Ames Systems

IMPORT QA · NOTHING-LOST VERIFICATION · FIXED PRICE

The import finished.
Now prove nothing disappeared.

Friday's migration "went fine" — the wizard said so. But nobody can say whether all 8,000 contacts actually arrived, or whether some arrived twice, or whether phone fields got quietly mangled on the way in. You send us the export from before the import and the export from after. We show you exactly what was lost, duplicated, or changed — with counts that reconcile to the penny.

Sound familiar?

The three questions nobody
wants asked after an import.

Is it all there?

"The count doesn't match."

You exported 8,214 records. The CRM says 8,207. Or 8,220. Everyone shrugs — maybe it deduplicated something? Maybe some bounced? Nobody knows, and the answer is buried somewhere in importer logs nobody kept.

  • We list every record that went in but never came out — by name, not by guess
  • And every record that appeared without your permission
Did it change anything?

"Why does this phone number look… shorter?"

Importers trim, reformat, drop trailing digits, mangle dates. One rep notices a weird record; how many others look like it? Checking by hand means opening records side by side for a week.

  • Cell-by-cell comparison of matched records — every difference listed before → after
  • You see which field changed, in which record, from what to what
Can you defend it?

"Trust me" is not a migration report.

Your boss, your client, or legal will eventually ask: how do you know nothing was lost? A screenshot of the importer's success screen is not an answer.

  • A scripted verdict: counts, findings, and totals that reconcile exactly
  • Written so a non-technical stakeholder can audit it alone — and a second engine recomputes the whole comparison from your raw exports
What we do

Two exports in.
One defensible verdict out.

We never touch your CRM. The whole check runs on the file you exported before the import and the file you export after — sandbox-safe by construction.

What we compare

MISSING
records present before, absent after — data lost in transit
EXTRA
records that exist after but never existed before
CHANGED
same record, different values — field by field, before → after
DOUBLES
records the import created more than once
TOTALS
numeric columns summed per side; every unit of difference must be explained by a finding above, or the verdict fails

What comes back

Verdict
PASS or FAIL in plain words, with the exact accounting behind it
Finding lists
one CSV per category — every row names names
Reconciliation
the sums balance: pre-import total − losses + additions = post-import total
Audit path
a README a colleague can follow without talking to us

Honest boundaries

Exports only
we compare the two files; we don't audit the importer itself or fix your CRM
Declared keys
you tell us which column defines "same record" and which columns correspond across systems — we never guess a match. Rows without a key — or sharing one ambiguously — force an INCONCLUSIVE verdict: we refuse to certify what cannot be proven
Similar ≠ same
near-miss records are flagged as ambiguous for YOUR decision, never auto-joined
Data handling
NDA on your paper; both exports deleted on request — verified absence, register-tracked, confirmed in writing
What a FAIL looks like

This is the report you want before anyone else finds it.

IMPORT QA VERDICT: FAIL
pre-import rows : 8,214
post-import rows: 8,213

records lost in transit      (MISSING):   2
unexpected additions         (EXTRA)  :   1
silent cell changes          (CHANGED):   3
duplicates created by import (DUPPOST):   2

totals[lifetime_value]: pre=$412,330  post=$411,180
expected-from-findings=$411,180 → RECONCILES

Every number traces to a CSV listing the exact records. A PASS reads the same way — zeros everywhere, totals reconciled — which is precisely what makes it worth having.

Fixed price

One comparison,
one price.

Free sample

$0

One pre/post pair up to 1,000 rows, inside 24 hours. See the verdict format on your own data.

Standard comparison

$300–500

Up to 10,000 records per side. Declared key + column map agreed upfront.

After a failed migration

Scoped

Found losses? We re-run the identical check on your corrected re-import so you close the incident with proof.

$0Free sample · one small pre/post pair · inside 24h

Have both exports? Start with the free check.

Anonymize them if you like — the method shows either way. Standard comparisons run $300–500 per pair ≤10k rows, deliver-then-invoice.

Reply within 24h · prefer email? [email protected]

Related

Where this fits in the journey.

Before the import

Moving systems?

Clean and map first — migration prep beats discovering losses afterwards.

Before the migration

Inherited mess?

Cleanup with a full change log, so only clean data ever enters the new system.

Two systems

CRM vs billing disagree?

Scripted cross-source reconciliation — exact rules only, ambiguity flagged.