Market research firm data inventory: records by system

What a market research data inventory records#

A market research data inventory is a metadata list: each row names a system, a type of record it holds and the facts that decide how that record can be used. It copies no data, so it can circulate among the COO, compliance and IT without exposing respondent or client material.

Research firms need more columns than most businesses because their records carry promises in two directions: to respondents, through consent and privacy notices, and to clients, through master agreements and statements of work. The template keeps both visible on every row, so nobody has to remember which studies came with which strings.

The template: systems, record types and key columns#

Start with these rows and adjust the systems to your own stack. The values shown illustrate how to fill each column; they are not recommendations for your firm.

How to fill in consent basis and client ownership#

Use a short list of fixed values for the two judgment columns so rows can be sorted and compared. Free text drifts quickly; a fixed list forces each question to be answered or marked unknown.

Unknown is an acceptable first answer. A row marked unknown shows exactly where the next contract or notice review is needed, which is more useful than a confident guess that later turns out wrong.

• 1Consent basis: study-only consent; consent includes further research use; membership terms allow reuse; not respondent data; unknown.
• 2Client ownership: firm-owned; client-owned deliverable; client confidential but usable in aggregate; mixed by project; unknown.
• 3Retention: name the trigger event and the source of the period, such as report delivery plus the contract period.
• 4Notes: record exceptions, such as a client that prohibits third-party AI tools on its studies.

Supporting columns worth adding#

The five core columns decide what a record can be used for. A few supporting columns make the inventory useful for planning a migration, answering a client audit or scoping any later licensing review, and each one can be filled from a conversation with the system owner.

Systems research firms often forget#

The systems most often missed are the ones where research files land by accident. Collaboration tools are the biggest gap. Microsoft states that Teams chat messages are stored in a hidden folder in the mailbox of each user in the chat, and channel messages in a similar hidden folder in a group mailbox, reached through eDiscovery rather than directly. Files shared in Teams usually end up in personal OneDrive or team SharePoint storage, so a data file dropped into a chat may not sit in the project folder at all.

• 1Personal OneDrive and desktop folders where analysts keep working copies.
• 2Old survey platform accounts left open after a move to a new platform.
• 3Transcription and translation vendor portals that still hold files.
• 4Facility streaming and video portals from in-person groups.
• 5Incentive and payment platforms holding respondent contact and payment details.
• 6Freelancers' and subcontractors' own storage.
• 7Email attachments carrying data files and banner tables.

How to build the inventory without moving files#

Build the inventory from interviews and admin screens, not exports. Each step uses information that system owners can give from what they already see, so no respondent or client file has to leave its system.

• 1List systems from the IT asset register, single sign-on and expense reports.
• 2Interview each system owner about record types, the date range and rough volume.
• 3Pull consent versions from survey and screener templates, with the dates each version was in use.
• 4Pull ownership answers from master agreements and a sample of statements of work.
• 5Copy retention rules from the retention schedule, and note any tool that deletes automatically.
• 6Review the draft with operations, compliance and one senior researcher before calling it final.

Illustrative: a B2B research firm maps its records#

Illustrative: a fictional research firm runs B2B studies for software and industrial clients, combining online surveys, a recruited database of professionals and remote in-depth interviews. The COO's first draft lists the obvious systems: the survey platform, the recruitment database and the shared drive.

Interviews with system owners add an old survey platform account still holding several years of studies, a transcription portal nobody had closed and a Teams channel where data files were exchanged with a fieldwork partner. Each becomes its own row.

The consent column shows that recent interviews used a form with a separate choice about reuse in methods research, while older ones did not. The ownership column shows that most master agreements allow aggregated use of results. The firm takes the exports it is allowed to keep, closes the stale accounts and schedules a contract review for rows marked unknown.

How SourceX uses an inventory like this#

SourceX starts every engagement with metadata, which is exactly what this inventory holds. In the Supply step of the SourceX five-step transaction, the inventory shows which record families exist, how far back they go and which rows are blocked by consent or client ownership.

The SourceX Enterprise Data Value Framework then rates the remaining rows qualitatively, on drivers such as uniqueness, domain expertise, human-generated signal, recency, rights and privacy burden. Nothing is shared during the initial assessment, and the firm approves every later step.