Every Data Migration Ever
A brief summary of the conversations over the last month:
Me: “So, how much of your data do you need to migrate?”
Client’s Head of IT: “Should just be some person records, some company records. That about right [Operations Manager]?”
Client’s Operation Manager: “Yeah, not even. Just a subset of that.”
Me: “So it’s just flat data? Like one row for one person, no linked tables?”
Client’s Head of IT: “Correct. And we don’t even need much there, just the basic name, address, phone number, etc will do.”
Me: “How clean is the data? Are you sending all of it and expecting us to clean it, or are you sending just the stuff you want to keep?”
Client’s Head of IT: “Oh, we definitely don’t want that in the new system, so we will just send over the parts we want.”
Me: “Are you sure? Are you absolutely doubly sure? Pinky promise, no take-backsies?”
Client’s Head of IT: “Yeah, but tell you what, let’s have a call next week with our data guy.”
Today:
Data Guy: “Yeah, so we have two unique databases we need to merge, one in India and one in England. Hundreds of thousands of people and client records, millions of contact log records. For each worker, there will be around a hundred unique fields that need to be mapped, and for each worker, around a thousand records for previous work history and communication logs, an unknown amount of documents, but let’s say at least 20 PDFs per person. There are around two hundred directly relevant tables, but a lot more that could be useful.”
Me: “Do you want some of this or all of it?”
Data Guy: “…yes? Obviously everything. We need this import so that you can perform a data cleanse, fix duplicates, fill in missing info, sort it properly, etc., as we don’t have the capacity to do it ourselves.”
I should know better at this point, I fall for it every time.

Clients From Hell