Real reviews
Real reports
These are the documents the panel issued. The file was not rebuilt or polished for the website, so its hash still matches our registry — download it and check for yourself on the verification page. Every report on this page was written in English by the panel; Russian-language reviews are shown on the Russian site.
Your report will not end up here
Customer cases are published with their consent: the founder asks each of them personally and the permission is recorded in our registry — without that record a report is never published, however good it looks. The bot never asks.
The documents carry no identifiers by design: the service
never asks for a company name, registration details or personal names, so
the panel never receives them. What happens to your materials —
in the privacy policy; you choose the
retention mode with /privacy in the bot.
Assets / real estate
One problem property. Sell the entire €1.48M portfolio at a 22% discount?
One distressed asset did not automatically justify selling three.
The owner of three mortgage-free Barcelona flats had one of them occupied without consent while vacant between tenancies; counsel estimated a civil route of 8–14 months at roughly €390 a month in charges. A fund offered €1.16M for all three, occupied flat as is, 30-day close — against a 2024 valuation of €1.48M. The panel does not oppose selling. It opposes selling three assets because of a problem affecting one, at one blended price that has not been allocated property by property: the two let flats bring in €3,050 a month, and no evidence produced shows their risk is comparable.
What the panel found
First test. Send the fund one written request: a binding allocation of the €1.16M across the three flats, the buying entity, all conditions, deposit and proof-of-funds position, and its all-cash offer for the occupied flat alone on the same timetable.
What changed after the review. The question «is waiting the bigger risk?» was split: it may be for the occupied flat, it is not yet shown to be for the two let flats. The all-three sale is not evaluated until a binding per-flat allocation, a standalone offer for the occupied flat and a written note from Spanish counsel on the actual legal route are on the table.
Business / M&A
€7.8M on the table. Sell 65% of the company — or keep control?
Headline valuation was not the same thing as the economics of control, earn-out and retained ownership.
A founder, 47, with more than 80% of his wealth in a €6.2M-revenue B2B SaaS company, was offered €7.8M for 65%: €5.8M at closing and up to €2.0M of earn-out tied to ARR and EBITDA after costs the buyer would control, a 36-month CEO lock, a 7× EBITDA put/call on the remaining 25% with bad-leaver clauses, and 21 days to sign. His goal was liquidity, less risk, less operating load and a share of the upside. The panel did not attack the goal. It found that the structure reliably delivers only the first: the earn-out, the leaver definition and the deadline were all unwritten or untested, and it sent the package back for rework before anything is signed.
What the panel found
From the findings, verbatim:
- The earn-out and the put/call both refer to EBITDA after costs and decisions the buyer will control: budget, CFO hire, pricing, shared-services allocation, and part of marketing.
- Bad-leaver mechanics apply only to voluntary early departure. The client has not seen the contractual definition. UNVERIFIED
- The stated 21-day deadline is unverified, untested, and the client has not asked whether a 45-day timetable is possible.
From “Resolved”: Send a written request to the buyer's counsel for the earn-out EBITDA calculation methodology, including shared-services allocation, intra-group charges, marketing allocations, accounting policies, measurement timing, seller information rights, audit rights, and dispute procedure. Deadline: 48 hours.
First test. Send one written request to the buyer's counsel within 48 hours for: (1) the earn-out EBITDA calculation methodology, including shared-services and intra-group cost treatment; and (2) the complete good-leaver/bad-leaver definition, including removal and termination provisions. The buyer's response or refusal is the most informative currently available fact.
What changed after the review. Before review: "accept the strategic offer and sign the term sheet — am I right?", framed as a price-and-liquidity decision on a €7.8M headline. After review: not principally a headline-price decision but a control-and-contract-definition decision. The reliably specified amount is €5.8M cash at closing; the €2.0M earn-out and the value of the retained 25% remain contingent on provisions not yet seen in writing.
Personal / career
A 55% pay rise abroad. Move the whole family — or test the move first?
The decision was not necessarily "move or stay": some of its irreversible parts could be separated and delayed.
A VP Product in London, 41, on £155,000 plus bonus, was offered a senior role in Singapore at the equivalent of £240,000, with housing and school allowances, a three-year contract, six months' probation and 14 days to answer. His wife, an architect on £92,000, had no job there; the children were 8 and 13, the elder ten months from a key exam year; a 76-year-old mother lived nearby. The couple had discussed only two options: everyone moves, or decline. The panel did not oppose the move. It found the plan commits four people irreversibly on eight load-bearing facts, none of which had been checked — the contract's exit terms, the staged-start option, the wife's actual red line, the household arithmetic — and sent it back for rework.
What the panel found
From the findings, verbatim:
- The contract that governs the entire downside — probation notice, severance, clawback of relocation, housing and school money — has not been seen; only the offer summary exists.
- The choice being decided ("everyone moves or we decline") was never tested: no staged-start question to the employer, no direct question to the wife about whether her condition is a limit, no extension request on a deadline of unknown origin.
- "Build capital" is, in the client's own words, "an impression, not a spreadsheet": the £92,000 spousal income loss, the school shortfall and the UK property drag have not been netted against the uplift.
From “Resolved”: Request the full employment contract text — specifically the clauses on probation notice for both parties, termination, severance, bonus eligibility and timing, and any repayment obligation for relocation, housing, school or sign-on money on termination inside 12 and 24 months. Deadline: 3 days.
First test. One written question to the hiring manager — "Is a delayed start after the elder child's exams, or a single-status first year with the family joining in year two, available?" — plus a request for the full contract text. Both inside 48 hours, both cost nothing, and together they settle the largest open assumption in the report.
What changed after the review. Before review: "accept the offer and move the whole family in four months — am I right?" After: the question is not "go or not go" but "on what sequence, and under which contract clauses" — and neither the employer nor the wife has been asked the questions that determine the answer.
Verdict, panel size, counts and dates come from our hearing log, not from prose written by hand.
How the format changed: named roles before 19 August, a single file before 21 August
Reports issued before 19 August 2026 gave the panel's roles synthetic professional profiles such as “former operating director” or “15 years in due diligence”. Those were model role instructions, not biographies of real people and not a claim that human experts participated. The panel builder itself worked that way, and we found it in our own audit. Since that day a panel member is a mandate — Mandate “Name” — described by what it tests.
Hearings before 21 August 2026 were delivered as one document.
Now there are two: the report carries the whole review, and the
transcript holds the raw statements of each mandate before
synthesis. The transcript belongs to the higher plans and is delivered
to the buyer via /materials; it is never published on the
samples page, and its fingerprint is printed inside the report.
We do not silently edit the old PDFs: changing a word changes the file hash, and the file stops being the original that was delivered.
What “original” means
A printout, screenshot or copied passage may be useful for reading, but file verification concerns the issued PDF. A matching hash proves that the file is unchanged. It does not prove that the analysis is correct.
Why the flaws stay visible
If an early report contains a formatting defect, a superseded method rule or an error we later identified, the original file remains unchanged and the correction is documented next to it. An immutable record should not become a polished retrospective.
Where Opponentura gets it wrong →
Want your own decision checked the same way?
Start a review — first verdict free First check whether you need one