How We Compare Launchpad Tools
This site compares classes of tool rather than products, which is an unusual choice in a market where every other page is a ranked list. This is the method: how a class is defined, what counts as evidence for a claim, how cost gets described, and the four things the desk will not do at any price.
Why classes rather than products
A ranked list of named products is the obvious way to write about this market and the least useful. Products in launchpad tooling appear, rename themselves, change operators and disappear on a timescale of weeks, so a ranking is stale before it is indexed. Worse, a ranking implies the writer tested the products, and almost nobody who publishes one has.
Classes are stable. A bundler in a year will still buy ordering at a launch moment, a volume tool will still consume fees per leg, and a wallet manager will still decide who holds keys. Writing at that level means a page stays true for as long as the behaviour exists, and it hands the reader a way to evaluate whatever product is in front of them today rather than the one that existed when the article was written.
It also removes the incentive that ruins most writing in this space. A site that ranks products has something to sell to the products it ranks. A site that describes classes has nothing a seller wants to buy, which is the point.
How a class gets defined
A class earns a page here when it has a distinct answer to four questions. Marketing categories that collapse into an existing class do not get their own page, however many products use the name.
- What does it produce? A concrete object or record: a supply position, executed trades, funded wallets, posted messages. Not an outcome like attention or momentum.
- When is it relevant? A specific moment or period. Several classes expire at the launch, and a class whose timing cannot be stated is usually two classes wearing one name.
- What does it consume? A cost shape, not a price. Which fees are unavoidable, what is locked rather than spent, what the seller's margin sits on.
- What is left visible? Whatever a reader can find afterwards. If the answer is nothing, the class is probably a read-only tool and belongs in a different category entirely.
Applying that test is why this site names five classes rather than the dozen a search will return. Trending pushes collapse into visibility tooling because the object produced is the same manufactured signal. Copy traders, alert bots and dashboards fall out because they read rather than act, which changes the risk profile enough that grouping them with the rest would be misleading.
How a claim gets tested
Claims on this site fall into three buckets and each is handled differently. Keeping them separate is most of the discipline.
| Kind of claim | Standard it has to meet | If it cannot meet it |
|---|---|---|
| Protocol behaviour | Documented publicly and checkable by a reader against the same source | Stated as a general property without the number, or left out |
| Market behaviour | Describable as a pattern a reader can verify by looking themselves | Written as an observation with its limits attached, or dropped |
| Effectiveness | A method the desk actually ran, with the method published | Not published at all, because we have not run one |
The third row is the reason there are no success rates, conversion figures or benchmark tables anywhere on this site. We have not measured how well any of these tools work, so we do not publish numbers about it. Where a figure appears, it is either a protocol constant such as the base transaction fee, or clearly labelled illustrative arithmetic built from round numbers you are invited to replace with your own.
How cost is described
Every class note describes a cost shape instead of quoting a price, and the reason is practical rather than coy. Prices in this market vary by seller, by week and by how the conversation is going, so a published price is wrong immediately and useful to nobody. The shape of the cost, which parts are network fees, which are venue fees, which are locked capital and which are the seller's margin, changes far more slowly.
Where arithmetic appears, three rules apply. The inputs are stated so a reader can redo the sum. Invented figures are labelled as invented. And the conclusion drawn is about the ratio rather than the absolute number, because the ratio survives a change of inputs and the absolute number does not.
What counts as evidence
The standard applied to sellers on the buying checklist is the standard this desk applies to itself: anything asserted should be checkable by a reader without our help. Protocol facts point at public documentation. Market patterns are described so a reader can look for them in an explorer. Judgements are labelled as judgements, in the same sentence, rather than in a disclaimer at the bottom.
Ordinal judgements do appear, particularly in the scoring sheet on the buying checklist. Those are the desk's opinions about relative weight, stated as opinions. The distinction between "this is heavier than that" and "this is 3.7 out of 5" is the whole difference between a defensible judgement and a fabricated measurement.
What this desk refuses
Four refusals, written here so a reader can hold the site to them and notice if a page ever drifts.
- No ranked vendor lists. No leaderboard, no top ten, no "best of" ordering of named products. The comparison work happens at class level and the vendor decision stays with the buyer.
- No paid placement. No page position, mention, ordering or wording is for sale, and no seller has been offered any of those things.
- No ratings, review counts or testimonials. Nothing on this site claims that a number of people evaluated anything, because no such evaluation happened. Fabricating one would be the same act this site criticises the visibility class for.
- No operating guidance for deceptive tooling. The comment and visibility page describes what the class is and how readers detect it, and carries no setup advice, no seller checklist and no help making the output convincing.
Where links point and why
Some pages link to commercial tools where the paragraph is already about the thing the linked page addresses. Those are ordinary editorial links to public pages, placed because a documented product is a more useful reference point for a buyer than a description written from the outside. A link is not a recommendation, no seller pays for one, and nothing on this site should be read as a verdict on whether a particular product suits your situation.
Links to protocol documentation exist for a different reason: so a reader can check a claim at its source instead of taking our word for it. Where a page states a constant or describes how something works on Solana, the documentation root is linked and the reader is expected to use it.
Corrections
Errors get fixed in the text rather than patched quietly, and a correction that changes what a page claims is made plainly rather than presented as though the page always said the right thing. The most useful corrections name the page, quote the sentence and say what the actual behaviour is; those can be sent through the contact page.
Two categories of message get a different answer. Requests to be added to a ranking are declined because no ranking exists. Requests to remove a critical description of a class are declined unless the description is factually wrong, in which case the fix is the correction rather than the removal.
What this site is not
It is not a review site, not a directory and not a broker. It does not test products, does not accept products for testing, and does not have a relationship with any seller mentioned in a class note. It gives no financial, legal or tax advice, and nothing here is a judgement on whether any token, launch or strategy is a good idea.
What it does is narrower and, we think, more useful: it names what each class of tool actually does, describes what it costs in a way that survives a change of price list, and hands over the questions that tend to end badly when nobody asks them.
Common questions
Do you rank the tools you write about?
No. There is no leaderboard, no scoring of named products and no best-of ordering anywhere on this site. Products change too quickly for a ranking to stay true, and publishing one would create exactly the commercial relationship with sellers that this method is designed to avoid.
Do sellers pay to appear here?
No. No placement, mention, ordering or wording on this site is for sale, and none has been offered to any seller. Where a commercial page is linked, it is because the paragraph is already about the subject that page addresses.
Why are there no ratings or review counts?
Because no reviews were collected and no evaluation took place. Publishing a rating would mean inventing a measurement, which is the same act this site criticises elsewhere. Judgements appear as judgements, in plain words, without a decimal point attached.
Where do the numbers on this site come from?
Two places only. Protocol constants such as the base transaction fee and account rent minimums, which are publicly documented and checkable. And clearly labelled illustrative arithmetic built from invented round figures, which readers are told to replace with numbers from their own quote.
Will you review a product if it is sent to you?
No. The desk does not accept products for testing, does not take briefings and does not publish verdicts on named tools. Questions about a specific product are best answered by putting the buying checklist to its seller and reading the replies.
How do I report an error?
Through the contact page, naming the page and quoting the sentence you believe is wrong. Corrections are made in the text rather than quietly patched, and a change that alters what a page claims is made openly rather than presented as though it had always said so.