⇥Importing data

Where the data belongs: in Contacts and in Opportunities. Above each list, a button opens the import as a drawer — you do not leave the page.

There is no separate import page any more. What there is, is the Import history: afterwards it shows who imported which file, and when.

Enlarge

Enlarge

Excel files (.xlsx, .xlsm) and CSV. Please convert other formats first using “Save as”.

It also works without a file: select the cells in Excel, copy them and paste them into the field. The first row holds the column names, and there must be at least one data row.

Then the delimiter does not match. German versions of Excel usually save CSV with a semicolon, whereas the comma is common internationally.

Two ways out:

• Save as .xlsx in Excel and use that instead of the CSV — then the question does not arise.
• Or select the cells in Excel, copy them and paste them into the paste field.

If you stay with CSV: check in the preview whether the columns appear separately before you map them.

Usually not. Aurora Hive Sales detects the columns itself — from the headings and, where they give nothing away, from the values: a column in which 96 per cent of entries contain an @ followed by a dot is an email address, even if its heading says “Field 7”.

You can override every decision. The mapping shows the field in Aurora Hive Sales on the left and the column of your file on the right. Anything not mapped is left out — nothing is lost, it is simply not imported.

No. Detection is calculated, not asked — no value from your file goes to an AI or anywhere else.

With contact data this is not a detail: sending an address list through an outside service would be a transfer of personal data, for which you would need a legal basis and a contract. That is why detection has deliberately been built so that it manages without one.

At least one of: Email, Contact person, Last name or Company. Without one of these, a record could not be identified.

In practice you should always map the email if it is in the file: it is the feature by which existing contacts are recognised and added to, and the prerequisite for the contact appearing in a campaign later.

Yes, and you should always do so. The Dry run calculates the whole import without writing anything — the same calculation as for the real import.

The result is shown in figures: how many records are new, how many already exist and will be added to, how many will be skipped. If these figures do not match your expectation, the mapping is wrong — and nothing has happened yet.

They are recognised by their email address and added to rather than overwritten. An empty field in your file therefore does not delete the existing value in Aurora Hive Sales.

This is the property that makes regular imports from a maintained list safe: you can read in the same file twice without doing any harm.

Yes — every import can be undone for 24 hours. After that, no longer.

The undone run does not vanish from the history but stays greyed out in the list. Vanishing would be more convenient and poor evidence: even later on, it must still be possible to see that something had been read in and removed again on that day.

In the Import history. Every import remains listed with its result — even weeks later you can trace where a record came from.

This is more than tidiness: if someone asks where their data came from, this history is the answer.

One limitation concerns old entries: the list from the time before the run log only knows runs that touched Opportunities — pure contact imports were not recorded back then.

Enlarge

They are skipped if none of the identifying features is present, and otherwise created as well as possible. The dry run tells you beforehand how many are affected.

Skipped rows are neither an error nor a loss of data — your file stays unchanged. If the number of skipped rows is surprisingly high, take a look at the file: usually there is a header row in the middle of the data or a totals row at the end.

Three things, all on the same day — while undoing is still possible:

1. Spot check: open three or four contacts and check whether the values have landed in the right fields.
2. Classify: select the new records and set a category or tag. Without classification they are useless for campaigns.
3. Count: does the total match what the dry run announced?

Only then plan the campaign — not before.

Because the postcode has not been mapped. The map in the reports is calculated from the contacts’ postcodes; without them there is no location.

Check in the mapping whether there is a column for the postcode and whether it is assigned to a field. If the postcode is in the same column as the town (“01129 Dresden”), it must be split off in the file first.

Five steps that turn the import into a matter of minutes:

1. One header row, right at the top, with meaningful names.
2. One piece of information per column. Postcode and town separate, first name and last name separate.
3. No intermediate headings and no totals row at the end.
4. No merged cells and no filters.
5. Save as .xlsx if possible — then the question of the delimiter does not arise at all.

What you do not have to do: rename columns so that Aurora Hive Sales recognises them. Detection takes care of that.

Yes, and that is often better than one big file: each import is listed separately in the history, can be undone separately and assigned to a source separately.

In practice this means: one file per source — trade fair, directory, existing list — instead of merging everything in Excel first. Set the matching tag straight after each import; then you will know later where a contact came from.

Yes — that is one of the cases the import has been built for. Existing contacts are recognised by their email address and added to rather than overwritten; an empty field in your file deletes nothing.

Two things you should know:

• Changed email addresses create a new contact, because recognition depends on the address. You can find such cases via the company file.
• The dry run tells you beforehand how many will be new and how many added to. If the ratio does not match your expectation, something is wrong with the file.

Then the character encoding of the CSV does not match. This is a classic Excel problem: when saving as CSV, Excel does not always use the encoding that other programs expect — German umlauts are the typical victims.

Two ways out:

• Save as .xlsx and use that file. The problem does not exist there.
• Or select the cells in Excel, copy them and paste them into the paste field.

If it has already happened: undo the import within 24 hours and repeat it with the better file.

Rows with the same email address go to the same contact: the later row adds to what the earlier one created. So you do not get a duplicate contact — but the values of the last row win for fields that both fill.

If your file contains contradictory details for the same address, this is the moment to clean it up first. The dry run reveals it indirectly: the number of new contacts is then smaller than the number of rows.

The same procedure and the same preview before anything is written — only the fields differ: for opportunities it is about title, client, location, source, link to the source, deadline, period and volume. If the file also contains contact persons, additional contacts are created from them with the tag “Auftrags-Import” (opportunity import).

It is opened above the opportunity list with “↑ Import opportunities”. These runs also appear in the Import history; historically they are even the only ones that appeared there from the start.

Three things people drag along out of habit and regret later:

• Free-text columns with mixed information — “Notes” from an old system, in which appointments, ratings and phone numbers are jumbled together. What you cannot filter does not help.
• Dead addresses. Anything already considered undeliverable in the old system does not belong in the new list — it harms your sender reputation.
• Columns nobody can explain any more. If nobody in the team knows what “Status 3” means, it means nothing.

Anything not mapped is left out — nothing is lost, it is simply not imported.

For everyday use there is no noticeable limit — lists with a few thousand rows are no problem. What increases is the time for the dry run: it calculates the whole import.

With very large data sets, splitting them up still makes sense, but for a different reason: undo always affects the whole run. Anyone who imports a hundred thousand rows in one go and discovers a mistake undoes everything.