Skip to main content

Data Driven Governance Techniques for DAM Success

By May 24th, 2022Blog

If you’ve ever been in a conversation with people who run digital asset management (DAM) systems, then you’ve probably heard the term ‘governance’. That’s because, among the DAM community, users and vendors alike, it’s understood that long-term DAM adoption across the business is essential for the system to achieve its ROI objectives. But, what does that even mean?

In simple terms, governance is the policies and procedures which provide guidance to users (and rules for systems) that enable the DAM to operate smoothly. An effective DAM allows organizations to obtain the full range of business benefits they can derive from both the technology and the people who are employed to manage them. The phrase “people, process and technology” is widely used in relation to DAM. Governance can be thought of as the glue which connects each of these three factors together to deliver a successful DAM strategy.

Governance and the Role of Data

Although it is advisable to set up a governance best practice program right at the outset of the implementation of a DAM initiative, it is virtually impossible to anticipate every possible issue that might be encountered in advance. For this reason, a governance committee is required to establish new or updated policies to deal with all the different issues which will be encountered throughout the lifecycle of the DAM solution.

The task then becomes how to determine what the impact might be of making changes to procedures, configuration and other rules. Further, to discover what has happened since they have been modified. Any decisions made must be informed ones and, if it subsequently becomes apparent that they were not optimal, evidence for that should be readily on-hand so adjustments can be made.

Actionable data is essential for effective DAM governance. Although theoretically this can be collected externally from the DAM solution, for practical purposes, this should be built directly into the DAM. It is very easy for vendors and consultants to come up with platitudes and all-encompassing phrases that few will argue with, such as ‘governance is good’ and so on, but in the real world, DAM users must make potentially complex decisions and be ready to deal with the outcomes. In this article, I want to analyze how data can be used to support this process and enable far more effective DAM governance than might be possible if decisions are being made without empirical evidence to support them.

Example Governance Issues

Here are some examples of real world governance issues I have encountered in my consulting practice and how these got solved using data from the DAM systems in use at the time:

Issue #1

A group of stakeholders was finding they were searching for text they knew to be in sales proposals, but their results were coming up empty. The DAM system had the ability to index text in documents but this feature was turned off. Not long after the feature was enabled, the number of users accessing the system began to decline. The governance committee had to find out what had occurred.

Firstly, to even know that this sort of issue is occurring, the governance committee needs to be meeting on a regular basis (once a month is a reasonable minimum). They also should be reviewing basic reports such as the number of users logging into the DAM. The second item of data to look at is whether one group is over-represented. In this case, the stakeholders who had the problem before no longer encountered it because they were able to find what they were looking for. Their log-ins had increased, but there was now a decline in the corresponding number by those from other business units.

Some reports received by the vendor’s helpdesk revealed that more users were reporting the system ‘had bugs’ because they were getting lots of results which were not relevant to what they were looking for. Because of this, certain users were losing confidence in their ability to find anything useful in the DAM and therefore, not using it.

After some analysis of the asset results in test searches, it was discovered that the sales proposals included a lot of legal text and lists of countries which generated numerous false positive search results, especially for those who were not interested in that type of asset.

The resolution was to establish a metadata classifier for sales documents and ask the vendor to only enable text indexing of these documents. This could be applied immediately without any downtime or development work being necessary. The result resolved the problem and after an investment into more training and user awareness programs, numbers began to increase again.

Issue #2

The marketing personnel of a large professional services firm complained that there were not enough images of people for them to use in campaigns and the same photos appear all the time. They wanted to commission a photographer to shoot more images of staff for them to use.

An analysis of the digital assets held in the system indicated that there were a far larger number of photos of staff than many of the current users were aware of. The staff images that were being downloaded had higher quality metadata and fewer access restrictions, this meant they were more likely to get found and be used.

Rather than invest in more content creation, some picture researchers were employed to improve the metadata of the less visible staff images. The owners of the digital assets in question were also asked to allow more to be made accessible to those outside their own business unit and to remove download restrictions where they were not absolutely necessary. The candidate assets were assembled into discrete collections so the DAM tool’s batch modification facilities could be used to adjust the permissions as a group to avoid the necessity to make manual changes to each of them.

The result was that more images of people were available for the marketing department to use. This limited the need to spend money on commissioning new photography and enabled more efficient re-use of existing digital assets that the business already owned. This type of intervention was only possible because:

  • There was a system which was collecting data about digital asset usage
  • There were tools in the DAM to allow this data to be analyzed
  • The DAM made it straightforward to update digital asset permissions as a batch operation rather than having to do them piecemeal by editing each record

 

Without access to high quality data which was collected in the DAM itself, this would never have been feasible and the business would have spent money unnecessarily.

Issue #3

The ratio of searches to downloads on the DAM had recently increased (more searches, fewer assets being downloaded). The DAM governance committee was unsure why this had become the case.

This could have been happening for a number of reasons, for example:

  • The asset metadata quality might have been insufficient and users were not finding what they needed
  • There could have been a lack of the type of assets required by users at the time

In this instance, the problem turned out to be both the above. Firstly, a decision had been made to enable AI image recognition and use it as the default when assets were uploaded. This technology can have some benefits, but it also has risks and requires careful management. It is more likely that false positives will be generated in search results. In terms of governance, a safer approach is to allow AI image recognition to suggest keywords rather than insert them automatically. This will not apply in every case, but it was with this particular scenario.

As well as lower quality automated metadata, further consultation with users revealed that many were searching for a new product (and all the metadata associated with it) but very little had been added to the DAM. In this case, governance best-practice should coordinate different sections of the business so that key events (such as the introduction of new products) are anticipated and planned for in the DAM before it takes place.

This is another example where to even be aware that this issue exists, the DAM needs to be collecting rudimentary auditing data like searches and downloads. Secondly (and possibly more importantly), one or more human beings need to be looking at these reports as part of a regular governance review. This last point segues into the other key consideration when it comes to DAM governance and data.


Does your DAM tell a meaningful story about asset performance? Watch our webinar, DAM Data 101, to learn how you can get the most from your DAM data.

Human Beings Should Reality Check Data

As can be seen, a number of governance issues can relate to conflicting requirements between two or more groups of users. A positive advantage for one group could create negative effects for another. Getting to the bottom of why this is the case requires both quantitative and qualitative approaches. In addition, some awareness of what is happening across the business is essential. The data alone cannot provide this and there is a risk that it may tell the governance committee a story they want to hear rather than the one they should be paying attention to.

DAM initiatives cannot be managed in an entirely reductive fashion with reference to data alone. Feedback from human beings is also essential to reality check what the data appears to be suggesting. Ideally, empirical evidence should support a hypothesis made by the governance committee. In other words, the tail should not wag the dog.

With that being said, relying exclusively on user feedback is risky also. If one group of users is highly vocal, makes frequent representations to the governance committee and deems that the DAM is not meeting their objectives, they may force changes to be made which will have negative consequences for all the other users (who may be a larger, but more silent majority).

The two perspectives represent the ‘technology’ and ‘people’ aspects of the DAM equation. While the latter is undoubtedly the most important, the technology (i.e. what collects the data which represents what is really happening) should validate assertions made by human beings. ‘People’ is a description which groups together all the human beings involved, but this is not to say there are no conflicting interests within groups of them.

DAM governance will frequently involve some trade-offs and it is not possible to keep all users happy all of the time. The data collected (hopefully automatically) by the DAM, should provide a rational basis to make a more scientific analysis of whether a decision will benefit the majority of users or not. Further, it should help DAM users demonstrate to vendors where there is a proven need for more sophisticated and granular features such as being able to enable functionality for specific groups or types of assets.

DAM Governance Should Be Based on Data

Digital asset management is all about appreciating digital asset value – in both senses of the word. It is not just acknowledging that digital assets have a value, but also pursuing active strategies which increase it. In DAM initiatives, the whole is very much greater than the sum of the parts. A DAM is more than a collection of digital files on a server somewhere, the repository itself has a value which is both composed of individual assets and enriches each of them by virtue of being a part of a wider collection.

In order to satisfy both interconnected objectives, effective governance based on rational processes that are informed by data is vital. A data-first DAM is therefore essential for modern DAM governance processes, which (in itself) is probably the major factor behind the success or failure of an enterprise digital asset management initiative.


Ralph Windsor is Project Director of DAM Consultants, Daydream and Editor of Digital Asset Management News.