In a DAM, Speed is all that Matters

We are all impatient. Especially when sitting in front of a computer: waiting for the answer to your question is just no longer tolerable. We are all used to <1 second response times from Google, and we are mostly OK when the thing we are looking for is on page 5 of the results. Google has trained us, and we want the same responsiveness from our DAM.

There are three main areas where speed matters in a DAM system: input, searching, and output. Let’s look at each of them.

What if you are an entertainment company, or a clothing designer, and the Red Carpet is a very big deal for you: you are expecting photos to come in and you need to get them up on your company website and on social media pronto! You see the situation you need on live TV, but when does it pop up on your DAM screen so you can put it to work?

Let’s trace the course of the terrific photo you are waiting for, from shooting at the Red Carpet to your desktop. The photographer shoots it, adds caption information and uploads it via WiFi (via email, FTP, or some upload process) to your DAM. There it pops into a folder with stuff coming in from all your input channels, some of which arrived ahead of it (the backlog). Your DAM chews thru each preceding object one at a time, and for each one it has to extract the metadata, make thumbnails of different sizes, insert a record into the database, and index all the fields so it is searchable. How long does all this take: “it depends”. How big is the backlog, how slow is your thumbnail generation, how quickly is a new record indexed?

Best practice, by the way, for a special event is to setup a special high priority input queue just for Red Carpet material, so no other incoming traffic can slow things down. A good DAM can input a new object in ~250 milliseconds, so there should be no waiting.

Now we come to search. A well-designed DAM can let you mix things like text searches with date ranges (“Taylor Swift in the last 30 minutes”) with no performance penalty, and get you results in under a second. Other systems break that query into two different searches, then determine the overlap, and then return the results to you, which can take a lot longer.

Finally output. What if you need to take this great photo and output it in your three standard channels: in three different crops, three different resolutions, and put watermarks on two of them, and send them to three different folders so other people can put one on your company website, another on Facebook, and a third to Instagram? Do you have to output three copies, open each in Photoshop, do the crops and resizes and watermarks, and save each and then drag them to the right folders? How long does that take: 10 minutes? Or can you just drag that photo to an icon in your DAM, pre-programmed to do all those things automatically while you move on to the next thing (and all three corrected images go to the three locations in seconds)? Less mouse clicks, and faster response: a DAM should not be a silo in your workflow, but instead it should have lots of automation capability to take all the grunt work off your shoulders and make the computer handle it.

So speed involves different things at different stages of our Red Carpet experience. All require good design and very efficient software to keep up with the speed of you!

Leave a Reply