The Digital Signage Insider

Our friends at aka.tv recently published an opinion piece by Terry Scannell which added some constructive criticism to the digital signage budgeting article that we posted to the blog last week. Overall, Scannell estimates that our newly revised budget is between $1,800 - $2,600 short of what it costs to deploy a screen for three years. While we could go back and forth on any number of expenses and complications that can arise during the course of a retail media deployment, I thought it might be useful to revisit Scannell's comments and provide some additional detail, feedback, and explanation of how we arrived at our numbers.

Update: Since we write a new pricing article every year, we've created a page that lists all our articles about digital signage budgeting, cost estimates and price guidelines. This is also a handy page to link to if you want to point your friends or clients to our latest pricing data, since we'll be updating the page whenever we add another article in the series.

Player hardware: the importance of high quality parts

Scannell's first objection was that we kept the media player price steady at $1,500. He argues that:

The fact is that the price of player hardware has dropped significantly if you are using a standard PC or Apple player as opposed to a dedicated player like WireSpring. In addition, the power, functionality and small form factor you can get for this price is truly amazing. For that reason, I would reduce the budget allocation for player hardware to $1,000 as opposed to $1,500. So I have a saving here of ($500).

Actually, WireSpring's digital signage software does run on standard PC hardware, from the lowest-end Dell desktop to the highest-end workstations out there. However, that doesn't mean that all PCs are created equal for an always-on application like digital signage. High-quality hard drives, video cards, and other parts make a big difference in the long run, leading to better uptime and fewer on-site service calls. That said, our FireCast Media Appliance is simply an industrial PC with carefully selected and tested parts, and we often give an exact bill of materials to any customer who prefers to build their own PC. However, a number of our clients insist on purchasing a turnkey hardware/software solution, and there are basically two reasons why: they either value the long-term availability of a consistent platform, or they want one company to throttle if something goes wrong. As we noted when talking about the IBM Anyplace Kiosk, it's pretty hard to get a truly consistent hardware platform from the major PC manufacturers (Apple included), since they purchase parts from multiple suppliers and often go through many cycles of motherboards, chips, RAM, etc. in the course of a product's lifetime. However, consistency is absolutely key to managing and troubleshooting a large network, and from our own anecdotal observations, managing a homogeneous network of industrial computers (whether ours or somebody else's) can easily cost a third less than managing a same-sized heterogeneous network of white boxes or mainstream PCs. Similarly, when something does go wrong, many network operators value the short chain-of-command for identifying and resolving the issue, regardless of whether it appears to be caused by hardware or software. As Scannell said, "you have to know what business you're in," and most of our customers (and our blog readers) aren't in the digital signage business. They're in advertising, or marketing, or merchandising or something else altogether, and they want absolutely nothing to do with the nuts-and-bolts of the digital signage platform.

First level tech support: fielding questions from the stores

Scannell's next concern came when we suggested that many of our retail clients were opting to take on first-level tech support duties for the kiosk and signage networks. He says:

This area is where I really start to have a fundamental disagreement with Bill's budgeting. What he is saying is that the clients can pick up first-line network support. This of course saves money - in this case, $3600 over the three years.... Remember, though, that when talking price the next question should always be, "for what?" I have not talked to a lot of CTOs in recent years who feel their staffs do not have enough to do; few of them feel that what their IT organization needs is to run a private broadcast network in their copious free time.

This is an interesting subject, and I was quite surprised myself when first reviewing our own data. However, in the past 3 years -- and in particular over the past 18 months -- the companies who have deployed the biggest networks with us have all opted to handle first level support in-house (after lengthy discussions with us about what exactly that entails, of course). I stand by my original observation that those retailers who see the most value in their networks are the most eager to handle supporting it themselves.

To be absolutely clear, we're talking about first line support tasks, like notifying a particular store if a system misses its regular heartbeats, resetting network gear after a power outage, etc. -- the very sort of tasks that most help desks are already set up to handle for other products in the store, like POS systems and in-store radio channels. Any company deploying a digital media network as a profit center understands that downtime costs real dollars, and a well-run retail operation can address simple first-line problems faster than a 3rd party firm ever could. I'm not suggesting that this is the right model for everyone to follow, but it does certainly illustrate a trend that we've seen, and likewise has proven to be an efficient and cost-effective way of quickly closing out simple technical problems. As a related note, Scannell proposes eliminating the entire line-item for remote management and content distribution software and including this in the tech support fees. I don't really see any benefit in re-classifying it this way, but the tech support discussion is the same no matter which category the cost is placed in.

Installation and project management: planning and executing a smooth rollout

When it comes to installation and project management, Scannell says:

I really disagree with this one. The 2004 and 2006 blogs budget the costs for a one-location project management fee at $300. In fairness, if it is only one location $300 may be right. We also know that clients hate to see a line item for this.

But this is one of the areas where many projects go wrong - way wrong. Let's face it: it is as disruptive having the installers come to your store a day early as it is having them come a day late. Oh, and did we mention the fact that we need to get written landlord approval to put that satellite dish on your roof? And on and on it goes.

I raise the one-time project management budget by $500 to $800. I think that's realistic. Obviously, there are many exceptions, and just because it costs $800 to do one site does not mean you can take $800 and multiply it by 1000 locations.

I agree it may be misleading to look at this on a per-screen basis, since few people would hire a project manager to deploy a single screen. Our intent was for this to apply to the networks we see most often, which are around 50-500 locations with one or more displays in each. Based on data that we have from these networks, the project management cost on average comes out to about $300/screen. Our estimate here is data-driven, primarily from customers in the retail sector. With regard to installation costs, which Scannell subsequently revised downwards to $1,000 (from our estimate of $1,400), this is again a data-driven observation, however one which predominantly reflects installations in retail environments that had high ceilings and required numerous power and network drops. Keep in mind that your installation crew may need different licenses and insurance to run electrical cables, make modifications to ceilings and walls, and even to work at heights above 8', all of which can drive up the cost per day....and a multi-day installation in a large retail environment is not that uncommon.

Some final takeaways

Looking at the numbers, I think Scannell's total budget estimate is probably a safe bet if you're talking about a small network of 1-10 locations. However, looking back over the past two to three years of deployments that we've worked on, I feel that our revised 2006 estimate is reasonably accurate, at least for digital signage networks that span 50 to 500 locations. Deployments on the low and high ends of the curve will skew these figures (especially on the high end, when the network owners seem to delight in squeezing every last drop out of their suppliers), but even projects that fall well within the scope above will be different in different cases based on the needs of the client.

One last note: after numerous emails on the subject, I'd like to reiterate that these are not necessarily WireSpring's prices for the listed hardware, software and service items. We've certainly factored our own pricing into the equation, but we've also included partner and competitor data to simulate a more "average" price point for each item. With nearly 300 companies out there selling kiosk and digital signage systems, it's tricky to say how general our prices are, but we've based our estimates on some of the best known and widely-deployed hardware and software options out there. (Yes, $500 is a reasonable estimate for high-quality player software in all but the smallest quantities, but as Scannell says you'll probably pay more for a single license for a single install.) Thus, aside from the challenge of drawing an apples-to-apples comparison among some of the very different business models in our industry (i.e. the purchased server vs. hosted ASP approach), hopefully the additional info I've provided will give you a good idea of what to expect when it comes time to create your own digital signage budget.