The Science of Signage: A Digital Out-of-Home Primer

May 2nd, 2008 Daniel Parisien

I’ve consulted on hundreds of digital signage deployments and implementations globally over the last 5 years and a specific recurring structure has emerged with regards to technology, content and workflows for managing these networks. The purpose of this article is to identify the basics of a digital out of home network for newcomers who wish to skip the learning curve or for those that wish to focus on the core elements that make a digital signage network a money making success.

The sea of decisions that must be made before spending the capital to deploy a network can easily lead an operator towards focusing on technology instead of focusing on building an effective communication medium.

My team and I (BroadSign Solutions) have had the benefit of repeatedly following the digital signage learning curve vicariously through the hundreds of entrepreneurs that have evaluated BroadSign Suite for its application to their business. This includes many of the members of OVAB and other industry leaders and pioneers. While there are many variables involved and decisions to be made to develop a digital signage network, there are not so many that they can’t be listed in a blog article, such as this. In order to create an effective communication medium, arguably the only real purpose of a digital signage network, a network operator must make all the decisions in the design of a digital signage network from the perspective of the audience. Since the experience of the audience is first and foremost a function of the venue, let’s list some example venues so we can see the variety of digital signage applications and use them as examples:

• Transportation: trains, airplanes, buses, in transit, boarding, unboarding, arrivals, departures, ticketing, lounges, concourse, …
• Retail: supermarket, convenience store, big box, outdoor facing, entrance, checkout, aisles, departments, freezer, deli, …
• Healthcare: hospitals, doctors, dentists, veterinary, surgery, waiting areas, in rooms, …
• Hospitality: hotels, resorts, casino, outdoor facing, lobby, elevators, in rooms, …
• Food/Drink: food courts, quick-serve, restaurants, coffee, bars, clubs, buffet, restrooms, …
• Outdoor: pedestrian, automotive, train line periphery, downtown, highway, tourism destination
• Active: ski, golf, gym, skate park, chair lift, pro-shop, cart, lobby, change rooms, …

The above list is by no means comprehensive, so it is easy to see that digital signage has virtually unlimited amounts of applications. So if I am going to deliver on my promise in the intro of this article, I must answer the question:

What is the common framework for all the different applications of digital signage?

We can start with the workflows and processes involved in the growth and operation of a digital signage network, since this is universal to all of applications:

1. Deployment / Operations

a) Planning and Selection

Before a deployment project can begin execution, one must select the technology vendors and service providers. This usually involves standardizing on a software platform, a small set of playback devices models, display technologies and to select a team of integrators to perform on-site installation and maintenance. One good practice at this stage is to create an operating system image that is pre-configured with the playback software and system drivers so all player PCs can be treated as identical interchangeable clones, i.e. a toaster.

b) Survey and Profiling

A site survey is performed to identify the tasks required for permanent installation of the display on location. For example, locating parts of the venue’s structure which can withstand the weight of the display equipment and the location where the player PCs will be installed and the whole wired together and connected to a network. It is also at this stage that the display units that represent the channels at that location can be built up in the software and categorized according to geographic location, demographic or other criteria for later use in targeting advertisements. Typically, the catalog of criteria used for targeting/grouping is defined by marketing for usage by the ad sales team.

c) Installation and Commissioning

If the site survey is completed and comprehensively planned, installation should simply be step-by-step execution by certified professionals. One or more players are installed at the location and they are plugged into a power source and a network source. The display technology is installed and is connected to the player. Approximately 50% of BroadSign powered players are collocated with the display while an equal amount are stored in a dedicated location and wired to the displays via a video distribution network such as video-over-cat5. If the network supports audio, the same statistics apply to the sound system. Activating a newly commissioned site is often the easiest aspect of the deployment process which is associating the player PC system to the display unit profile created in the earlier step.

d) Monitoring and Troubleshooting

Once a new site is commissioned, all the players installed at that location are added to the monitoring and troubleshooting pool. In the field, several revenue-compromising issues could arise that need to be dealt with as quickly as possible. There are 3 main classes of issues:

1. Content issues: content is not downloaded yet, content errors such as missing codecs, content play time is truncated, etc.
2. Display issues: display is not powered on, display is not on the correct input, display volume is not functional, etc.
3. Player issues: player is missing in action, player is having difficulties polling, player has restarted unexpectedly, etc.

While most issues are either solved automatically by the software, e.g. content downloads, or for notification purposes only, e.g. player restarted unexpectedly, some may require you to send personnel on-site to replace units. One example of this is if the notification is the result of irreparable hardware failure. Since, from a remote monitoring system, it is impossible to distinguish connectivity issues from hardware failure (since both mean the player is no longer polling the server), programming monitoring thresholds is integral to proper issue escalation so you don’t spend an inordinate amount of money dispatching field technicians. Of course, another part of protecting against these costs is to select extremely durable hardware.

2. Media Placement / Advertising

a) Loop Strategy

A network’s content strategy is primarily driven by the nature of the audience at the venue. The first part of a content strategy is to determine the loop length, which should be equal to the average audience dwell time. The logic behind this is that each slot in the loop has an equal opportunity to be seen by the average viewer. In some cases, the loop length is based on other criteria, but this is typically combined with a default loop saturation (how many times the message appears in the loop) that achieves the same objective, e.g. a 6 minute dwell time location may have a 1 hour loop, but ads scheduled to play 10 times an hour by default, which means they play every 6 minutes.

The next step is to determine how much non-advertising content is necessary in the loop, where the rest of the loop is reserved for advertising (either promotional or for sale to third party advertisers). This non-advertising content should be beneficial to the audience and should serve the purpose of the real-estate owner, which may or may not be the network owner. For example, an office tower network may wish to have business news and a healthcare network may run health related content. Acquiring and licensing this content should be factored in as part of a network’s operational costs and is integral to its marketing strategy.

b) Ad Inventory and Rate Card

Typically, a media company will price its ad space by calculating its costs, adding a profit margin and then distributing that value to its prime day parts and locations. This means that, as long as that primetime is sold, the company will remain profitable. Assuming that primetime is desirable by advertisers and easy to sell, this also means that any remaining time that is sold is also pure profit. The next stage is to create a rate card for the network in a currency that allows advertisers to compare it to other media, for example a CPM value.

The rate card is intimately tied to the content strategy and campaign execution capabilities of the software platform. The rate card is a catalog of media packages chosen by marketing and sales to represent typical media buys on their network. As media packages are used to book campaigns, they carve out of the available ad inventor. Sales management is expected to monitor ad inventory, or avails, in order to focus the sales team on undersold venues and to resolve overbooking conflicts. Typically, a media package has the following components:

• Pricing, e.g. a CPM value
• Targeting criteria, e.g. a specific demographic or market
• Flight duration, e.g. 4 weeks
• Slot duration and frequency of appearance in the loop, e.g. 4 slots of 15 seconds each
• Other premium options, e.g. sponsorship, adjacency rules, exclusivity

c) Content Management and Campaign Execution

Digital signage is a communication medium and is one of the best media platforms for targeted campaign execution. The campaign execution workflow has 4 main components:

1. Campaign: this is the stage at which you reserve slots in a series of targeted display units, for a defined flight period, day part and frequency in the loop. Refer to the earlier section on rate card and ad inventory for more details.
2. Content: once content is created, it is uploaded into the system and undergoes quality control and approvals. It is at this stage that the content is ensured to playback properly (required formats) and will be of acceptable quality for the display (required dimensions)
3. Schedules: content is tied to a reservation via playback rules, such as play times (time range, date range, week days), e.g. the same campaign can have a schedule for the “coming soon” creative of a movie trailer and a different schedule for a “now playing” version.
4. Categories: another class of playback rule that helps define the content’s position in the loop. Categories can be used for separation, segmentation, preemption or synchronization, e.g. separate Coke ads from other ads in the beverage category or set a sunblock ad to synchronize with the beginning of the weather segment.

d) Effectiveness Measurement

In order to determine CPM, first one must conduct some formal research by a respected third party media research company, such as Arbitron. They perform compliance audits, which measures the effectiveness of content delivery to the display, and they also perform viewership analysis, which measures the effectiveness of audience delivery after the message leaves the display (and sound system, if applicable). These and their other services will help a media company’s sales and marketing efforts, for example to describe a media package as a compelling and measurable audience.

In an advertising sales environment, campaign execution is preceded by a proposal and contract negotiations. During this stage, an insertion order is generated which contains the price breakdown of each campaign in the media buy and the all-important planned number of impressions and ad repetitions that they will generate. During and after campaign execution, an affidavit report is generated which compares the original planned amounts with the actual amounts as measured by the software.

In a non-advertising network, effectiveness measurement is more directly measurement of sales uplift, which may involve correlating ad play times with sales records from the point of sale.

Effectiveness measurement (ad delivery, audience delivery) is key to getting on the media plan, but it is even more important for the media research to yield outstanding results; this is directly a product of great design decisions.

Let’s chunk down design decisions into 3 categories:

1. Audience Experience: the content and audio-video presentation must be built to harmonize with the reason the audience is at the venue, e.g. end-cap signage in retail brings awareness to the product and allows the consumer to dig deeper and find out more about the product if this is desired.
2. On-site Infrastructure: mounting, wiring and physical protection are a function of the venue’s construction details, e.g. in a train car, signage must be physically secured from vandalism and must be connected wirelessly to its network update source.
3. Software Platform: the capabilities of the selected software platform drive the integrity of the design and need to meet all operational requirements, e.g. if a software platform cannot support triggered content playback, then you cannot execute on a design that includes interactive signage.

Each layer or peel of the onion, so to speak, puts restrictions on the layer above it. For example, if the on-site infrastructure prohibits installing a display at eye level to the audience, this will filter down to the audience experience in terms of visibility. This being said, as mentioned at the beginning of this article, an effective communication medium is designed from the audience’s experience. While I will list details about on-site infrastructure and network architecture, my primary focus will be on answering the following question:

What properties of a digital signage application are part of the audience’s experience?

In short, the audience’s experience is primarily a result of the venue and a great digital signage application capitalizes on providing an easy to consume message that is relevant to the audience. Easy to consume and relevant means that it is easy to see (and hear, if this is applicable), that the message can be viewed in a measurable and predictable fashion, that it can be viewed without endangering an audience member (e.g. taking their eyes off where they are going), and that it does not disrupt them from achieving their objective at the venue. This is important enough that I wish to analyze it in a little more detail from 3 perspectives:

1. The Relevance of the Message :

It is imperative to remember that signage in an out-of-home network is secondary to the purpose of the audience at the venue. An easy example is that a consumer in a retail network is there to purchase products, so information about those products near where they can be added to a shopping cart. For an advertiser, being able to target their message to locations and times where it is contextually relevant is not only good for sales, but is also good for the brand since messages that are contextual are considered helpful and welcome.

Another aspect of message relevance is related to the non-advertising content, such as news updates. For example, elevators in business towers providing quick news headlines (business, local, national, world, sports, entertainment, weather) for the tenants of the buildings as well as visitors to those tenants is helpful and relevant. News content with audio works best in high dwell time venues such as transit applications such as trains or healthcare waiting rooms whereas RSS headlines are more appropriate in less captive environments.

2. The Content and Loop Duration :

The duration of the content and the loop determines how much of the message is received by the audience. For example, if the message is too long compared to how long the audience can look at the display, only part of the message is received.

Duration of Message: the more the audience is moving in front of the display, the less attention they can give to a display which gives you a shorter time to transmit your message. For example, a concourse application or a retail application in the aisles should design content to get the message across in 1 second. Due to this, even if the message displays for 30 seconds, as long as the entire message is persistent on the display for this period, the message will be delivered. Conversely, the longer they are immobile, the longer the content should be. For example, a message that persists on the display for 30 seconds will be boring to a captive viewer such as a sports arena or airport departures. In this case, a programming style similar to television becomes most effective.

Duration of Loop : as a rule of thumb, making the duration of the loop equal to the average viewer’s dwell time at the venue with line of sight to the display, then each slot in the loop has an opportunity to be seen by the average viewer. Typically, if the loop duration is much longer than the dwell time, then the media company will package several slots for a media buyer in a way that the frequency of ad plays equals the dwell time. For example, a retail location with a 6 minute dwell time may have a 1 hour loop, but all the advertisements run 10 times per hour, which means it plays during the average dwell time.

3. Visibility/Audibility :

The visibility of the display and, if applicable, audibility of the sound system determines how clearly the message is received by the audience.

Display: this is a function of display placement in accordance with the audience’s line of sight, the size of the display compared to the proximity of the viewer, and the brightness of the display contrasted to the ambient brightness of the venue.

• Although line of sight for a typical person at a venue greatly increases the possibility that content is seen, many venues have restrictions on installation that make this difficult to achieve.
• Being able to discern elements of an advertisement depends on the size of these elements and how close the audience is. For example, retail applications at the end-caps can have smaller writing since the audience is closer to read it, displays that must be further from the audience must be proportionately bigger so text can be read, etc.
• Plasma and LCD displays are the most prevalent solution as they represent over 80% of BroadSign Powered installations. These are mostly only applicable in an indoor setting, though outdoor casings are possible.
• LED boards are typicaly extremely bright and big which is good for outdoor applications. They are so bright that at night, the brightness must be tuned down. LED is used in less than 10% of BroadSign Powered installations and they are all in an outdoor environment, though indoor LED applications are also common such as in a sports arena.
• Projectors are typically not as bright as other signage and compete with ambient lighting which makes it difficult in a retail environment and impossible in an outdoor application. Projectors are used by less than 10% of BroadSign Powered installations.

Audibility: while some networks do not use audio due to restrictions at the venue, those that do must ensure the audio aspect of the program can be understood. Sometimes audio is secondary to the message such as background audio or radio, but often times it contains a large amount of the message. If the content relies on audio being understood, this is only possible in the venues that have low ambient noise (e.g. healthcare) or if the sound system can be extremely loud, e.g. directional audio in retail.

The above properties of a digital signage application are directly experienced by the audience, but what about those properties behind the scenes that have a direct affect on those properties. In other words, we must answer the question:

What are the types of requirements at a venue that can restrict the design of a digital signage application?

There are some venues that are more amenable to the introduction of an effective digital signage network. This is because the on-site infrastructure can limit what solutions are available. The largest challenge of digital signage design is to find a harmony between these limitations and the selection of digital signage application components that maximize the audience’s experience.

There are seven aspects of a permanent digital signage installation that will affect its design, i.e. which components can be selected:

1. Security: in an environment where the display or other aspects of the digital signage application are publicly accessible, all components must be locked down. For example, displays that are housed in subway trains must be encased in vandalism-proof glass. Another example is that flight information displays must be protected against people unplugging cables or hitting the power button. In some cases, professional grade displays have their external buttons disabled to prevent tampering with display settings. The larger the daily audience, the more the signage is at stake. The less surveillance is possible, the higher the chance for vandalism.

2. Regulations: depending on the jurisdiction of the venue, regulations must be respected which often means some certifications must be obtained before installation can commence. One example is in public transportation networks where signage is part of an emergency broadcast mechanism. In this scenarios, the display enclosure must be able to withstand the full blast of a fire hose. Another example is signage in a medical setting where wireless signals cannot be used and there is a threshold on how much electromagnetic radiation electronics can emit.

3. Cabling: the based requirement of a digital signage network is to have a display and a player PC connected together via some sort of video cable and the two must be connected to a power source. Additionally, in some circumstances, the display and PC are also connected together via an audio cable (50% of BroadSign clients) and/or an RS-232 cable (over 80%) for remote device control and monitoring. In a situation where a venue has a back-office server room, the PC and display may be connected via a AVC (audio-video-control) distribution network such as a video-over-cat5 installation. In situations where the display is co-located with the PC, cabling is less of a concern, except oftentimes power and Internet cabling become the biggest challenge. Depending on the situation, you may need to use different types of video cabling, e.g. VGA (70% of BroadSign clients), VGA-Over-Cat5 (20%), DVI (10%), RF wireless (0%), etc. It is important to validate the video quality is acceptable from those technologies in a real-world test scenario before mass-deploying as this could be a lingering headache for the lifetime of the network.

4. Connectivity: in most digital signage applications, an Internet connectivity source is necessary in order to distribute content to the PC and to collect health and playback statistics. BroadSign clients are roughly using 85% broadband of which 4/5 are wired broadband and 1/5 is 3G cellular. Of the leftover 15%, 3/5 are using 2-way satellite with multicast, 1/5 are using 802.11 wireless and 1/5 are using dialup. The method of connectivity and the network topology has a large effect on what type of content can be scheduled and how often it can be updated.

5. Power: both the PC and the display must be connected to an electrical power source. In some cases, such as transportation applications, the vehicle’s DC power must be adapted to equipment which is tuned to AC power outlets (less than 5% of BroadSign clients). In other more extreme cases, displays must be powered by limited battery supply (0%). Power is nevertheless required in order for it all to work.

6. Temperature: temperature and heat dissipation are important when planning a digital signage application. Heat is the leading cause of hardware failure in this industry and proper, real-world endurance testing could prevent a lot of expensive maintenance down the road. The flip-side of the heat problem comes into play in outdoor networks where the equipment may be subjected to a wide variance in temperature. All digital signage equipment comes with documented operating temperatures. One very cold or very hot day could mean you need to replace all your equipment. In this case, a controlled temperature housing could be required to protect your capital investment.

7. Construction: the construction of the venue may prohibit the permanent installation of an eye-level display. Displays and mounts can be heavy and require reinforced structures at the venue to support them. Regulations may come into play and one has to think of withstanding natural disasters such as earthquakes and emergency situations such as fires. Some venues are just more amenable to signage that can meet the needs of the audience, e.g. a train car specifically fitted for signage, and some are much more challenging, e.g. product shelves in the aisles have no power sources for displays. All these challenges can be solved with construction, but not all construction projects will be welcome by the venue owner.

Hopefully restrictions at the venue have not restricted the selection of components to the point where the experience of the audience is compromised. The final set of design decisions are primarily driven by the software platform and network connectivity method, which begs the question:

What effect can the selection of a software platform have on the effectiveness of my network and what restrictions can it place on my business model?

It may seem overly biased for the head of consulting at a digital signage software development organization to answer this question. At the same time, who could be better qualified to synthesize industry demand and answer this question? Hopefully, the following seven aspects of a digital signage software platform will speak for themselves and will not seem like a commercial for BroadSign Suite:

1. Workflow Scalability: the two major workflows are deployment/operations and media placement/advertising. In terms of workflow scalability, you will want to look at what it takes to configure the software system on the PC once it comes out of the manufacturer. Is it compatible with mass cloning and what is the process for activating and licensing the software? You will also want to look at how easy it is to build a loop of independently targeted ad campaigns, i.e. a mix of local, regional, national and international messaging including editorial content and advertisements. One final area where workflow scalability is often compromised is in the area of proof of display reporting. Are summary reports readily available or must they be constructed using custom database work each time one is needed?

2. Supported Content Formats: some of the creative received by agencies and content houses are not in a format that is suitable for digital signage, e.g. analog tape, digibeta, Panasonic DV25 or uncompressed video. These must be converted to high quality-to-compression ratio formats such as MPEG-4 or WMV. Sometimes compatibility with MPEG-2 is necessary in order to support legacy content libraries. Typical industry standard formats are MPEG-2 (TS or PS), MPEG-4 (DivX, H.264), WMV (VC-1, WMA), MOV (Sorenson, H.264) and FLV (Sorenson and VP7). If audio is used, it is often either MP3, M1L2 or AAC. Other important content formats are Adobe Flash (SWF) and Web content (HTML, Javascript, etc.) permit richer applications such as interactive content, automated templates, third party database integration and more.
3. Software Security: you will want to look for a software solution that encrypts all control-level data and protects against tampering techniques such as data injection, man-in-the-middle. Typically, this is done by using SSL as the transport for this data. Another aspect of security is that the playback software should not be listening on any ports which could make it vulnerable to remote control via an API or through exploits such as buffer overflows. Finally, some level of content integrity should be verified before playback in case content was intercepted on its way to the played. Typically this is done by comparing a cryptographic fingerprint calculated on the player with one provided by the server over the secure SSL connection.

4. Advertiser Accountability: playback reporting is crucial for the invoicing stage of the media buy post-execution. There are many layers of accountability: proof of play on the PC, proof of display on the screen, proof of click for interactive content. Typically, these are provided in customizable reports which are easy for a media buyer to understand, but can become more and more detailed and technical if comfort is not yet established between the network operator and the media buyer. For example, an affidavit report may display performance details of a campaign as a line item, but each campaign may also be described in a multi-page report which has per-venue or per-day breakdowns of playback measurement.
5. Campaign Capabilities: different campaigns have different objectives:

• Broadcast: many clips are grouped into segments and broadcast across all locations or to a class of display. This is typically done for editorial content or automated content such as RSS templates.
• Targeted: ads are targeted to a group of venues based on geographical proximity, demographics, aspect ratio of the display, language of the audience, legal restrictions, competitive restrictions, day parts, week parts and combinations of all of these.
• Interactive: microsites (aka kiosk applications) are broken down into a series of components where clicking certain links triggers their playback. This allows one to track the usage of different components of the application.
• Sponsorship/Adjacency: have one clip immediately follow another for sponsorship or co-marketing reason.
• Saturation/Separation: have slots of the same category be spaced apart as far as possible in the loop. This may be to space out competitor or to space out ads from the same content provider.
• Synchronization/Triggered: have slots from one campaign trigger the playback of another campaign in another zone or on another set of displays. This trigger can also come from third party technology such as a GPS geo-location device.
• Multi-Schedule: have one slot play a different message depending on the time of the day, day of the week or calendar date. For example, a movie trailer may have a “coming soon” version that plays until the day before the premiere and a “now playing” version at the day of the premiere and afterwards. Another example is the breakfast menu in the morning and the regular menu after 11 am. Finally, opening hours may change depending on the day of the week.
• Customizable: a default message may be customizable by the manager at the local venue, e.g. custom pricing or templated message.
• Preemptable: a message my fill up the loop, but is removed from the loop if its slot is needed by a higher priority message. For example, filler content can be used to make all loop the same length, but higher popularity areas will have less filler in their loop as it has been preempted by paying advertisements.
• Emergency: a message needs to be pulled immediately or some form of emergency message, e.g. amber alert, needs to take over the display.

6. Self-Healing Capabilities: due to the high cost of sending technicians on-site to troubleshoot problems, it is important for the software platform to self-heal as much as possible in the face of problems that compromise ad sales revenue. If it is possible to do so, it may also be helpful to have remote access to the operating system of the player PC, though this is often restricted due to stringent IT requirements. Examples of situations where you want software to self-heal are: deadlocks, memory issues, operating system crash, player crash, disk space issues, codec errors, display issues, and network issues such as high packet loss, high latency, late downloads or data corruption,

7. Compatibility and Extensibility: new and interesting technologies are constantly coming to market in this space, so it is important for your software platform to be able to work with these technologies in order to not be left behind when the next killer app enters the fray. Examples of new technologies include audience measurement cameras and mobile interactive applications. Extensibility is also important in order to develop a competitive edge. This can be done via content with software development platforms such as Flash and HTML, or via player control features such as halt/resume or triggers. Another important aspect of compatibility is the emergence of technical standards in the industry.

Why is audience-oriented design so important?

The networks that make the wrong selections will not last as the networks that are designed as effective communication mediums quickly take promenence in the market and thusly take a larger piece of the ad pie. If an entrepreneurial network operator has an exit strategy that involves being acquired by a larger media company, having a solid measurable medium that is reliable and is also easily consumed by its target audience will mean a much higher valuation.

Hopefully, this text has been helpful to you. If you made it this far and have questions or comments, do not hesitate to add a comment here or contact me directly daniel.parisien(a)broadsign.

Entry Filed under: How to: Digital Signage Tips, The Big Picture, Digital Signage Evolution, Uncategorized