Article
Tracking digital signage effectiveness
Advanced tools can measure how well content is working.
May 9, 2007 by James Bickers — Editor, Networld Alliance
Editor's note: This story is excerpted from "Measurement and analysis for digital signage," a free how-to guide from Digital Signage Today. To download your free copy, click here.
Retailers, banks, restaurants and any other business that deals with human beings in person, at their location, must always be aware of traffic. Declining foot traffic is a harbinger of things to come — ShopperTrak, which provides technology tools for measuring footfalls, estimates that a steady decline in foot traffic over a given period will result in decreased sales in about 13 months. And being aware of decreases as they happen allows the business to move dynamically, enacting campaigns to bring more people into the space.
Laura Davis-Taylor, president of Retail Media Consulting, said there are three primary types of tools that do the actual measuring of customer activity:
- Simple traffic counters, such as laser beams across the entrance to a store
- Video recognition systems, camera-based technology that counts the number of people that walk by a certain space, if they stopped and for how long, etc.
- Traffic tracking tools, ceiling-based cameras that assign a unique numerical ID to each customer that enters the space, creating a log of that customer's activity in the store (movement, dwell time, patterns). That information can be turned into "heat maps" that are accessible in real time.
While the basic traffic counter is limited to some very rudimentary numbers, the other two options can be extended to provide data on the effectiveness of digital signage. Traffic data from zones within a store featuring screens can be correlated with POS data, to establish relationships. If the tool is sophisticated enough, it can track the effectiveness of individual campaigns on the screens — for instance, did that video promo for Oreos result in an increase in sales? That POS data can also be analyzed alongside play logs, allowing operators to see which digital assets had the greatest effect on sales.
| ||||||||||||||||||
New technology on the horizon
Two new technologies are on the way that promise an even greater focus on specific customer data.
The two largest companies that track consumer exposure to media and marketing, Nielsen Media Research and Arbitron, have both announced services for measuring in-store media, in addition to their traditional radio and television ratings. In December 2006, Nielsen announced its "follow the consumer" service Nielsen In-Store, a joint venture with the In-Store Marketing Institute. The forthcoming service is built around a measurement model called Pioneering Research for an In-Store Metric (P.R.I.S.M.), which aims to allow in-store marketing effectiveness to be compared alongside traditional media.
Arbitron has scheduled roll-out of its Portable People Meter over the next three years. The PPM is a pager-like device that a program participant wears throughout the day. The device, which will take the place of the hand-written diaries that participants usually fill out and mail in, listens for special identification codes within the audio portion of any given media (a radio broadcast, for instance, or a television commercial or in-store promo). At the end of the day, the participant places the device in a recharger/docking station, which sends back to Arbitron complete data on what media that person was exposed to that day, and at what times.
Both companies continue to offer more traditional, low-tech solutions for traffic measurement, like exit interviews and manual counting.
Making a good impression
Arguably the most important metric for companies honing their in-store media is the impression. Unfortunately, not everybody uses this term in the same way.
In the simplest usage, an impression occurs any time a message is displayed. Some companies require a person to view the message in order for it to count as an impression; and some companies will make a distinction between a gross impression (any time a message is displayed and one or more people were within viewing range) and a net impression (any time a person demonstrates recall of a message). Complicating matters further are alternate terms – "ad view," for instance, or "opportunity to see," commonly used in the U.K., both of which are synonymous with "impression."
Other numbers that need to be tracked include reach, or the number of unique people who are exposed to the message; and frequency, the number of times each of those unique people was exposed to the message over a given period of time.
Multiplying reach by frequency creates a figure called the Gross Rating Point, or GRP. "A GRP, as defined by Nielsen Media, is a percentage point of the total audience size," said Bill Gerba, president of WireSpring Technologies. "So, for example, in most Nielsen research, which revolves around television advertising, one GRP equals one percent of the total number of TV-viewing households."
Finally, there are proof-of-play statistics, which are created and managed by the software running the digital signage network. Nurlan Urazbaev of BroadSign International called proof-of-play numbers "the basis for accountability in digital signage" and a crucial component of an in-store marketing program.
"In contrast to broadcast TV, a modern digital signage system should be able to record every instance of an ad displayed on each screen — the player level is not enough," he said. "At the end of a campaign a proof-of-play report compares the number of planned ‘ad plays' — we call them ‘ad repetitions' — with the achieved ones. Having these statistics, it is easy for the ad sales department to show what exactly advertisers paid for, reconcile invoices, etc. Solid proof-of-play also facilitates creating rate cards and campaign planning and budgeting."
The level of specificity provided by a proof-of-play report (also referred to as play log, billing log or performance log) varies from one software package to another. Perhaps the most important distinction to note is whether or not the software measures proof-of-play at the player level or the screen level — in other words, if the message was sent to five screens, but one of them was not functioning correctly, is that counted as five ad plays or four? Being able to measure proof-of-play at the screen level — knowing, in this example, that the screen was off, and only four ad plays should be recorded — is an attractive capability, and one that will be especially important if the screen deployer wants to court advertisers.
