Discovery cost calculator
What will discovery cost you, and how long will it take?
Enter the size of the estate, its languages and the discovery plan you have today. The calculator prices manual code archeology two ways, estimates a PYD ENTERPRISES timeline and shows every formula it used.
INPUTS
RESULT
Enterprise PortfolioYour planned manual discovery
$480,000
6 × 4 mo × $20,000
PYD ENTERPRISES benchmark range
$440,000 to $825,000
1,100,000 weighted LOC × $0.40 to $0.75
PYD ENTERPRISES timeline
11 working days
vs 84 working days planned
Projected discovery savings
$395,000
$480,000 minus Enterprise Portfolio $85,000
Edge-case exposure in your manual plan
PYD ENTERPRISES model score 65/100 · High
With PYD ENTERPRISES, every executable path is accounted for before build: extracted as a rule, classified as plumbing, or flagged for review. What remains is the review list, which your analysts close.
HOW EACH NUMBER IS CALCULATED
Manual archeology cost, two ways
planned = consultants × months × rate = 6 × 4 × $20,000 = $480,000
benchmark = LOC × factor × $0.40 to $0.75 = 1,000,000 × 1.100 × … = $440,000 to $825,000
Your own plan is the baseline for savings. The benchmark range comes from PYD ENTERPRISES' engagement benchmarks; if your plan sits far below it, check that the plan covers every program, not a sample.
PYD ENTERPRISES timeline
parse = ceil(weighted LOC ÷ 1,000,000 LOC/h ÷ 8 h) = 1 d (1.1 h)
review = ceil(weighted LOC ÷ (25,000 × 6 analysts)) = 8 d
total = 2 d setup + 1 + 8 = 11 working days
Parse throughput is PYD ENTERPRISES' figure of about 1M LOC per hour on a 32-vCPU node. Review assumes your consultants become reviewers, capped at 8. Manual timeline uses 21 working days per month.
Savings
savings = planned - license = $480,000 - $85,000 = $395,000
License tier picked by LOC: Enterprise Portfolio, per year, up to 3,000,000 LOC across 5 systems. Analyst review time is a cost in both scenarios and is not netted out.
Edge-case exposure score
load = weighted LOC ÷ (consultants × months) = 45,833 per consultant-month
score = 100 × load ÷ (load + 25,000) = 65
A PYD ENTERPRISES model score, not a measured probability. It rises as each consultant must cover more weighted code per month, which is when manual reading turns into sampling.
Email this estimate
We send the inputs, every formula and a discovery risk report you can forward to your program sponsor. No source code is requested.
Language factors
Not every line costs the same to read.
The benchmark range of $0.40 to $0.75 per LOC is calibrated on COBOL batch. Other stacks are weighted by these PYD ENTERPRISES planning factors.
- COBOL batch + JCL×1.00Baseline: copybooks, REDEFINES, PERFORM THRU, JCL COND chains.
- COBOL + CICS online×1.20Pseudo-conversational state in COMMAREAs and BMS maps adds hidden flow.
- Java EE / Struts×0.85Typed and readable, but logic is split across EJBs, ActionForms and XML descriptors.
- VB6 / WebForms×0.90Event handlers, global module state and ViewState carry rules.
- PL/SQL / T-SQL×1.10Triggers and dynamic SQL produce side effects far from the caller.
- Informix 4GL×1.05Screen logic and database access interleaved in one program.
FAQ
About the numbers.
Where does the $0.40 to $0.75 per LOC range come from?
It is PYD ENTERPRISES' own benchmark from discovery engagements: the fully loaded consulting cost of reading code and producing specifications, divided by the lines in scope. It is not an industry survey figure, which is why the calculator also prices your own plan from consultants, months and rate.
Why are the language factors different?
They reflect how much hidden control flow and side-effect logic each stack tends to carry. CICS pseudo-conversational programs hide state in COMMAREAs; PL/SQL triggers fire far from their callers; Java EE is more readable but spreads logic across descriptors. The factors are PYD ENTERPRISES planning figures and are shown beside each language.
Is the exposure score a probability of failure?
No. It is a PYD ENTERPRISES model score from 0 to 100 that rises as each consultant must cover more weighted code per month. It is a way to compare plans, not a prediction. The formula is printed under the result.
Does the PYD ENTERPRISES timeline include our review time?
Yes. Most of the timeline is analyst review of extracted rules, not compute. The calculator assumes your discovery consultants become reviewers, up to eight per system.
What should I do with the estimate?
Email it to yourself or your sponsor, then start a 14-day trial on up to 100,000 lines in your own VPC to check the timeline against a slice of your real code.
Test the timeline on your own code.
Start a 14-day trial on up to 100,000 lines in your private VPC and compare the review effort with the estimate above.