Have you ever set out to make a minor improvement to a process that has always been in place… only to end up transforming the way you work from top to bottom, to the great benefit of several colleagues? It's a rare situation, but it's exactly what happened during a recent development for one of our clients in the publishing industry.
At a glance
- Client: a publishing house, with a custom FileMaker business application supported by IU DATA for more than 10 years.
- Problem: weekly manual re-entry of upcoming titles on the distributor's platform.
- Solution: XML export in ONIX format, manual and automatic submissions, distributor feedback imported back into FileMaker.
- Result: 180 hours freed up in the first year, up to 280 hours per year afterwards; return on investment estimated at under 18 months.
The starting point: re-entering data every week
Our client, a publishing house, uses a custom FileMaker business application that we have been developing and maintaining for more than 10 years.
Every week, part of the sales team had to send the distributor information about upcoming releases. In practice, this meant manually re-entering, on the distributor's online platform, all the information about the "products" due for release — a product being a specific release of a work (an edition, a format, a translation…). This re-entry covered one or several weeks of releases, sometimes far into the future.
The hidden cost of this task was significant:
- time lost every week by one or more people;
- risks of errors and omissions, requiring regular proofreading;
- records to be completed or corrected as information became available;
- two tools to master instead of one — the FileMaker application and the distributor's platform — with different structures: not all information is entered the same way or in the same place.
The client's request: a first step towards automation
A little over a year ago, the person in charge of the application came back to us with a simple change request: the ability to export a basic Excel file, with just a few columns, listing the products due for release soon.
The idea was to email this file to the distributor, whose staff would import it into their own platform. Additional information would still have been re-entered by the client's staff, as before.
But an opportunity came up at the same time: the distributor was reviewing how publishers could send it their data, gradually introducing the upload of XML files in ONIX format — the international standard for exchanging book metadata, created in 2000 and recommended in France by the Syndicat national de l'édition, Electre and Dilicom.
Goodbye Excel export, hello structured XML export!
The challenges: constant evolution
Talking about XML may come as a surprise at a time when APIs are everywhere in exchanges between online services. But every industry moves at its own pace, and the book supply chain has long been organised around ONIX — including at the largest French distributors.
An ONIX export requires more work than a simple JSON API call. The documentation sometimes left grey areas: a lack of precision on a given type of information, an unclear expected structure for a specific tag… Many back-and-forth exchanges with the distributor's team were needed to resolve these ambiguities.
The export therefore evolved almost every week, through meetings bringing together our team, the client's and the distributor's, over several months (nearly a year including a few breaks). Several factors explain this timeline:
- The richness of the standard. ONIX has existed for more than 25 years and offers a very large number of fields. This allows great precision, but increases the complexity of the development accordingly.
- An integration still in progress on the distributor's side. Not all of the standard was supported at the start: some fields considered important by our client had to be added along the way.
- Rules that vary by product type. A new release does not require the same information as a translation, for example, and some fields become mandatory or optional depending on the case. Other fields were added as new needs emerged from the client or the distributor.
- Richer submission modes. From manual, record-by-record submissions, we moved to batch submissions of several products (selected with a FileMaker find), then to automatic weekly submissions of upcoming products.
- Two-way exchanges. After each submission, the distributor sends files back to the FileMaker application, making it possible to check that the data has been properly integrated and, above all, to retrieve the reference number assigned to each record, stored on the corresponding FileMaker record.
Finally, we added a prior validity check: a visual indicator shows at a glance whether all the mandatory fields of a product have been filled in before it is sent as ONIX.
What the development delivered
After several months of work, the complete process is in production and used every week, for both manual and automatic submissions.
Measurable gains. According to the client's estimates, the automation frees up about 180 hours of work in the first year (around 26 person-days), then up to 280 hours per year in the following years (around 40 person-days). The difference reflects the adoption period, the adjustments planned for the first year and the related monitoring checks. These estimates are based on sending about 10 to 15 ONIX files per week — whereas, since going live, many more products are sent automatically every week. Again according to the client, the return on investment is estimated at just under a year and a half.
More peace of mind. The data is now identical on both sides, with no gaps or omissions, since re-entry has disappeared. And the information automatically sent back by the distributor removes the need to keep checking its platform to see whether the data is finally available.
Stronger security. Files are automatically uploaded to a secure SFTP server, instead of the occasional email exchanges used for special cases — which had also been planned with the initial Excel export.
Unexpected benefits. Faster, smoother submissions to the distributor also mean faster distribution of this information to the associated sales platforms (online bookshops, Amazon, etc.): titles get commercial visibility sooner.
A win for the distributor too. It was its first ONIX integration with a publisher outside its own group. All the cases encountered allowed it to test its integration in real conditions and to anticipate many situations it will now be able to handle for its future publishing partners.
In conclusion
The lesson from this project: a change request, however small, deserves a look at its "why". By questioning the real need, you can step out of the "tunnel vision" that sets in when a process is so ingrained in daily routine that you can no longer imagine improving it — and discover a feature far more useful than the one originally requested.
The time invested may seem long at first. But the hours saved every week for years to come, and the peace of mind that comes with them, make it a real day-to-day asset.
Do you also have a process that relies on re-entering data? Let's talk: that's often where the quickest wins are hiding.

