For a while now I've been studying how different measurement tactics can be used to determine the value of emerging media, including digital signage. And for the second time in recent memory, a post from one of my favorite blogs helped me see a new way of approaching an old problem. Simply put, I think the digital signage industry is missing what Adaptive Path would call the "Value Metric." It's the way that we take data from our measurement tools and relate it to the ultimate value of the system, such as generating ad revenues or increasing product sales. Sounds confusing, right? Let's take a look at a handy illustration and some examples.

We can connect behavior to outcomes with "Linking Elephants"

Adaptive Path, one of the premier user experience companies out there, feel that this "Linking Elephants" illustration is vital for helping clients determine ROI. In fact, they kept it hidden from the public for years, only making it available in pricey research papers and reports. Recently, they decided to share their findings in the context of how organizations approach the user experience. This worked out nicely, since that pesky block #4 -- the Value Metric -- is exactly what I've been trying to identify and work out for the digital signage industry. It's essentially the measured component that attaches a hard value to the original customer behavior being measured. That's a mouthful, so let's start by investigating how the concept of Linking Elephants applies to traditional media.

Traditional media focuses on audience size

For television, newspapers and magazines, the whole chain is pretty easy to visualize:
  1. Business Problem: How much do you sell your (limited) ad inventory for?

  2. User Behavior: Reading/viewing your medium

  3. Behavior Metric: Number of viewers/readers, based on subscription rates or data from Nielsen, etc.

  4. Value Metric: Demographic makeup of viewers combined with historical pricing and total inventory availability

  5. Financial Metric: CPM, the cost per thousand viewers
A hard cost to deliver a message to a thousand viewers (CPM) is simply based on the number of people consuming the content, as estimated by subscription rates or measurement agencies, and some factor accounting for market forces like supply and demand. The actual value of the ad inventory is thus directly proportional to the measured audience size.

Online media looks at measurable actions

Internet advertisers tried to adopt CPM for their own use in the mid 1990s, but quickly discovered a more accurate -- and valuable -- way to monetize their space. Because they could measure not just who had an ad served to them, but also who responded to that ad by clicking on it, they decided to price their inventory on a per-click basis. It doesn't matter if 10,000 people view your ad: if only 1,000 click on it (indicating engagement), you get billed only for that 1,000. Thus, their chain of metrics might look like this:
  1. Business Problem: How much do you sell your (unlimited) ad inventory for?

  2. User Behavior: Clicking ads on your web pages

  3. Behavior Metric: Ad clickthroughs

  4. Value Metric: Auction-based determination of the value of a contextually relevant "warm lead"

  5. Financial Metric: Cost per click
The rate for each click is much higher than it would be for a mere view, since the click is perceived as being more valuable.

Emerging media is stuck in the middle

Simple enough so far, right? The problem with out-of-home ads is that in terms of engagement and measurement, we're somewhere in between the old stalwarts -- TV, print and radio -- and the new heavyweight, Internet. Consequently, the best way to link measurement and return together is still a topic of much debate. (This is one of the reasons why the value of measuring digital signage remains uncertain, even in the context of behavioral analysis.) The main question is this: Are our customers viewing, engaging, or acting?

So far, I've pinned down what steps #1 and #5 should be in the diagram. Our Business Problem is just like TV and print: "How much do you sell your (limited) inventory for?" Our Financial Metric is ultimately going to be the cost per spot. On most digital signage networks, this would be the price for an individual playback on an individual screen in the network. But the User Behavior, Behavior Metric and Value Metric are all up for grabs. For example, the user behavior might be "glancing at the screen" (whatever a glance is). Or maybe it's "engaging the screen via mobile phone." Or it could be something totally unrelated to the screen, such as picking up a coupon or moving in a different direction inside the store. Because there are so many possibilities, the Behavior Metric is equally difficult to lock down. If a glance is the desired User Behavior, then tracking cameras and software might be the best way to generate a "number of glances", "glances/spot" or "glances/hour" metric. Of course, that information is less useful if the preferred behavior is interacting with the screen some other way.

Many questions remain

We're still left scratching our heads about the Value Metric. Should it have some demographic component? What about the kinds of purchases typically made at the venue? Can it account for "soft value" things like customer experience or trip history and other measures of loyalty?

From my analysis, there is no one single User Behavior nor one Behavior Metric nor one Value Metric. Thus, we probably won't be able to create a single chain of behaviors and metrics like Adaptive Path uses. But we can have a flowchart! In an upcoming article, I plan to chart out all of the common user behaviors that a digital signage network operator might pay attention to. Then, I'll try to create a map of available Behavior Metrics and Value Metrics for each one. To make the list as comprehensive as possible, I'm asking for your help with the following questions:
  1. What's the most common goal for digital signage at-retail??

  2. What kind of "interaction" is the best to measure?

  3. What would that measurement be worth to you?
Let's get a great list of answers together, and hopefully we'll see a natural Value Metric fall out. To join the discussion, leave a comment below.


+1 # Franois Reeves 2008-09-26 13:32
Funny how digital signage installations and networks should be burdened with exact measurement while radio, television and OOH got away with inaccurate reporting for years. Clients and providers are not playing the same tango in DS just yet. It will come. We have skimmed the subject before. Nielsen just released a very thorough measurement method aiming at establishing a concordance between CPM an in-store metrics (fuel pump transactions, cash register records, restaurant confirmed guests, traffic auditing, etc.) For myself, I calculated an Excel matrix for 5 second exposures (multiple thereof, 5 sec being the minimum ad exposure required for acknowledgement) to work the problem backwards. What kind of traffic do I need to reach acceptable CPM costs. You'd be surprised by the results. It turns out that in-store installations provide the worse results at $5 CPM. The best results are achieved for type A & B installations, obviously offering more eyeballs. Middle range results are attained for waiting room installations. Eyeball traffic is your finite matrix driver assuming you have learned all of Bill's lessons on message effectiveness. Sorry for the long post. I'll gladly calculate CPM assessments for your projected installations. Write me an email.
+1 # Franois Reeves 2008-09-26 13:34
CORRECTION: The best results are achieved for building classes A & B installations
0 # Bill Gerba 2008-09-26 14:30
Hi Franois, You make a very important point: while retail advertising networks get the majority of the press and hype in our industry, they're not necessarily the best pick. As you noted, advertising in other venues can produce very good results (see Captivate Networks as an example of how to do it right in buildings), and non-advertising uses of the technology continue to outpace advertising uses in terms of both number of screens and dollars spent.
0 # Francois Reeves 2008-09-26 16:53
Unless I missed something, Captivate Networks is selling impressions. I believe this is the wrong way to go as you can have a thousand impressions with no one in front of your screens! I was referring to the cost of reaching 1,000 people. I think CPM should stand for the cost to reach people not showing an ad one thousand times. If their screens are third party audited I guess it amounts to some sort of equivalence.
0 # Laura Davis-Taylor 2008-09-29 18:08
(1) What's the most common goal for digital signage at-retail?? Kind of hard to answer b/c it depends on the intent of the installation. If ad sales is the driver, it's media revenues. If not, sales tends to be the most common goal (with all tactics ultimately supporting it) followed by the softer, more fuzzy goal of brand affinity. (2) What kind of "interaction" is the best to measure? Maybe a pipe dream, but I'd love to be able to one day measure glance, any kind of interactive touch, physical behavior (path or pulling a coupon, etc.) sales and "post impression" sales (they saw the ad one week and bought the next). You of all people know the barriers to making this happen but I can still dream! :) (3) What would that measurement be worth to you? As a very famous retailer said to be recently, "one qualified view is worth 10,000 unqualified views". Value should be reflective of this. The hard part is that if the retailer is not creating the creative, they can't be responsible for a bad creative execution not pulling response. ..the brand had the chance and should still pay for exposure. But if they have no way of knowing what does and does not definitively pull response, how can they learn the most effective creative approach? Thanks for taking this on's incredibly gracious of you!
0 # Franois Reeves 2008-09-30 09:41
Laura Davis-Taylor says: "but I'd love to be able to one day measure glance" Laura, I think we are near a solution that would do just that. SONY has incorporated in their new video cameras a face recognition technology whereby faces are detected among objects and isolated as white squares (Bill is happy, anonymous) on the screen. The next step would be to combine that with a counting algorithm and voil!
0 # Bill Gerba 2008-10-01 02:59
Laura: great insights, as always. I'll definitely be using that (unattributed) "one qualified view is worth 10,000 unqualified views" quote in the future :) I do wonder about your last sentence, though. After all, retailers are often selling products that they're not directly responsible for, and they've been doing fine with that for centuries. Why would the management of the media designed to sell those products be any different? (not saying that it should or shouldn't be, but I could see the argument going either way). Franois: So you get some hardware/software together that works well enough to give you say 99.9% accuracy (we've tested a half dozen packages, and so far they're all **well** below this mark). What then? Who decides how long a glance should be? There's still no benchmark for identifying an engaged viewer in my opinion.
0 # Roahre Jansen 2008-10-01 03:03
Hi Bill. Hope that you fella's are all well in the States. I thank you for the opportunity to share my thoughts regarding one of the final frontiers that we (digital) face, with much awaited anticipation of the digital revolution wave consuming the modern retail world. The issues that you raise are so very pertinent to the progress of digital in South Africa. As a matter of fact we are currently facing identical barriers of entry. In direct response to your mail here are my thoughts. 1. What's the most common goal for digital signage at - store?? Digital signage has to stand on the foundation of a win ,win, win formula. We have to ensure profitability to the client( Brand owner) through means of futuristic,reliable, interactive digital technology.Because they obtain the financial rewards they are no 1 on the winning podium.The retail store is next in line as far as the win win formula is concerned as they are ultimately responsible for permitting the (motor) to turn. The retailer's rewards are also financial due to the rate of sale and the margins attached to product.We (Digital) are the last to take our place on the win win platform. This is my understanding that we need to ensure that through digital in store advertising the financials that are presented are based upon a proven track record.(For the majority of us involved in digital) This track record is currently still categorized as works in progress.Having said that Advertising works and we as pioneers need to re- MIND our target of that. That track record speaks for us and they are irrefutable.In closure, financial growth is ultimately the first prize that digital has to ensure. 2.What kind of interaction is the best to measure? In my 24 years of experience i Know that money buys the whiskey !!! The best factor that we could incorporate into this cost calculation is to track the spend as advertised through the retailers cash register.That i believe is the conclusive measure to obtain a pulse on the heart of retail and what happens for the brand and it's owners.This as it happens i believe supports both my views and understanding that i have conveyed in your first question. 3.What would that measurement be worth to me? At this current moment we are in pursuit of this answer, as this is one of the barriers that prohibit the market from engaging with all that they have and we all know that those cash bundles are literally endless. This is for us at White light a serious factor that we are monitoring very closely because it will ultimately give us a financial factor that we can proceed with. Due to the fact that digital in store is still in its upstart the answer to this is still out there and if i had that today i would know what a brand owner is prepared to direct financially at digital advertising.
0 # Franois Reeves 2008-10-02 10:00
I think you are being hard on yourselves with the notion of measuring digital signage's effectiveness with its impact on the cash register. A sale might not materialize right away. Long term exposure to your messages might prevent a buyer from going to the competition. etc. There are other aspects that DS can deliver well and is not measurable with dollars and has meaning. Top of mind, customer awareness, brand exposure, customer relations (providing contents and entertainment at the point of sale is a nice customer gesture, a distinguishing factor over the competition). I doubt POP displays (printed) were ever measured with so much strain on the cash register's behaviour. I can see your point in that it provides a final argument to help deploy networks but to reduce DS to only that aspect for measurement is limiting.
0 # june 2008-10-03 18:36
We're not in a retail setting, but rather, hospitality (hotels, welcome centers, restaurant waiting areas). Therefore, we can't use POP sales lift measurement tools. We're now experimenting with on-screen coupons and matching map/coupon print pieces that tie-in. But I'm skeptical that either will deliver the hard data that a gazillion coupon print run will deliver. (And we shouldn't be held responsible for badly done creative that doesn't motivate them to buy.) Billboards and general broadcast TV don't have their feet held to the fire as much, but then, they've been out there longer. I hope that we'll be able to deliver better measure metrics down the road.
0 # deepak jayaram (deejay) 2008-10-06 12:35
good point to debate on bill . . given that this has been a vexed question that has been taking up so much mindspace . . also most new media seem to have the extra burden of having to prove themselves to get a significant level of spends while existing media options need metrics more to determine the level of spends. Personally believe it boils to the level of metrics existent in the marketplace . . the probable answer is agree on the "relevance and role of digital networks" and to subsequently relate the spends to the same. The grid you hv put up could be an interesting starting point . .
0 # Dwight Epp 2008-10-14 18:12
I'm not sure if I'm out of place or not. I have been in the outdoor advertising industry for only five years. I have thought of how to measure the metrics for our billboards and have so far come up empty. I believe that it is the adverisers that can answer the question for us.(testimonials, surveys etc) While some companies do surveys to find out, " how did you here about us", internet, billboard, yellow pages, TV, radio,etc. This may determine how they advertise in the future. Billboard companies, Lamar, Clear Channel etc although hard to measure are a good example that outdoor advertising works. They are currently going digital. Digital is greener than paper and you can change your ad instantly with scheduling. I'm sure you all know this. 1. branding and sales for the advertiser 2. surveys with rewards(testimonials from the surveys) 3. depends on the answers Seems outdated but what else is there. Measuring someones glance doesn't do it PS. advertised political party in Saskatchewan and they are going to write me a testimonial from what they have heard from the public. Thanks Dwight Epp
0 # Bill Gerba 2008-10-27 19:48
All, Thanks for the continued thought and conversation on this topic. I'm very much looking forward to putting together the digital signage version of the Linking Elephants "Real Soon Now" :)

Subscribe to the Digital Signage Insider RSS feed

Looking for more articles and research? Our newest articles can always be found at Digital Signage Insider, but there are hundreds of additional research articles in our historical articles archive.

You may also be interested in M2M Insider: our blog about M2M and the Internet of Things.