ORPHANWARE The three prices, and what each one buys
Recon Report: $45. A full inventory of your binary: what the program does, its command line interface and flags, the file formats it reads and writes, and an honest feasibility verdict on whether full recovery is realistic. Delivered in about 3 days. This is where every engagement starts.
Source Recovery: $120. Everything in the recon, plus readable recovered source code and a recompiled working binary with compile verification. Delivered in about 5 days, for scope a recon has already agreed.
Verified Recovery: $250. Everything in Source Recovery, plus the full differential test report: the rebuilt binary run against your original across thousands of generated cases, outputs compared exactly, with documented edge cases and known limitations. Delivered in about 7 days, for scope a recon has already agreed.
There is no hourly billing and no open-ended quote. If your program is larger than a single fixed tier covers, or you have several programs, the recon produces a written quote for the larger scope before you commit to it. You approve the number first, or you walk away with the recon and the documentation it contains.
What drives the price of any recovery
Whether you hire us or anyone else, these are the factors that decide what a recovery costs. Anyone quoting you should be able to say which of them apply to your binary and why.
- Program size. A 40,000 line utility and a 400,000 line business system are different jobs. Size is the single biggest driver, and it is measurable from the binary before work starts.
- Language and toolchain. Some compilers and languages leave more recoverable structure than others. Visual Basic 6 and managed code recover differently from optimized native C++, and an honest quote reflects that.
- Symbols. A binary that still carries debug symbols is faster to reconstruct faithfully. Most production binaries are stripped, which is normal and expected, but it is real work.
- Packing and protection. Standard commercial licensing protection is ordinary. Deliberately obfuscated binaries need an extra stage before decompilation can start.
- Verification depth. A rebuild that is read and spot-checked is cheaper than one that is differentially tested across thousands of generated cases with the results documented. Verification is most of what separates a recovery from a decompiler dump, and it is priced accordingly here: it is the difference between the $120 and $250 tiers.
- Documentation. File format documentation, a behavior report, and build notes take time to write well, and they are a large part of what makes the result maintainable afterward.
Why most recovery pricing is a mystery, and why ours is not
Most reverse engineering is sold the way consulting is sold: describe your problem, wait for a per-project quote, and learn the price after you have explained yourself to a salesperson. The quote is high partly because the seller cannot see your binary yet, so the unknown risk gets priced in, and the buyer carries it.
The recon-first structure exists to remove that risk from your side of the table. For $45, the unknowns about your specific binary get measured: its size, its structure, its formats, its realistic recovery path. After that, the price of the full job is a known quantity quoted against facts, not a guess padded against surprises. If the verdict is that recovery is not realistic, you learned it for $45 and you still hold documentation of what your program does, which every alternative path needs anyway.
What to compare the price against
Recovery is rarely the only option on the table, so compare it against the real alternatives, not against zero:
- Migration or rewrite. Moving a legacy application to a new platform is a project, not a purchase. Published pricing for even simple Delphi migration projects runs to $10,000 and up, and a rewrite of a business system costs more, before counting the business rules that get lost because they existed only in the old program.
- Replacement software. A modern substitute means licenses, data conversion, retraining, and usually a changed workflow. Sometimes that is the right answer. It is never the cheap answer it first looks like.
- Doing nothing. Free until the day the machine, the operating system, or the one person who understands the program fails. Then the same recovery question returns, under time pressure, possibly with a damaged binary.
Against those numbers, a $45 verdict and a fixed-price recovery in the low hundreds is the inexpensive door. It is priced that way on purpose: the pipeline is automated where automation is reliable, and a named operator directs and signs off on the result.
When recovery is not worth paying for
Honesty about the negative cases is part of the product. Recovery is probably not your best spend if the program is a small utility with a living modern equivalent, if the original vendor is still operating and will sell or escrow the source, or if the system is already scheduled for replacement and only the data matters, in which case data extraction is the real job. The Recon Report says so plainly when one of those is true. A verdict of "do not recover this" is a successful recon, not a failed sale.