Thirty Words of Scope: The Part of a Quote That Never Grows With the Price

One hundred listings, thirteen fields, no exceptions

Take the 100 service records published in the VIT MARKET catalogue, sort each record's field names alphabetically, join them into a single string and count how many distinct strings come back. The answer is one. Every listing carries the same thirteen fields — number, title, difficulty, category, description, priceRub, priceUsd, duration, slug, group, priceMinUsd, priceMaxUsd and featured — and not one of them has a fourteenth or is missing a thirteenth. For a buyer that uniformity cuts in two directions at once: whatever the catalogue can tell you about a $50 job is exactly what it can tell you about a $6,000 one.

Four of the thirteen are plumbing rather than information. The number is a position in a list, the slug is a URL, the group is a shelf, and featured is the seller's opinion about what belongs in the shop window. That leaves nine fields answering four questions. What is it — title, category, description. How demanding is it — difficulty. How much does it cost — four separate price fields. How long does it take — duration. Four answered questions is not nothing. It is also not a scope of work, and the distance between those two things is where a buyer's money quietly goes.

Two thousand nine hundred and twenty-two words for a hundred jobs

The description field is the only place in a listing where the work is described in sentences rather than labels. Counted as whitespace-separated tokens of the stored text — the descriptions are held in Russian in the source file — the hundred descriptions come to 2,922 words in total. The shortest is 18 words, the first quartile is 26, the median is 28.5, the mean is 29.22, the third quartile is 33 and the longest is 42. Half the catalogue explains itself in between 26 and 33 words. Rendering the file in English or Spanish would lift those counts somewhat, because Russian packs more into a single token, but it lifts every row together.

The two extremes are not where instinct puts them. The shortest description in the catalogue, 18 words, belongs to marketplace card optimisation for Wildberries and Ozon, published at $30 to $80 per card. The longest, 42 words, belongs to a weekly report for a company owner at $200 to $600 — one of the cheaper listings on the automation shelf. Add the titles, which run to 649 words across the hundred with a median of six words each, and the entire body of prose this catalogue publishes about a hundred distinct pieces of professional work comes to 3,571 words. That is roughly the length of two of these articles.

Paying more does not buy more explanation

If description length tracked anything at all, price would be the obvious candidate: a bigger job needs more explaining. It does not. Ranking all 100 listings by description length and separately by price midpoint — midpoint being priceMinUsd plus priceMaxUsd divided by two — and correlating the two sets of ranks returns a Spearman coefficient of −0.0088. Run on the raw values instead of the ranks, Pearson returns −0.0779. Both are indistinguishable from no relationship whatsoever, and both lean very slightly negative, which is to say that the dearer listings are, if anything, a shade terser than the cheap ones.

Group averages state it more plainly than coefficients do. The ten most expensive listings by midpoint average 28.3 words of description; the ten cheapest average 28.9. Cut the same hundred rows a different way and the result survives: the eleven listings whose floor price is $1,000 or more average 29.8 words, while the twelve whose ceiling price is $300 or less average 28.8. A commission that starts at four figures is explained in one word more than one that finishes below three. Whatever the rising price buys, it is not a fuller written account of the work.

What one word of scope costs

Divide a listing's midpoint price by the number of words it spends describing itself and the result means nothing in isolation and a good deal in a distribution. Across the catalogue the median is $26.79 per published word of scope. The first quartile is $12.11 and the third quartile is exactly $50.00. So a quarter of the shelf is asking more than fifty dollars for every word of explanation it prints, the middle of the shelf is asking a little under twenty-seven, and the cheapest quarter is still asking more than twelve.

At the floor sits YouTube SEO at $1.16 a word — a listing priced at $15 to $50 per video that spends 28 words on itself. At the ceiling, among rows whose band is an ordinary project price, sits a custom Shopify theme at $166.67: two to six thousand dollars of work and 24 words of description. One row scores higher still, a marketplace template at $208.73, but its published band runs from the price of a single copy to a lifetime total across all sales, so its midpoint is not a project price and the number is quoted here only in order to be set aside.

Twenty-one listings put a number in the prose

Twenty-one of the hundred descriptions contain a digit of any kind. That is the entire population of numeric statements in the catalogue's prose, and reading them one at a time shows that most are illustrations rather than commitments: a sample price alert in a chat window, a report that arrives at nine on Monday morning, an inbox lightened by 50 to 70 per cent, a builder subscription whose saving repays the work in 6 to 12 months. All of it helps a reader picture the output. None of it helps a buyer decide whether the output arrived.

VJOURNAL's reading — and this is a reading, not something the data counts on your behalf — is that four of those twenty-one state something you could genuinely test on delivery day. A landing page promising 90+ in PageSpeed. A speed job pinned to INP under 200 ms, LCP under 2.5 seconds and CLS under 0.1. A first-line answering bot that handles 60 to 80 per cent of typical questions in the company's own words. A backup routine that keeps 7 daily, 4 weekly and 12 monthly copies and mails a confirmation every morning. Four listings in a hundred publish an acceptance criterion. The other ninety-six leave acceptance entirely to the brief.

Seven estimates that keep running after handover

The delivery estimate is the second place a listing leaks scope information, and seven of the hundred estimates describe work that does not stop at handover. Answer-engine optimisation quotes 3 to 5 days to start and then 3 to 5 hours a week to run. Rank monitoring quotes one day to connect a client and then 0 to 30 minutes a week thereafter. A site migration quotes 2 to 5 days of preparation plus a month of support. A traffic-drop investigation quotes 1 to 3 days for the diagnosis and states outright that the remedial work is a separate job.

Every one of those four is priced with a single pair of numbers. Only one says in dollars which part of the work the pair covers: the investigation is labelled per diagnosis, at $500 to $2,500. The migration's $700 to $3,000 sits directly beside an estimate that contains a month of support, and nothing in the record says whether that month is inside the price or beyond it. This is not a contradiction and it is not a trick — it is a price list behaving like a price list. It is also the precise point at which a brief has to take over from a catalogue.

Nine questions the fields cannot answer

Set what a signed engagement needs against what the record actually stores and the gap turns out to have a fixed shape. Nine questions have no field anywhere in this catalogue, and not one of them is exotic. How many rounds of revision are included. What counts as done. Who supplies the content, the data and the access. When the money falls due, and how much of it before work starts.

Then the second half. Who owns the output, and under what licence. What happens after handover, and for how long. Which recurring third-party costs — hosting, subscriptions, model usage — the buyer carries once the work is delivered. What happens when the scope changes halfway through. And who is actually doing the work, given that a catalogue lists services rather than people. Nine questions, thirteen fields, no overlap at any point.

Two of the nine surface in the catalogue's own prose as absences rather than answers. Search the hundred descriptions for hosting, a subscription, an API or an operation limit and seven come back, four of which raise the subject only to remove it: no hosting, no monthly builder subscription, no perpetual plugin updates. A fifth cites a hosting failure as the disaster it insures against. Only two look forward at a bill the buyer will keep paying — a custom module written against the operation limits of an automation platform, and an audit whose entire purpose is to cut a model API bill by two to five times. The catalogue is fluent about the costs it removes and quiet about the ones it starts.

What these counts cannot tell you

Every figure above is a fact about published text, not about delivered work. A 28-word description does not imply a 28-word conversation, and this dataset holds no record of what a buyer is told once they write in. There is no field for the brief, the call, the proposal or the signed agreement, so nothing here demonstrates that the nine missing questions go unanswered anywhere. It demonstrates only that they are not answered on the shelf, before contact, at the moment a buyer is deciding whether to make contact at all.

Nor can this analysis conclude that the descriptions are too short. A catalogue entry is an index, not an agreement, and 28 words that name a deliverable precisely can be worth more than two hundred that hedge. What the near-zero correlation does establish is narrower and still useful: the quantity of written scope a buyer receives for free is effectively fixed at about thirty words, so the quantity they must go and obtain for themselves grows with every dollar the price climbs.

The brief that makes a price mean something

The practical form of all this is short. Before a published band means anything, put the nine questions in writing and send them alongside your own description of the job — longer than thirty words, because you are the party who knows what done looks like. Where the delivery estimate names more than one phase, ask which phase the price covers. Ask for one testable number, the way four listings in a hundred already volunteer one without being asked. And ask what you will still be paying after the final invoice clears.

A price band is the answer to a question the seller has already chosen. A brief is where the buyer gets to choose the question instead. Nothing in this catalogue is concealed: the fields are uniform, the prices are published in two currencies, the estimates say exactly what they say and stop. It is simply that a listing is thirteen fields wide, and an agreement is not.

Decision framework: what to include in a website brief

A usable website brief defines the user problem, desired outcome, audiences, current evidence, required capabilities, constraints, owner, acceptance measures and known exclusions. Page count alone is not scope.

A short brief can be precise, but thirty generic words cannot distinguish a landing page from an operating service. Missing constraints return later as discovery work, change requests or quality disputes.

Before requesting a quote, add one sentence for each of nine fields: problem, user, outcome, evidence, functions, content, integrations, constraints and acceptance; label unknowns instead of inventing certainty.