The Digital Signage Insider

A while back, a colleague asked me to vet a business plan for an ad-funded digital signage network to be installed in a major US grocery chain. It had all of the "must haves" for such a network, including a good-sized pilot, plenty of screens in desirable DMAs, and a nod to social networking with some quasi-interactive content. In short, it looked like just about every other big-box or grocery chain network out there. That got me thinking: there have been lots of failures in the grocery space. LOTS. And there's currently a ton of mediocrity. But I have a hard time pointing towards anybody who has truly succeeded in making a great digital signage network for grocers. So when speaking with my colleague about past issues and my concerns about the plan, I laid out an alternate, best-case scenario that I thought would offer the greatest chance for long-term success. We've done these thought experiments with a number of customers over the years, in dozens of different industries. So if you find this sort of analysis useful (or at least entertaining), perhaps I'll make this type of article a regular feature on the blog.

If I were starting a supermarket digital signage network, here's how I'd do it

To begin with, grocery and big-box networks fall into one of two general categories. The first group is internally-funded networks, where the venue owner owns and operates the signage network outright, such as experience, branding and private label networks and private label merchandising networks. The second group is externally-funded networks that are monetized by advertising revenue, where a dedicated company owns (in whole or in part) and operates the network, even though it is physically located in another company's venues. There are numerous examples of both types of these networks, but for the purposes of this exercise I'm going to assume the latter case, since they seem to be more popular with the folks I talk to. In these scenarios, we know that getting venue buy-in is absolutely critical. Thus, the first things that I'd do prior to developing a new network would be to:

1. Get the grocer to pony up most or all of the funds to equip a statistically significant number of stores (at least a dozen, preferably 25 or more) with screens. Instead of blanketing the store, target one specific area for now, preferably one where the retailer has a good number of private label products in place.
   
   
2. Get the grocer to commit to providing data (ideally register receipts, but at the very least info on sales of the advertised products and average basket size) about those stores with the screens, and an equal number of similar stores that don't have the screens for a control group.
   
   
3. Get the grocer to select 5-6 of their private label products that are being sold near the area where the screens will be placed.

I, the network owner, would then:

1. Develop 2-3 different socially-enabled ads or featurettes for each of the products, as well as prepare any secondary content that may be necessary to support the usefulness or attractiveness of the screens. I'd obviously be working from our extensive list of content best practices.
   
   
2. Deploy the content to the screens, and prepare different experiments. For example, run one type of ad on half the screens, and another on the other half during the course of the pilot (ideally 3-6 months).
   
   
3. Gather the grocer-supplied sales data to determine if any product sales lift is evident, or if there was any change in average basket sizes.

Why this approach makes sense

By starting out focusing on just one area of the store, and by using the private-label products that the grocer controls all of the data for (including their profitability), I'd stand the best chance of getting information about performance early enough to make changes (since I certainly wouldn't find the most efficient mix of content on the first try). Also, by starting with private label products (which are much higher margin than name brand products), I'd stand a better chance of enticing the grocer to expand the pilot if the results are good. And finally, since the pilot is deployed in such a small number of stores, I know I would have a very hard time attracting third-party advertisers -- and it would be like pulling teeth trying to get any information back from them (or the grocer) about results. The grocer would have no such concerns in our hypothetical pilot scenario: they're not going to care about DMAs if it's their pilot, since they use their own internal groupings. And because they have skin in the game (in the form of partly or fully capitalizing the cost of equipment), there's a much greater chance that they'll contribute the necessary resources to make the pilot -- and any subsequent expansion -- a success.

Finally, if the results do look good, you can choose how you'd want to expand, either by outfitting the same specific area in the rest of their locations, or by adding more screens to the other areas of the pilot stores and doing a second pilot phase.

For what it's worth, if you or someone you know is planning a grocery or big-box project at the pilot stage, I'd be happy to give you a great deal on software and media players in exchange for getting access to the resulting performance data and being able to write about it here on the blog. (After all, my last article about a grocery store pilot was more than 7 years ago!) I love my research, you know, and seeing how a properly-run supermarket pilot progresses would help me prove (or disprove) a number of theories that I've been pushing around.