What You’re Getting Wrong About Digital Signage – And What To Do About It
January 14, 2019 by guest author, sixteenninewpadmin
Guest Post: Geoff Bessin, Intuiface
You’ve been careful. You’ve studied all the angles. Thought about your audience, quality content, the project’s goals, hardware. You’ve given your digital signage project everything you have. It’s now deployed, left on its own like a child leaving his/her nest to face the world.
Ask yourself this. What did you get wrong? You can’t expect to be 100% correct about everything. That would be news-making. “Company X deploys digital signage and gets nothing wrong!” So what did you get wrong?
Here’s a scary thought: you don’t know! What are you doing to assess effectiveness? Put simply, what are you measuring? And further, what do those measures tell you? Without data there can be no insight. And without insight, you’ll have no idea what to do to improve the success of your signage deployment.
That’s signage with the lights off.
Here’s where we need to cut you some slack. What’s the old saw in psychotherapy? It’s not your fault….
Classic digital signage are emitters, broadcasting information to a faceless audience. They’re passive, which means there are no direct measure. How can you determine the level of effectiveness, the level of success for a deployment, when there is nothing that can be directly measured? Worse, secondary measures suffer from the correlation/causation conundrum.
Luckily, advances in technology have made it possible for signage to pull in information as well as, if not better than, its ability to push information. Signage is becoming a sensor, thanks to the introduction of human-machine and machine-machine interaction. Interaction leads to insight as it gives you direct, first-hand correlation between signage content and its appeal and effectiveness.
Interaction could be active, meaning the result of a conscious action on the part of your target audience (touch, speech, RFID tags, etc.). There are also more passive types of interactivity where anonymity can be preserved while information is still gathered about the user, like facial analytics and eye tracking. Then there are machine-to-machine interactions that bring context to the interaction (e.g. weather), commonly mediated by Web services or connected objects
Building an Analytics Competency
If signage is a sensor then there’s a lot of data out there. The trick is getting your hands on good, actionable data. Here are the steps you need to understand and embrace to ensure your ability to collect actionable data that leads to real, meaningful insight.
1. Identify Business Objectives
Start by asking yourself (your team), what does success mean for this signage project. What is/are the objectives?
Analytics aside, you should be doing this anyway. When you create a specific objective, you give your team a greater chance of achieving that objective because they know precisely what they’re working towards. And further, it justifies the expense in time, cost, and personnel, finite resources that you could apportion elsewhere.
What makes a good objective? They should be S.M.A.R.T.
- Specific: No room for misinterpretation. Think of the five w’s (who, what, when, where, and why).
- Measurable: Clear measures of progress.
- Assignable: No ambiguity about who owns what.
- Realistic: Achievable with available resources.
- Time-bound: Commitment to a timeline for achieving results
Objectives tend to require the same level of iteration one would expect from project delivery. They often start either too ambiguous or, upon reflection, insufficiently relevant or meaningful. In addition, different members of a project team might have different perceptions of objective priority. Can they coexist or are there trade-offs?
2. Define KPIs
Key Performance Indicators – KPIs – are measures used to evaluate the progress and success of a project. Further, they’re not just any measure, they’re actionable. They shine a light on which objectives are being met, which are not, and what you could do to improve the outcomes of both. They get you through that swamp of data we wrote about earlier, focusing on the information that matters.
The hardest thing you will have to do is identify the right KPIs for your signage deployment. Just because you can measure something doesn’t mean you should. There are lots of measures that don’t, by themselves, tell you anything about progress towards an objective. Worse, other measures are clearly linked to objective success but they just don’t give you actionable insight
Sit down with your well-iterated objectives and imagine what the ideal KPIs would be. To help you think things through, there are three broad classes of KPIs to consider:
- Design: How well is the project design supporting delivery of a given objective? Is there anything about the content, layout, sequencing, location, etc. that can better lead toward success?
- Operational: How well is the underlying hardware and software supporting delivery of a given objective?
- Business: What does management need to know about the project to be sure the objectives are being fulfilled?
Let’s say you’re creating a POS kiosk for a movie theater. An obvious objective would be Sales per Kiosk. (Pick a dollar amount that makes sense for you.)
- Design KPIs
- Average dwell time (i.e. amount of team each visitor spends at the kiosk)
- Conversion rate (i.e. when the visitor actually buys a ticket)
- Conversion rate correlated with gender and age range
- Frequency of request for help for each screen
- Operational KPIs
- Total downtime
- Average response time
- Business KPIs
- Revenue per kiosk (correlated with the weather at that location)
- Ratio of new sales vs. pre-paid pick-up
- Percentage of overall ticket booth sales
- Which movie generated the most sales and on which kiosks
3. Collect Data
You’ve identified your KPIs so now it’s time to collect the data. To do so, you somehow need to instrument the software running on the kiosks to report the required data when relevant events occur.
It’s akin to Web analytics but harder to do because there is no universal kiosk software. Google, for example, can make many assumptions about what is happening under the covers when visitors are browsing a website. As a result, it is easy for them to identify and collect a host of information about those visits.
With kiosk software, you’ll have to be proactive and identify products enabling instrumentation. Ideally, your search for a signage CMS option would treat interactivity and analytics as first-class requirements. Doing so will ensure your access to the greatest level of flexibility and depth for data collection, information about target audience preferences and the surrounding context that may influence those preferences.
How instrumentation occurs is up to the software platform. Our recommendation is to bias yourself towards options that don’t require any understanding of the underlying technology. Creatives should be as capable of instrumentation via the chosen platform as a more technical member of the team.
4. Visualize Data
Now you need to turn the collected data into insight. Typically, this is what one thinks of when hearing the word “analytics” – the analysis of the data. It’s visualization using the charts and dashboards so a story can be told and next steps can be identified.
Yes, you’ll be using bar/line/pie charts and value indicators – and many more visualizations – but the choice can be fluid. What’s important is to understand that there are four types of analytics.
- Descriptive: What happened. Either a snapshot or a trend, it looks backward and thus doesn’t answer the question Why? Most data falls into this category.
- For example, average daily sales per movie theater kiosk
- Diagnostic: Why it happened. Could be assisted by secondary, non-KPI measures
- For example, sales per movie theater kiosk in the context of weather, location in the theater, time of day, city, day of the week.
- Predictive: What could happen with the status quo. It’s a forecast.
- For example, expected monthly sales
- Prescriptive: What could change the status quo for the better? It’s essentially automation of the manual labor applied for Diagnostic analytics. Often involves machine learning and AI.
As the era of big data is upon us, there is no shortage of data analysis software in the marketplace. You can use everything from Excel to IBM Watson. Your decision will be primarily influenced by the level of analytics skill your team already possesses. If you’re like most people, it is preferable to start with simple tools oriented to novices. The cost will be lower and you’ll have time to build skill.
5. Act on Insight
You’ve identified objectives, defined KPIs, collected data, and visualized that data. If you’ve done it right, you should be armed with knowledge outlining the steps you need to take to either accelerate or correct project progress.
Maybe you need to revisit how the target audience is encouraged to interact. Maybe the screens need to be moved to a new location. Maybe the screens are too small to be noticed. Maybe certain content should be removed, other content emphasized. Maybe new types of interactivity would be more effective. Or maybe the project needs to be put on hold and reassessed because it’s just not delivering.
As noted, few projects are a roaring success out of the gate. Changes will surely be needed required, both minor and major. Your stakeholders know this and simply need the assurance that your team is in a position to succeed. They will appreciate the clarity of your reports and will rest easy knowing the clarity of your purpose.
So iterate, as you have throughout the process, making changes to your signage and measuring the outcomes. Lessons learned will not only improve the likelihood of success, they will also influence future projects, increasing the speed with which you can deploy, monitor, and improve them. It may even encourage the adoption of projects with higher technical complexity, projects that would have once seemed intimidating.
Conclusion
Analytics has always been important. It is no more critical today than it was in the past. The problem was with the ease with which analytics could be applied to digital signage. Frankly, it was a huge pain unless you had a huge budget.
This is no longer the case. The most innovative of digital signage CMSs are treating data collection and analysis as first class concerns. Further, adoption of these capabilities has been greatly simplified. It’s a good thing, as most of us have little exposure to the world of big data.
Now go start measuring! Your signage will thank you.
Leave a comment