What do you actually get back from a Titleman run?

# What do you actually get back from a Titleman run? **A written answer in the room, almost always. A file, rarely.** Measured over the 1,427 runs that finished in the 60 days to 2026-09-17: 1,202 of them — 84.2% — produced no file at all. The answer was the room. That single sentence is the one most people are missing when they ask this, and it is worth saying before anything else, because the expectation it corrects is a deck. ## What a run returns, by the files it actually left behind Every figure in this table comes from one query against production, run on 2026-09-17, over runs that finished in the previous 60 days. It counts the files themselves — the blobs attached to the run — and not the run's own completion flags, which is a different and less reliable thing to count. | what came back | runs | share of finished runs | |---|---|---| | no file at all — the answer is the room | 1,202 | **84.2%** | | a PDF | 185 | 13.0% | | an Excel workbook | 40 | 2.8% | | a PPTX deck | 38 | 2.7% | | a Word file | 0 | 0% | **This corrects something we used to say.** An earlier reading of our own data counted those completion flags rather than the files, and concluded that PDFs happened three times ever and PPTX never. That was wrong, and it was wrong in the direction that matters: it had us telling people we could not hand them a file that was already sitting in their room. PDFs are in fact the commonest file we produce. They are still the exception, and we do not promise one in advance — but "never" was not a true word and we have stopped using it. ## What comes back in the room In the last thirty days, about six runs in ten produced an interactive report — figures laid out on a canvas with their sources attached, rather than a wall of prose. The rest returned written analysis. Either way the result is readable where you asked for it, and the figures carry where they came from. ## Which products produce which file A file is tied to a product, not to effort or price. As measured on 2026-09-17: - **A workbook comes from Value Opinion.** It produced 8 in the last thirty days and 6 of those went to clients. - **A slide deck comes from BOV at Scale**, and there it is about one run in five. - **A PDF can come from most things**, and does on about one finished run in eight. - **A shareable link does not exist.** Report sharing is switched off across the product. Six links were ever created and two of those are revoked. If you were sent a Titleman link before, it is dead. We do not promise any of these up front. The honest order is: state the job, get the answer, and take the file if the run produced one. ## Scout, since it gets asked by name Scout returns **a longlist of properties across a territory** — not a single address, and not a live listings feed. It ran 243 times since 1 January 2026 and 103 times in the last thirty days, so it is among the most used things we have. Its median run took 37 minutes, with the slower tenth running past 81 minutes (measured over the last 30 days to 2026-09-17). **Its honest limit:** Scout does not cover territory outside registered US states. A search outside that coverage returns nothing rather than guessing. ## "Light" and "Heavy" These come up in the same question often enough to be worth a straight answer: they are not names on our product list. The products we publish are Scout, Comps, Value Opinion, Underwriting, Best Use, BOV at Scale, Feasibility, Excels, DD, Legal and Pitch Deck. Whatever tier a job was described to you as, what comes back is the shape above — a written answer in the room, and a file only where the table says so. One naming point we are strict about for a reason: our opinion of value is called a **Value Opinion**, never a valuation and never an appraisal. We hold no appraisal licence, and in the United States valuation is licensed work. The word matters more than it looks. ## "Can I see one?" Not as a sample file. We removed the sample reports from the site deliberately on 2026-09-15 and their links now return 404, so anything you find pointing at them is stale. We also do not point at our own best work as evidence of what your run will produce — those were chosen and published; a run is a run. What we will do instead is run your actual job. That is the only honest demonstration, and it is the same work either way. ## How long it takes Most jobs land within the hour. Medians over the last thirty days to 2026-09-17 ran from about 11 minutes for a default chat to 48 minutes for a Value Opinion. The tails are long — the slowest tenth of Value Opinion runs passed 104 minutes, and a default-chat run has once reached 24 hours. If you need a commitment, ask for the slow end rather than the median; we would rather quote it than miss it. --- *Every figure on this page is a measurement with a date, taken from our own production data on the date shown. Two different measurements appear here — files attached to runs, and run durations — and they are reported separately and never added together. Where an older reading of ours disagreed, the correction is stated above rather than quietly dropped.*