In common with most types of business oriented software, DAM systems need to have business case that can demonstrate an opportunity to gain a positive ROI (Return On Investment). While all this is obvious stuff that you don't need an MBA to understand, proving an ROI case is harder than it might superficially appear.

In this article I am going to critique one of the more popular methods that is often used and discuss alternatives which I believe are more realistic and useful.

Statistical Follies And Quantitative Farces

The kind of ROI data that corporate bean counters like is typically numerically oriented and reads as though someone has gone away and scientifically measured activities using a precise and rational method. Statements like "we can save x per employee" usually go down well.

Most pre-implementation statistical evaluations will be composed of some ROI propositions -- with search related activities being prevalent (as you would expect for DAM). A common method is to then work backwards from each of these benefits and estimate how long they take as an average. For example:

Average Search Time Per Hour x Employee Cost Per Hour = Current Hourly Cost To Business

Theoretically, this tells you how much it costs to search for assets without a DAM. Various analyst supplied stats about how much time you can save by using a mythical generic DAM system then get quoted (30% seems to be popular). The discount factor is then applied to tell you the ROI.

This sounds reasonable, but there are assumptions and fallacies baked into this technique that, in my opinion, render it unfit for purpose.

Firstly, it is unlikely that anyone ever measured the length of time required to find assets before they planned to implement DAM. So that poses the question, how has the average search time been derived? The duration required to find assets is non-trivial to predict.

For example, the artwork you were reviewing this morning you can probably find in seconds. Something from last week, maybe a minute or two. Photos of discontinued product from several years ago could take hours or even days. On other occasions you might be persuaded to cut your losses after an unsuccessful search and use an alternative you can find, even if not exactly what you were looking for.

Some might argue that you can still average these durations out to identify a mean figure. That might be correct, but when are you going to start and end your sampling period? If a succession of hard to find assets are required, the average figures can trend upwards (and vice versa too). To get a properly representative sample, you need to carry this out over a period of many months -- usually far more than the limited time available to decide whether you should be investing in DAM or not.

Properly measuring the time needed to find digital files is difficult to do accurately at scale. Kitting out every prospective DAM user with stopwatches and something to record the results is hardly practical.

Instead, what usually happens is that staff get asked to make an estimate (or more likely a guess) about how long they think it takes them to find assets on average. That prompts a series of HR issues also since they might be unwilling to admit that they spent hours searching for something trivial because they (or a colleague) filed it in the wrong folder and no one noticed until they tried to find it again.

While DAM systems give you some tools to allow you to aggregate assets and deal with them as though they were generic commodities, the reality is that each asset itself is usually a one-off instance with some characteristics that might make it uniquely valuable. You cannot appreciate exactly how much until you need to find them. If it were the case that assets were commodities, you would not have to sift through thousands of records to find the exact one you need and worry about topics like findability etc.

This raises the other issue, who comes up with these average percentage savings made possible by DAM and what is their exact method for working them out? How do they explain why a generic average can be applied in all cases?

Software applications are mechanical in nature and merely having a licence to use one is not sufficient to generate a cost saving. There are many additional factors that can have a bearing upon the productivity improvement available. A less versatile DAM can generate a higher ROI than a more fully featured system that is managed by those who lack the expertise to use it properly. In other words, ROI is heavily influenced by characteristics of the implementation rather than the technology itself.

The Hidden ROI In DAM

There are a range of hidden ROI benefits to DAM that cannot be easily measured pre-implementation because they are only available once a DAM solution is in place. Here are a few:

Centralized Access

One of the advantages of DAM is the ability to put everything into a single location so everyone knows where it is. The amount of time saved by not needing to switch from different storage locations all adds up. This might only be a few seconds, but it will be repeated daily hundreds or thousands of times.

Efficient Asset Distribution

You might have a fast internet connection that can ship megabytes of data around quickly, but someone still has to find the file, attach it to an email (or upload it to a third party file sharing service) and then send it off. With most DAM systems that can all be automated so you can quickly build a collection of assets and dispatch them direct from the system without the associated upload/download rigmarole.

Once an automated system for keeping track of everything is in place, it is much easier to get precise and accurate usage figures. The intelligence gathered can inform a wide range of operational activities in relation to digital media that you would not have known otherwise.

There are many more examples. The potential to enhance productivity with DAM is not under question, it is the method by which a generic calculation gets applied before any testing has been carried out for a given usage scenario.

Alternative ROI Techniques

If many of the current methods of ROI calculation are unsatisfactory, what else can be done? Asking senior managers to take a leap of faith that DAM will produce ROI because you said it will is unlikely to be well received. The answer is to revise the strategic approach taken towards DAM implementations and focus more on risk management and the objectives or outcomes the organization is seeking to achieve.

Risk Management

The traditional method for procurement of DAM software and services has some fundamental flaws that place businesses at a higher risk of failing to achieve ROI. This is because to get a DAM initiative underway, the sponsors are required to negotiate a budget that includes the full expected cost up-front. Going back to ask for more money isn't a pleasant experience and calls into question all the other claimed benefits of the DAM initiative and the skill of the managers involved too. There is a built-in motivation to get a bigger budget than might be needed -- just in case.

Further, once the word gets around that funds are being set aside for DAM, lots of other parties then try to get their pet projects loaded into it too. So the scope (and cost) of the implementation increases exponentially. This top-down method transforms an ROI analysis into an internal PR and marketing exercise which helps no one.

Analyzing ROI from a risk management perspective suggests that a more effective technique is to start with the smallest feasible implementation that aims to address just one single critical issue, test the results and adapt implementation priorities based on real data.

This is a fact that is often missed about DAM, having a system in place can help you collect ROI data to help direct future investments into it. A risk management oriented ROI strategy enables you to avoid excessive capital expenditure to subsequently find much of what was supplied was unnecessary or didn't work out as expected.

Outcomes-Oriented ROI

Another method for improving DAM ROI is to focus on targets the organization needs to achieve where digital media assets might get used. For example, the ability to find assets more quickly is not the ultimate goal of most organizations, it is some other business related objective such as delivering marketing campaigns faster and more efficiently.

This works well with the risk management methods described above as the organization can start doing DAM using some very basic (and cheap) solutions first, evaluate the benefit and then use the data collected to decide if a further investment into DAM might yield results or not.

If it does not, the strategy can be re-assessed. It might be that spending more on DAM will not yield any significant gain. It could be that some other related activity (e.g. improving findability) could produce positive results or maybe adding further features and building out the solution might help.

Conclusion

An observation made by a number of people is that "roi" is coincidentally the French translation of the English word "king." It certainly is true that ROI should be the ruling concern when implementing DAM. It is important, however, to remember what the French did with Louis XVI and make sure that your ROI case doesn't get accused of lavish self-indulgence and dispatched to the guillotine when the time comes for it to be judged.

Image courtesy of Anthony Berenyi (Shutterstock)

Editor's Note: To read more by Ralph, see his The Digital Asset Management Value Chain