Last Wednesday, 3M launched the fourth generation of its Visual Attention Service (VAS) scanning tool, which includes new video capabilities. According to 3M’s release, VAS is viewed by marketers as an “alternative to eye tracking that enhances the visual impact of professional design across advertising, marketing, and digital and print media.” Read the release.
Kelly Canavan, Marketing Manager of 3M’s Commercial Graphics Division, discussed the announcement and 3M’s approach to the business of advertising.
AdExchanger.com: Can you talk a little bit about 3M’s strategy in addressing opportunities in digital, in particular, and advertising and marketing, in general?
KC: 3M tries to find [new] businesses that are adjacent with existing 3M businesses. Once it decides whether it will get into a space, the company asks itself: Will it create products for that space as a supplier?; or, will it do things itself?
What we're doing related to the [digital advertising] space is, we have entered into digital out-of-home.
Our initial approach to digital out-of-home is as a supplier. As it relates to advertising and marketing, our first step was to get into the business of providing the software and networks that allow people to advertise and do marketing in physical environments like retail stores, hotels and such.
What we’ve realized is that there are some critical needs to take digital out-of-home (DOOH) over the hump. It's a tiny fraction of the advertising world right now –and, DOOH may be a bit more important when it comes to merchandising.
Just to be clear about the difference: advertising in DOOH is where the insurance company, GEICO, plays an ad in a grocery store. You can't buy GEICO at the store, of course. But if P&G advertises in the grocery store, you can buy things from P&G in the store - merchandising.
DOOH is getting more of a foothold in the merchandising category. Yet, there are different worlds of people who are focused on advertising versus merchandising in DOOH. The market today, though, uses advertising and in-store media without really making that distinction.
It’s evidence of the nascent nature of this little slice of marketing that still hasn't really defined itself.
So what went into 3M choosing digital out-of-home as a market it wanted pursue?
We have a couple of established businesses and the one that's most related is our commercial graphics business. That division produces large format vinyl that's used for outdoor advertising and for signs for companies - the actual sign out on the road, window graphics and all sorts of wall coverings. We even have architectural market products that are used in the interior of retail spaces. So we've got some adjacent products, and the thought was, "Well, we're helping people with premium quality graphics products to express their brand. Let's help them with the digital version of that."
So - do you think you can do audience buying in digital out-of-home right now?
I think you can try to buy audiences on the digital out-of-home networks but it's a drop in the bucket. For now, it’s, "Let's just learn how this thing works." Because you can't do it at scale, certainly not until somebody comes in and aggregates the different little networks out there to form one. Right now, people just need to learn about [DOOH] and get their hands dirty with it.
There are things you can do along the way that are beyond what the obvious audience buying which relates to a new software product we have. It came from a team of 3M neuroscientists where one day one of them came in and talked to us regarding something completely unrelated and said, "Hey, we are working with software that predicts what people are likely to see. Would that be interesting?"
That's how cool connections happen here at 3M.
So, how is 3M addressing audience buying, or perhaps a precursor to that world in digital out-of home?
I don't think we are going to address audience buying until later. We're going to focus more on the behavior, more on the merchandising side of the equation. We're focusing on digital in-store marketing where people can buy products. We’re looking at behavior change as ROI rather than impressions.
Got ya. First, what is it that needs to happen “later” to fuel audience buying in DOOH?
It’s really all about scale. Networks will be put in for both [advertising and merchandising] goals. And, there will be networks put in public spaces that are really about advertising. And then there will be networks put in retail locations - and I use the term very broadly - where companies want them to buy or do things [in-store]. When there are enough of those networks and they get a feel for how they can collaborate, then those networks can combine and be part of a larger national or ultimately, international network. Think about it in terms of moving the Internet to the physical world.
Then [network players] will start to get a feel for the ground rules. Right now - and when there's a new medium out there like DOOH - people really get a little concerned about what the rules are. Where should you start and stop?
Everybody in advertising is thinking about digital networks as an advertising network. Everybody who is in promotion and merchandising is thinking about it as a promotion network. So they're going to have to mature a little bit and realize that they're really the same thing and it can be used for both simultaneously.
I can see, in a grocery store setting, marketing messages for General Mills alongside marketing messages for NBC Universal. But, in a hotel environment that wouldn't necessarily work, as the hotel might think that advertising is a bad thing because it's against their brand experience. As digital out-of-home matures, some will allow more and some will allow less.
Can you talk a little bit more about the Visual Attention Service (VAS) tool and how it helps predict behavior?
When you think about the challenge of getting advertising noticed on television or on the Internet, consumers are paying attention... somewhat. But when you get out in the real world, you’re driving or you're walking through a store and you have other things on your mind - kids in tow, etc. Trying to get somebody's attention becomes a lot harder.
What the [VAS software] model does is break down the image or video into its elements. We use vision’s [elements] - the same elements that designers use: color, contrast, spatial relationships, shapes and motion, and so forth. VAS then breaks it down, analyzes the data and puts together a prediction of what people are likely to notice when they're first scanning that image in the first few seconds.
In a way, you could say “That's all it does?”, but that's an amazing first step of figuring out what is a visual hierarchy within an image or a video. If you're a designer, you can use it while you're designing your creative and adjust accordingly.
What you're doing is tweaking your design so the thing you wanted to get noticed first has a fighting chance.
This sounds like it could be for TV, digital out-of-home, display ads online – thoughts?
It has universal application. Frankly, in commercial or non-commercial settings.
One of the retailers that we've talked to is really excited to do a walk-through of their space. So the retailer will be able to walk through the store, up and down the aisles and get valid feedback on what people will notice and that they normally would have to gather themselves with a long, drawn out eye‑tracking study.
An interesting distinction, though - it doesn't tell you whether a particular marketing method is going to work, that's the job of the creative. But it does help you get noticed and that's a huge first step.
How is the tool made available?
Right now there are three ways that it's available. First, you can go sign up online and analyze photos and creative layouts.
Another way is that it's built into Photoshop so designers can use it while they're developing the creative itself.
The third way it's available, is as an iPhone or iPad app. The Android version is in the works, too.
So, if I'm a marketing or merchandising person for a company that's out in the field, or maybe I work for an outdoor advertising company, I can take shots from the field and get a quick analysis right on my iPhone or iPad.
Let me take you back to digital out-of-home in general. Where are we with success metrics?
Yes, that’s a story in itself. Again, there are two different categories of [DOOH network] users - advertisers and merchandisers. Metrics in advertising are about the audience: "What is your best estimate of the people walking by and what do you know about them?"
For merchandising, the goals are, "Did I sell more, or did I sell different? What behavior change did I impact amongst the people who saw the content?"
So with advertising, it's about impressions, and looser, because it's always been that way.
The bigger problem I see is in the merchandising side because it wants to figure out if there was a change in sales -a harder metric. That industry is kind of a mess. The way that people measure today is by looking at same store sales (before and after signage installed) or they put digital signage in a certain number of stores and compare the numbers [to stores that don’t have the signage].
But this is the best that business and marketing can do in these correlation‑based approaches. You look for correlations between the content and sales.
Having worked here at 3M and working side by side with scientists, that method of trying to measure ROI is really fraught with problems. Because you can't control for what’s happening out in the real world, like your competitor's advertising. Or, maybe there are weather events or road construction, etc.
The exception of that in marketing is the Internet because it has cookies and clicks and can measure scientifically and you can conduct experiments.
Because of the dollars involved in DOOH networks and because of the complexity of something so new, ROI measurement needs to move away from the conventional way of looking for correlations and move towards the Internet’s way - looking for a real cause and effect.
That's what we're pointing ourselves at - creating a quality measurement capability and embedding that into the operating system that you use to do marketing. You won't need to be a market research expert – the intelligence will just be built right in.
Email This Post