[Data Import] Walkthrough importing (relational) data

Showcase Tips and Hacks
1

Hi everybody,
I’m in the process of a huge import because I’m migrating basically all my company operations to Coda.
This post is intended to share some best practices but also to ask for few recommendations.

I tried to provide a self explanatory document here (just focus on the Importing folder):

I won’t go into the details of all the sections, but if you have questions/doubts/suggestion, you’re welcome to let me know.

I wish to have some hints from experts here regarding importing time: I played with a full import a couple of month ago to understand the overall timing and - by that time - it took several hours to import the full set of tasks dataset (around 2K rows with ~ 40 columns, several lookups… but not a big deal).

I know Coda improved a lot since then (and I got much more familiar with it) therefore I’d expect an overall improvement, but I appreciate if you have some useful experience to share.

Anyway, I hope you find it useful: enjoy!

4 Likes
2

Nice process, although a bit overcomplicated IMO :slight_smile:

Could’ve just juggled some column types and copy-delete-pasted some data so that Coda would parse it:

ezgif.com-optimize (60)
ezgif.com-optimize (60)904×480 969 KB

Same with lookups btw, changing the type to the proper table lookup for the Area column just works.

3

Well, @Paul_Danyliuk… yes: it is overcomplicated if you rely on this data samples… :grimacing:
As specified, it’s an over-simplification of the actual scenario.

For Tasks only, I have several date-time fields (created, started, completed, due date, etc…).
Similar for Goals. The same for Invoices, Expenses, Marketing/Sales activities, …
Doing cut/paste for 2K+ rows for each date column doesn’t seem to be so efficient and - moreover - replicable;

Further, there are lookup tables, (Goals, Contacts, Accounts, Areas, …).

Then, reference tables, i.e. relational consistency with the previous context; I hardly see a straightforward way, here.

Finally, transitioning from one system to another implies to have everything ready: people have to switch from one to other ideally seamlessly.
Creation of an automated process to import data allows to minimise potential pitfalls by testing all the importing steps.

Perhaps my use-case is uncommon; I’d like to hear if anyone else is thinking/experiencing something similar.

Cheers

4

We have migrated our data from couple different tools to coda. It took a while and the process wasn’t perfect, but it is definitely worth it. I have made several mistakes, that made the transition more complicated and these are some tips from my personal experience:

  • analyze your data and decide which information you really need;
  • Make a scheme of the relations between all your tables. Make sure you don’t enter the same data more than once. This is very important.
  • I would take extra time to set up all the date and time fields, because there we had some problems. Different tools could use different formats.
  • differentiate the “raw data” from the the data, which is created with formula and lookups. You don’t want to import that data. You will need to make the formulas manually. The good news it needs to be done just once for every row.
  • don’t rush the process. If you have a lot of data and you rely on it is better to delay a bit rather than creating sloppy experience and have bugs all around.

I think now is a good time to reinvent your system. I don’t mean start from scratch, but reconciler the processes and have in mind the super powers of Coda.

I hope this was helpful.

1 Like