Move from Affinity to Attio without losing your network

This matters more for a fund than for a standard sales team. Affinity organises most work around lists: deal flow is a list, the portfolio is a list, LPs are a list, and each list carries its own fields. It is flexible, but the same company can sit in three lists with three different sets of data attached and none of it modelled as a real relationship.

Attio works the other way. Portfolio companies, LPs, funds and co-investors become first-class objects with their own fields, connected by real relationships. For a fund that has run on Affinity for a few years, this exercise usually cuts the field count hard and turns three overlapping lists into one clean object model.

• Which objects you actually need: People and Companies certainly, Deals for the active pipeline, then the custom objects that carry your firm's logic — Portfolio Companies, LPs, Funds, Co-investors.
• Which fields matter: for each Affinity list, which fields feed a view, a report or weekly work, and which are dead columns nobody fills.
• Which relationships matter: a founder connects to a company, a company to a round, a round to a fund. Map these before you touch the export.

Affinity's moat is the relationship graph. It reads the firm's entire email and calendar history and scores who has the strongest connection to any person or company. That score is the product, and it is also the thing that does not export.

You can export the contacts and the raw interaction records. You cannot export the computed relationship strength, the intro-path ranking, or the years of automatic capture that produced them. No migration moves that intact, to any tool.

What you can rebuild in Attio is the practical version: connect the same email and calendar accounts so interaction history syncs and stays current, then add attributes and automations tracking last contact, frequency and owner, so the team can still see who has been talking to whom. It is not Affinity's algorithmic graph and we do not pretend it is. For funds leaving on cost and breadth, the rebuilt version is enough. If systematically working a warm network is the core of how you win deals, weigh this before you commit.

Two paths. CSV export per list and per view: open a list, filter, export — do this for Organizations, Persons and Opportunities, then for each deal-flow, portfolio and LP list separately so the list-specific fields come with them. Or the Affinity API, which is cleaner for larger funds or when you want opportunities, list entries and field values pulled in bulk with IDs intact.

• Keep the Affinity IDs in every export. They are how you relink records on the Attio side if anything needs a second pass.

Organizations become Companies, Persons become People, Opportunities become Deals. Recreate the Affinity deal-flow stages as statuses on Deals, same order and same names, so reporting continuity holds. Then build the custom objects you decided on and the relationships between them.

• Deduplicate: Affinity often holds the same organisation across several lists. Collapse to one record before import.
• Standardise enumerated fields — one spelling per stage, sector, round and owner — and normalise dates and amounts, especially round sizes and valuations.
• Drop anything that did not survive the data-model exercise.
• Import Companies first, then People linked to Companies, then Deals linked to both. Portfolio Companies, LPs and Co-investors come after the companies they reference already exist.
• Map the Affinity ID to an external identifier during import, and spot-check 10 to 20 records per file.

Affinity pulls enrichment from premium data partners. Attio replaces part of that with AI attributes that fill themselves: summarise a company, classify it into a tier, or run a research agent that pulls funding stage and headcount from outside sources. Connect call intelligence so meetings transcribe and write back onto the record.

This is the step where the move starts paying for itself, because an analyst stops doing the manual research and data entry the CRM can now do on its own.

Affinity reports do not migrate. Recreate each one a partner actually opens as an Attio view or dashboard widget; most funds find a large share of the old reports were never opened.

For a typical fund under 30 users with a portfolio and an LP base, budget two to four weeks end to end. The data movement takes a few days; the time goes into the model rebuild, the custom objects for portfolio and LPs, and standing up the rebuilt relationship tracking and AI enrichment. Heavy list customisation and external integrations push it to five or six weeks.

• Treating it as export and import. The data move is the easy part; the model rebuild and the honest handling of the relationship gap are the work.
• Carrying every list across. Three overlapping lists rebuilt as three overlapping lists recreates the mess in a new tool.
• Expecting the relationship graph to transfer. Plan the rebuilt version on purpose instead.
• Losing Affinity IDs in the export.
• Cancelling before the parallel week. Affinity contracts are annual, so time the non-renewal rather than expecting a mid-term exit.