Getting started
The first Excel import
Lenders send their objects as Excel lists, each built differently, with names in abbreviated forms. In this exercise you take such a list into a loan request, with images. Then you clear up the artists’ names that Moliri could not assign unambiguously, and check the result. The exercise takes about half an hour.
It uses the demo holdings: the project Leihanfrage Moderne 2026 and the list Leihliste_Hellweg_Nachtrag.xlsx from the invented Stiftung Kunstbesitz Hellweg, seven paintings with images in the column Bild. It works the same way with a lender’s list of your own.
Write access to the project Leihanfrage Moderne 2026 and the Excel file on your computer: Leihliste_Hellweg_Nachtrag.xlsx (demo data, invented). The first exercise helps, but you do not need it.
1. Look at the list
First open the file in Excel. One header row, below it one row per object: Inventarnr., Bild, Künstler, GND, Werktitel, Datierung, Technik/Material, Maßangabe, Höhe (cm), Breite (cm), Leihgeber, Provenienz (inventory number, image, artist, GND, title, date, technique/material, dimensions, height, width, lender, provenance). No subheadings, no total rows. Moliri can read a list like this directly.
Look at the column Künstler: Kirchner appears there once as Kirchner, Ernst Ludwig and once as Kirchner, E. L.. These are exactly the cases you clear up in step 5.
2. Read in the file
-
Open the project Leihanfrage Moderne 2026.
-
In the sidebar, open the Data In- & Export tab, click Excel under Import and choose the file.
-
Under Range:, check what Moliri has detected: the Sheet Leihgaben Nachtrag, the columns, the first data row and Header row 1. If something is wrong, correct it and click Re-read.
3. Map the columns
Moliri maps the columns to the project’s fields by itself, based on how similar the names are. Each mapping has a confidence rating: sure, check or open.
-
Go through the suggestions: Inventarnr. to Inventarnummer, Werktitel to Objekttitel, Künstler to Künstler/in, Technik/Material to Material/Technik, Bild to Titelbild (cover image). Change by hand whatever is wrong.
-
Leave the column GND without a target. How to use it to assign the artists straight away by their identifier is described under Excel import .
-
Look at the column Künstler. It shows a badge … to resolve: Moliri cannot find these names unambiguously in the library. Such cells are parked instead of written (Not a valid reference, parked for resolving instead of written). This way, no wrong authority record is created.

4. Choose the mode and import
-
Expand Mode and options. As the mode, choose New objects: a new object is created for each row.
-
Switch on Import images so that the images from the column Bild become the cover image, and Pin objects so that the new objects are together at the top afterwards.
-
Read the footer. Before you start, it tells you what will happen: how many columns are mapped, how many new objects will be created and how many cells will be parked for resolving.
-
Click Import … values (the button shows the number).

Mode and options expanded: for the additional delivery, New objects stays selected.
5. Clear up the artists
After the import, Resolve (…) appears at the bottom of the import window. Clicking it opens Resolve references. You can take the same route at any time under Check › Resolve references. Each affected object also shows at the top which values are still waiting; Resolve leads to the same dialog.

-
The list shows each parked term once, sorted by frequency, not every cell. If a name appears in twenty rows, you decide here once for all twenty.
-
Choose Kirchner, Ernst Ludwig. On the right, matches from the library and from sources such as GND and Wikidata appear. Choose the match for Ernst Ludwig Kirchner and click Accept.
-
Do the same for Kirchner, E. L. and the remaining names. A value that cannot be assigned at all can be taken out with Remove without resolving (key
I). -
Click Write (…) to write the decisions into the objects.
Moliri suggests, you decide. Moliri remembers that decision: if Kirchner, E. L. appears again in the next list for this project, the import assigns the name without asking. The same goes for a cell with an identifier, such as a GND number, that exactly one authority record in the library carries. A different spelling is parked again and ends up here again.
6. Check
-
The seven new objects are pinned at the top. Above the gallery, click Limit selection to … pinned objects: now you see only them.
-
Switch to Table and, in the view settings (cog), set exactly the imported fields as text fields: Inventarnummer, Objekttitel, Künstler/in, Material/Technik, Höhe, Breite.
-
In the Filter tab, group by Künstler/in. Both Kirchner paintings are now in one group, even though the list had two spellings. For text fields, the same grouping shows variants: a spelling with a trailing space forms a group of its own and stands out at once.
-
Sort by Höhe (height). A value of the wrong order of magnitude, such as millimetres in a centimetre field, will then be right at the top or right at the bottom.
Errors that affect many objects in the same way are corrected with bulk edit : Trim whitespace for spaces, Divide by 10 for millimetres, Search & Replace for a wrong spelling.
7. If everything went wrong
An import can be undone as a whole. In the sidebar, open the Editing History tab, view Batch operations, find the import and click Undo Batch. The dialog marks conflicts, that is fields that have been edited again since, and lets you select them individually.
What is clear now
| Observation | Where to go next |
|---|---|
| Columns are mapped by name, with a confidence rating | Excel import |
| Unclear names are parked, not guessed | Resolve references |
| Decisions are made per term, not per cell | Resolve references |
| Grouping finds data errors faster than reading | Sort and group |
| An import can be undone as a whole | Trace changes |
Next
- Excel import : all modes and options
- Assign files in bulk : assign more images by file name
- The first Word template : turn the loans into an object sheet