open as an ongoing reference as the market changes. If you build an AI-native product that belongs here, you can also submit it for review. The goal is not to win one search session. It is to keep a current, credible shortlist you can return to whenever the index today looks different from the index next month.