Less hype. Better software decisions.
Plainly stated software research for freelancers and small service businesses: verified pricing, documented limitations and an evidence level on every tool.
What readers can expect
Independent
No sponsored winner and no affiliate-first ranking logic.
Practical
Recommendations by workflow, business type and budget.
Honest
Limitations and evidence gaps are part of every useful review.
Why this site exists
Most “best AI tools” pages are long lists that never help the reader decide. They rank by affiliate payout, restate marketing copy as analysis, and quote prices that stopped being true months ago. The reader finishes the page knowing more products and no more about which one to buy.
The Plain Stack is built as a decision system instead: a structured database where every tool carries an explicit evidence level, focused comparisons that end in a verdict, and pricing that is either verified against the vendor's own page or visibly masked.
Who this is for
Freelancers, consultants, small agencies and service businesses — people who buy software with their own money and feel every unnecessary subscription. The recommendations assume a small team without a procurement process, not an enterprise with a software budget.
What we publish
- Tool pages with verified pricing status, documented limitations and an explicit evidence level.
- Head-to-head comparisons that name the audience each tool wins for, rather than declaring one winner.
- Guides for building a small, coherent stack — including when the answer is to buy nothing.
- An open record of what we could not verify, and why.
How this is funded
Currently, it is not. The site carries no affiliate links, no sponsored placements and no advertising. If that changes, the affiliate disclosure will be updated before any monetised link goes live, and every affected page will carry a visible notice.
The editorial rule we intend to keep is the one that matters: affiliate availability never determines whether a tool is included or how it ranks. Two tools in our database currently have no verifiable pricing, and they stay in with that stated openly rather than being quietly dropped.
What we get wrong
Two things, predictably. Prices go stale — every price here carries the date it was checked, and AI vendors change tiers faster than any review site can track. And we do not hands-on test everything; where we have not, the evidence level says so rather than implying otherwise.
The methodology page sets out exactly what each evidence level means. If you find an error, the contact page is where corrections will be handled once the mailbox is live.
The database today
20 tools are currently researched, 18 of them with pricing read from the vendor's official page. Growth is deliberately slower than the verification work — a larger database that nobody has checked would be worth less than a small one that has been.