ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
WHAT IT COSTS TO BUILD AND RUN THIS DATASET
2
Scenario B · one vertical, national — A single vertical covered nationally. The default sizing. Hard costs only: headcount, data contracts, credits, infrastructure.
3
4
CostNote
5
BUILD — one-time$853,912Everything before a single usable record exists
6
Run — year 1$1,344,106
7
Run — year 2$1,392,338Universe growth, escalators, salary inflation
8
Run — year 3$1,442,455
9
RUN — three years$4,178,898This is the number that gets left off build proposals
10
3-YEAR TOTAL COST OF OWNERSHIP$5,032,810
Build once, run for three years. Put a vendor quote on this same clock, never on the build alone
11
12
WHERE THE MONEY ACTUALLY GOES (three-year totals)
13
Category3-year costShare of totalWhat sits in here
14
People$3,525,26270.0%||||||||||||||||||||||||||||Build task table, the operating team, on-call, and the model-migration tax
15
Data contracts$803,12516.0%||||||Vendor licences plus their one-time onboarding fees
16
Credits & processing$512,16610.2%||||
Scraping credits, LLM judgments, verification APIs, continuous QA — including the first full pass
17
Infrastructure$184,7583.7%|Storage (rent), compute (taxi meter), movement (tolls)
18
Golden dataset & other setup$7,5000.1%The hand-labelled truth set, plus any displaced roadmap work you chose to price
19
TOTAL$5,032,810100.0%Should read 100% — if it does not, a line has been added without being categorised
20
Reconciliation check (should be zero)$0Categories vs. the headline total
21
22
UNIT ECONOMICS — what you get for it
23
ValueNote
24
Records delivered (contacts in scope)500,000Locations × contacts per location
25
Accurate, in-recall records per year318,750Contacts in scope × accuracy × recall — the ones a rep can actually work
26
COST PER ACCURATE RECORD PER YEAR$5.26
Total ÷ three years of usable records. The only number that compares two options honestly
27
→ excluding people$1.58
Drops the build team, the operating team, on-call and the migration tax. Keeps data contracts
28
→ credits + infrastructure only$0.73
Pure machine cost. This is roughly the number a build proposal computes — and why it is several times too low
29
Cost per record touched, per refresh cycle$0.0090Blended scraping + LLM cost per signal-refresh
30
Year-1 run cost per record$5.38What one location costs you to keep current for a year
31
32
GUT CHECKS
33
ValueTargetNote
34
Run over 3 yrs ÷ build cost4.9at least ~2×Below that and the long tail is underestimated. Collection is the cheap part
35
Build share of lifetime cost17.0%~30% or lessThe other 70% is refresh, breakage, source churn and QA
36
Infrastructure per month$5,000≥ ~$5,000The floor for one vertical at 95% precision. Below it, something is unscoped
37
Implied average team size during the build4.32-8 FTETotal FTE-months ÷ calendar months. Outside that range, revisit the task table
38
People as a share of total cost70.0%expect 60-80%If credits and infrastructure dominate instead, the pipeline design needs another look
39
40
HOW TO USE THIS
41
ScopeThis calculator prices the hard cost of building and running the dataset. It does not price what bad data costs your sales team, or what the wait costs you — those are separate arguments.
42
43
1.Open Inputs. Every shaded cell is editable; nothing else in this workbook is.
44
2.Fill the BUILD task table in FTE-months. The task list is fixed — the headcount is your call.
45
3.Calibrate the credit tables from a real pilot: 250-500 accounts for scrapers, 500-1,000 for a full run. Cost every signal, not just the website.
46
4.Get an engineer to price storage, compute and movement — they grow on three different things.
47
5.Use the Scenarios tab to see how all of it moves with TAM size.
48
49
StorageRent. Grows by keeping more records — not by usage.
50
ComputeA taxi meter. Grows by adding always-on processes — not by more queries or more data.
51
MovementTolls. Grows with refresh frequency: weekly instead of monthly is 4× on the same data.
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100