Self-assessment

Score yourself

Answer each question honestly. Yes = 2, Partly = 1, No = 0. Keep a running total.

Your score

0/ 24

0 of 12 answered

Not scored yet

Finding it

0 / 6

01 Could you list, today, every place personal data lives: systems, spreadsheets, exports, email, backups?

02 In the last quarter, has anyone actively hunted for new or forgotten stores: shadow IT, old exports, test copies?

03 Do you know which vendors and partners hold your customers' or employees' data?

Tagging it

0 / 6

04 Is there a written scheme that separates low-risk data from high-risk data?

05 Is every important dataset tagged, with a named owner for each tag?

06 Do your strongest safeguards, encryption, access control, logging, actually follow the highest tags?

Mapping it

0 / 6

07 For each main process, can you trace personal data from the moment of collection to deletion?

08 Does each flow record why you hold the data and on what legal basis?

09 Do the maps show where in the world the data sits, and every party that receives it?

Keeping it true

0 / 6

10 Is there one register that pulls this together, with named owners and review dates?

11 Does every new project, vendor or system update the picture when it starts, not at audit time?

12 Could you produce everything you hold on one customer, across every system and partner, within days?

Readiness bands for this self-assessment
Score Where you stand What to do next
0 to 8ExposedDiscovery is not done, so every downstream duty rests on guesses. Start with Move 1 this quarter. Interviews before tooling.
9 to 16HalfwayThe picture exists but has holes. The section with your lowest marks is the next project. Re-score after closing it.
17 to 21Strong baseShift from building to proving. Run the live test at the end of this page, collect the evidence, tighten the review cycle.
22 to 24LeadingYour risk now is drift: new systems and vendors quietly outrunning the register. Guard the intake gate.

Any single "No" is a named gap, whatever the total says. Nothing leaves your browser.

What these three things are

Data discovery finds personal data across the enterprise: core systems, apps, documents, email, data warehouses, vendor tools. Not the data the org chart admits to. The data that actually exists, including forgotten exports, test copies and shared drives nobody owns.

Data classification tags what you find by sensitivity, so protection scales with risk. A press release and a customer account list do not deserve the same controls.

Data mapping records how personal data moves: from which collection point, into which system, on to which partner, into which report. Discovery tells you what you have. The map tells you how it moves.

One clarification: the DPDP Act does not grade personal data into sensitive and ordinary classes; it treats all personal data alike. Classification is still worth doing, because safeguards, breach priorities and duties under sector rules do scale with sensitivity.

The framework: four moves

Move 1: Find. Run discovery as a project with an owner and a scope. List sources from the IT asset register, the software subscription list from finance, and the single sign-on app list; shadow IT shows up in expense reports. Then walk business units through their real work, because only people can tell you about the spreadsheet exported every Tuesday. Scan with tooling, confirm samples with evidence, record every finding. Repeat on a cycle: new systems, vendors and exports appear every quarter.

Move 2: Tag. Four levels cover most enterprises:

LevelWhat it isExamples
PublicMeant for anyoneWebsite content, marketing material, published reports
InternalHarmless if seen insideInternal policies, project plans, announcements
ConfidentialNeed-to-know; damage if disclosedCustomer lists, financial records, contracts, most personal data
RestrictedSevere damage if compromisedPasswords and keys, payment card data, health data, large sets of personal data

For each level, write down what belongs to it, the damage if it leaks, the minimum safeguards, and who approves access. Business owners assign the level; the security team sets the handling rules; tooling applies labels where it can. Keep the scheme stable. Re-labelling the world every year teaches people to ignore labels.

Move 3: Map. Build flow by flow, not system by system. For each process, interview the owner with a short questionnaire, trace the data end to end including integrations and exports, and validate with IT against real systems. The twelve fields to capture per flow are in the mapping sheet below.

Move 4: Maintain. Land everything in one register with named owners and a review cadence. Make updates a byproduct of change: every new project, vendor or report updates the register at intake. A map refreshed once a year is archaeology.

Template 1: the data inventory

One row per dataset. Copy this and fill it in.

DatasetOwnerWhere it livesWhose dataWhat it containsLevelSourcePurpose and basisKept untilWho can accessWho receives itReviewed on
Example: loan applicationsHead of lendingCore lending app, Mumbai region; nightly backupCustomersName, ID proof, income, bank detailsRestrictedCollected directlyProcess the application; consent3 years after closureLending ops, credit teamCredit bureau, collections partner

A full inventory covers every dataset the discovery found, not only the core systems. Spreadsheets, exports and email folders get rows too.

Template 2: the data mapping sheet

One sheet per process flow: customer onboarding, payroll, marketing campaigns, loan servicing. For each flow, record:

  1. The process and its owner.
  2. Whose data: customers, employees, candidates, vendors' staff, website visitors.
  3. What data: the categories, and whether any of it is sensitive in your context (financial, health, children).
  4. The source: collected directly, received from a partner, or generated by you (logs, scores, inferred profiles).
  5. The purpose and the legal basis: consent, or one of the permitted non-consent grounds. This is where the map feeds the notice and consent work.
  6. Where it lives: applications, databases, file shares, email, paper. Include backups and archives.
  7. Where in the world: hosting location and any remote access from outside India. See the data sovereignty page.
  8. Who can reach it: internal roles and privileged accounts.
  9. Who receives it: processors, partners, group entities, authorities.
  10. How long it stays: the retention period, its trigger, and the deletion method. See the retention and erasure page.
  11. The safeguards in place: encryption, access control, logging.
  12. Risk flags: large scale, profiling, new technology, children's data. These trigger an impact assessment. See the record of processing and DPIA page.

Both templates are available as downloadable files.

What good looks like

A current inventory with tags, flow maps for every significant process, one populated register, and a working update routine.

The live test: pick one real customer and one date. Produce everything you hold on them, where each piece came from, who it was shared with, and what they were told. Time it. If it takes weeks, the groundwork is not done.

Where enterprises go wrong

  • A one-time inventory that is stale within a quarter.
  • Mapping only the core system while marketing, HR and analytics copies multiply elsewhere.
  • Scanning tools trusted as the whole answer. They miss spreadsheets, exports and paper.
  • A classification scheme with ten levels nobody remembers, or one bucket for everything.
  • Treating the exercise as an IT project. The business owns the data, IT owns the systems, and the map needs both.

The decisions to make

  • Interview-led discovery, tool-led discovery, or both. For any real estate, both.
  • Whether to cover paper-digitised records and legacy stores.
  • Depth of mapping: inventory-level, or field-level lineage for the critical flows.
  • The classification scheme, and who assigns the levels.
  • How the register stays current: an owner per activity, a review cadence, project intake gates.

Next step

The record of processing and impact assessments, the register this groundwork feeds. If you need help with any of this, write to us and we will help.