A few years ago, I completed a Product Management course under Cornell University, and one of the projects was to create a prototype of a new product or feature and test it with real users.
Holup, before we get to that, we were required to do customer research interviews. I worked with DAM administrators and Creative Operations Leads one-on-one, and we talked about the different asset-related struggles they had and the real challenges they faced. One theme that came up back then was inconsistent tagging.
It's not uncommon to get a request to find that needle in the haystack, or rather, the exact 13 needles that I specifically remember we have, which may or may not be scattered across 5 or 6 haystacks in various barns. If you have a meticulous filename, foldering, and tagging system, that's great! But most teams don't have that and are overworked and under-resourced.
Case one: the lone tagger.
I remember one of the designers I worked with regularly had a backlog of many hundreds (even thousands) of assets that she needed to individually open, inspect, and tag. And sometimes that tagging took a bit of cross-referencing and research. The advantage of having a single human do this is quality and consistency, but the down-side is time suckage.
Many teams will bring on a summer intern, and one of the things they have them do is manual tagging, since 1. it gets them familiar with the content, 2. it can likely be done at a decent level with little training, and 3. no one else wants to do it :(. In this case the quality is low, but at least the assets get tags on them.
Case two: tagging by committee.
In a third case, the asset-tagging work is shared among different people on the team. Since they are experienced team members, the tagging should be high quality, but there can be consistency problems. One person may tag in a slightly different way than another, because they probably have a slightly different point of view on the asset lifecycle. One person may spell things slightly differently than another. Sometimes it's a capital letter, or a hyphen, or an underscore, or a compound word, and every possible variation of this single field value can exist in the system.
The difference between finding all 13 needles and finding only 10, and the one you missed was the important one.
So just govern it?
To resolve this case, we need a way to govern the tagging and the tag taxonomy. We need a controlled vocabulary and a council to enforce it. Unfortunately, this is difficult to implement in real-world scenarios. Imagine you are one of 5 people tagging your assets, and every time someone thinks they need to add a new tag to the taxonomy structure, a meeting has to be called and the viability of this new tag has to be discussed. It is utterly impractical.
Have you felt any of these problems in your team? What can we do? I'll save that for the next journal entry.
Field notes from VDAM - written by a human.