The DOOH Advertising Paradox: Better Spots Are Harder to Sell

Author: Bill Gerba on 2009-11-17 11:21:25

I received a trio of comments and another trio of emails over the past few days that converged on two topics: pricing/buying/selling ad space, and picking the appropriate size and number of screens for a given venue. In fact, people ask me about both of these things quite frequently. While my in-person answers generally take into account the unique situations of my questioners, I thought it might be useful to lay out some of the basic arguments and logic that I use to make my recommendations.

How screen size, shape and placement impact ad performance

Image credit: Amy Gizienski

Reader Lisa writes:

What do you think would be the best use of digital signage within a small store such as a convenience store, supported by advertising? From an advertiser's point of view, is option 1 or 2 better in your view?

Option 1: A number of smaller (13"-15") screens at shelf-mount or endcap at just around eye level

Option 2: One or two 40"-46" screens at the back of the wall hanging off the ceiling?

This one is pretty straightforward. Except in some very specific scenarios, shoppers simply will not look up. We've studied so, so many networks where screens were dropped from the ceiling (and let's face it, that's how most of them have been done up to this point), and the uniformity of our findings has been pretty remarkable. As we noted in our article on Digital Signage Screen Placement: Modeling Consumer Behavior, people like to look down on their purchases: we pick things up, hold them in our hands, and generally look down at them while doing so. Additionally, in our article about Digital Signage Screen Placement: Angle, Height and Text Size, we found that putting screens high up can increase the raw amount of time that the screen would be visible to the average shopper. However, it actually decreases the amount of time that the screen will lie in the shopper's zone of attention.

So here's my short answer. From an efficacy standpoint, you'll almost certainly see better results with the smaller screens on the endcaps and/or on the shelves, if your host venue will let you put the displays there. That's prime sales space for them, so the typical retailer is loathe to give it up for screens. But you might get lucky, or you might be able to reach some kind of compromise. The second part of your question, though, is which approach advertisers will prefer. That brings up another set of considerations.

The best-performing DOOH ad placements may be harder to sell

How do you price and sell ad space on a digital signage network? This is, hands down, the #1 question I get asked, which is ridiculous considering that I've never sold an ad in my entire life. However, having worked in advertising for a while, and having been party to a number of transactions over the years, I've carried away a number of observations that leave me scratching my head whenever anybody asks why their ads aren't selling, or what they could do to improve sales.

What do I mean? Well, let me give you an example using the convenience store scenario that Lisa asked about. I'm 99% positive that a number of smaller endcap or eye-level screens will perform better (i.e. relay more information to a greater number of shoppers) than a few screens placed high above eye-level. We simply have too much data indicating that. However, depending on who your advertisers are and how they like to buy spots, the performance advantage might not actually matter (initially, at least). That's because big advertisers are still likely to use a very broad metric like CPM when making their purchases. Since you can probably make the case that having big screens up high will result in a much higher number of viewers, that translates to a much lower CPM if the cost-per-spot remains the same. So, it may seem like a better deal to your CPM-happy advertisers. As a hypothetical example, say that you, the network owner, wanted to charge $100 for an ad. You might find yourself running into a problem like this:

Screen placement	Number of monthly viewers	CPM	Face-value "deal" to the advertiser	Likelihood the ad will be effective
15' above floor (ceiling mount)	10,000	$10	Good	Low
10' above floor (wall mount)	6,000	$16.67	Fair	Medium
5' above floor (endcap)	3,000	$33.33	Poor	High

Basically, the entire digital out-of-home industry has a big problem: the best performing ad placements are usually harder to sell, because the raw number of viewers is low and the CPM is high. But on the flip side, if nobody watches the ads on the poorly-placed screens, nobody will act on them, so the advertisers will eventually pull the spots anyway for being ineffective. Because most media planners (the guys in charge of big, name-brand media buys) aren't generally able to "follow" a shopper from ad view to eventual sale, they can't know if a spot on a digital signage network is worth a higher CPM unless they try it out -- or somehow glean data from other advertisers who might have tried it out before them. Consequently, many networks rely on smaller advertisers to fuel the bulk of their sales. These guys are not only more accessible than a Coke or a Nike, but often they're better able to track their media investments (because there are fewer spots to keep tabs on).

So as the owner of a DOOH advertising network, your choice of screen placement may have a direct impact on your ability to price and sell your ads. However, relying on a placement that yields a lower CPM can be detrimental in the long run, especially if the screen is placed in such a way that the ads you do sell won't be effective. If you have the luxury of placing screens at various heights, you might be able to convince your advertiser to run some split tests to prove that some screens actually work better than others. But let's face it: that's probably not going to happen.

There's a lot more to this story, so next week I'll share some of the do's and don'ts that many of our clients have employed to get themselves in front of a larger number of advertisers, bring better advertisers on board, and work cooperatively -- not competitively -- with other networks that were facing many of the same problems as they were.

Have a question about screen placement or ad performance? Leave a comment and let's discuss it!

Comments (10)

Subscribe to comments for this article | Trackback

2009-11-17Dave Haynes writes:

Good post Bill

Where I stumble a little is on the screen placement and viewers thing. Why would screens at 15 feet up be three times as noticeable as screens at five feet? (And do people really put them THAT high??? If so, that's loopy). The OVAB guidelines (if you take two or three weeks and read them) talk about presence and notice, and I would probably argue 10,000 may be in the presence of the screens way up high, but only three in 10 people would notice them ... if that.

I agree entirely with the fundamental point that screens way up high, even at eight feet, are largely a wasted investment.

2009-11-17Stephen Ghigliotty writes:

As somebody who is actually selling ad space and working on DOOH campaigns, I couldn't ask for a more timely post...

Retailers are indeed loathe to give up an endcap space for a screen; even when it is designed for them. And that leaves us with too many networks using ceiling mounts and screens way too high.

What does mitigate that to some degree is synchronized content. The Walmart check out screens are very hard to ignore, even though they are still a little high; the combined visual impact of nine or so screens all running in sync demands attention.

2009-11-17Bill Gerba writes:

Dave: The 15/10/5 thing was just an example/not really indicative of what people are or aren't doing. But I've definitely seen screens hung at least 14" off of the ground, so I don't think it's too far off of the mark.

I agree the OVAB definitions are a good step in the right direction. What I have a problem with is that a) there aren't any widely-used systems in place for measuring "notice" or anything similar, and b) advertisers haven't exactly been clamoring to add OVAB to their planning suites, so we're often still stuck with the least-common denominator, CPM.

Stephen:

I agree, while it seems a little counterintuitive to some to have several screens all running the same thing, the effect (when done properly) is very hard to ignore. I expect to see more of that kind of thing in the future, especially now that more people are willing to experiment with smaller screens, screen combinations, etc.

Great points, both. Thanks!

2009-11-18Peter writes:

You can use ICapture (for celling and wall mount) and ICap (for endcap) systems from TruMedia, to generate all viewer statistics. It cost but you have data from every screen: who watching, how long,etc. For tests you can use this same content for all and it helps what place and screen size is effective...

2009-11-18Social comments and analytics for this post writes:

...This post was mentioned on Twitter by vjbagman: RT @NEOCAST Thoughtful comments from Bill Gerba on screen placement and advertising: http://bit.ly/1gIRoN...

2009-11-19Chauncey writes:

Bill:

Good post. Important to note that Lisa's original question regarded a convenience store. They see about 600 to 1000 shoppers per day per store for a five minute trip.

Your point on placement is derived from the two variables in the CPM metric. The C is defined by engagement and targetability and the M by sheer numbers of shoppers who can see the placement. It's an inherent tradeoff. But it's a non starter when you only pull 600 potential viewers into a site in a day vs cost of operation and equipment.

The only folks getting $30 CPM are in unique captive dwell like theaters. Transient dwell like the checkout monitors mentioned above are only pulling a fraction of that. Do the BE analysis and you can see the traffic you need before you even think about how many you actually convert into viewers. Icapture will help you confirm that you started with a tough business model.

This business is not rocket science. But it never fails to amaze me after a decade the assumptions that get made on traffic, cpm, and engagement.

2009-11-19Bill Gerba writes:

Hi Chauncey,

Yes, you're correct, my CPM numbers above are high. They weren't meant to be indicative of what a convenience store network might actually get, they were just to make the example math easy.

In reality, even at a low traffic number of 500 viewers/day that would equate to 15K viewers/month. Assume that 2/3 of them have an opportunity to see your screen and you're at 10K viewers/month, which is probably fair for most above-the-counter screens. Per Dave's post above, my guesses about the relative percentages of people who'd see something on the ceiling versus the wall versus an endcap were just that: guesses. They weren't meant to be authoritative, and I haven't studied any data about traffic flow in convenience stores, so they're almost certainly way-off.

The point of the table above was to show the relationship (which in the real world is almost certainly non-proportional and non-linear) between screen placements and opportunities to see, which are the only valuable component in a classical CPM measurement. Engagement really has nothing to do with it at this point, since if we were tracking and measuring that, we'd also be tracking and measuring sales lift per ad, which is how digital in-store ought to be measured, but isn't.

Thus, the paradox.

2009-11-19Chauncey writes:

Ah you are in the sales lift per ad camp. That is an interesting place to go as I know a good deal about relative results. Imagine CPM as arbitrage. Arbitrage delivers rewards to those who have more information. CPM actually prices in MORE value to DS advertising operator than does sales lift. The industry needs to be very careful here and be cognizant of what has happened to pure media rates in digital vs. search on the web. If DOOH aspires to a sales lift justification, it will end up killing most of the models getting funding right now.

Completely get the conceptual nature of your post. But engagement is important as an eye-level screen will almost always get better engagement from a viewer (as measured by time of view and recall) than an overhead screen. Lots of research behind that.

2009-11-19Bill Gerba writes:

If DOOH aspires to a sales lift justification, it will end up killing most of the models getting funding right now.

There's a big difference between those models getting funded today and those that will ultimately be successful. Specifically, there are a lot of the former, but very few of the latter.

I'm ok with clearing the wheat from the chaff it it will move things along with big-time advertisers and planners.

2009-11-20John writes:

Good post Bill, never thought about ad space this way.