This article was originally published on Econsultancy, on 31st July 2019.
In the last few months, I’ve noticed a trend emerging at the analytics meetups and events I attend. People are talking more and more about digital dashboards. I’ve even heard people advocate throwing out monthly reports, as they hail the arrival of democratized data in the form of swish new dashboards.
Don’t get me wrong, I love digital dashboards. And in this article I’ll talk a little bit more about when dashboards are great. But as useful as reporting dashboards can be, they’re also a tool that’s commonly misunderstood and misrepresented. Like any power tool, used incorrectly or irresponsibly, they can be dangerous.
What even are ‘digital dashboards’?
Just to make sure we’re all on the same page, when I talk about digital dashboards, I’m referring to a reporting dashboard which pulls data from multiple sources in real time, letting users view live updates on whichever metrics or data sources they’ve chosen to represent in the dashboard. These dashboards are typically interactive – you can hover over data to get more info, adjust the date range, and so on.
For example, here’s a dashboard created through Google Data Studio, which pulls data from Google Merchandise Store’s Google Analytics Account, to show ecommerce performance.
Dashboards have helped me and my team at Three Whiskey in a number of ways. They let us provide ad hoc reports in an instant. We can customize dashboards to let different stakeholders see the data points that matter to them. They’re efficient because – used correctly – they save time and let us aggregate data from multiple sources. We can adapt metrics using a WYSIWYG interface whenever we need to. And of course, we can always download the data in Excel for additional analysis.
But reaping all these benefits takes investment and expertise. Having a dashboard is not a replacement for having a data analyst. And that’s where I am worried many marketers are going wrong.
The rise of the dashboard
There are two factors that have driven the recent meteoric rise of digital dashboards.
Firstly, the concept of data democratization; wherein all data is available to everyone within an organization. In a data democracy, there are no gatekeepers, and everyone has the tools to understand and act on what the data is telling them. This idea and its potential benefits are now well established. And the technology to power it is evolving fast – complex machine analysis and processing of big data is giving users the power to draw insights automatically, in ways that were never possible before.
The second factor driving the rise of the dashboard is Google Data Studio. In 2016, the beta of this new Google product arrived. And just like that, a low cost and user friendly solution for digital dashboards landed at the feet of every marketer.
Data Studio has automatic integration with Google Ads, Google Analytics, Search Console, and all the rest. It uses Universal Data Formats (like JSON and CSV) over proprietary ones to allow easier integration, and it’s compatible with third party APIs. So whichever platforms you’re using to collect and manage data, you can probably plug them into Data Studio.
Sounds great… so what’s the problem?
I want to highlight six potential pitfalls of the digital dashboard – things you’ve got to be thinking about to make sure your dashboards are the business asset you want them to be.
1. The nature of digital dashboards is commonly misunderstood
Data Studio is to dashboards what Google Tag Manager is to analytics tracking. That is, marketers assume it is primarily there to save them time and make their lives easier. And while, yes, there are time saving aspects to these tools, they can only unlock a return for organisations if they’re backed by investment and expertise.
What marketers think about Tag Manager:
“I never have to speak to my development team again.”
What should actually happen:
“I need to work with my developers closer than ever to utilize the power of GTM and GA.”
What marketers think about Data Studio:
“I never have to use spreadsheets ever again.”
What should actually happen:
“I need to make sure I understand exactly what data I’m feeding into the dashboard, and what it means, in order to make decisions based on what I see there.”
2. The benefits are often overstated
If you have a look around for articles on the benefits of digital dashboards, most of them are produced by data platforms which have a dashboard product.
Agility, flexibility, efficiency, beauty…. These are just some of the frankly meaningless words you’re guaranteed to find in those articles. Even with a really well designed and accurately set up dashboard, you still need to invest time in data analysis, despite the vendors of dashboard products telling you otherwise.
3. Who is monitoring the data quality?
Is test data polluting live data? Are media campaign tags working correctly? Are UTM parameters being used correctly on campaign launches? If you’re setting up dashboards, these are just some of the questions you need to answer to be sure you’re getting clean and accurate data.
It’s very easy to modify the data presented in dashboards. In Data Studio, for example, you could wrap a Google Analytics metric in a calculated field and change it. Or you could obfuscate your results by leaving out a metric that contradicts the narrative you’re seeing (or hoping to see). Someone has to be responsible for monitoring all these factors.
4. A standard interpretation of data can’t be guaranteed
It’s easy to talk about ‘data’ like it’s objective. And the more you have, the more objective you can be. The truth is that two people can look at the same numbers and take something totally different away. This can be true of all the different categories of data that might be available to you. For example…
METRICS: What is bounce rate? Does everyone have the same understanding of this term? What’s a good bounce rate? There are so many contextual factors that affect what a good bounce rate looks like. And that’s before you even get to questions of how important this metric should be to your business, and why.
SEASONALITY OR STRATEGY: Maybe you check the dashboard in December and see a clear overspend on media. Is that a good thing or a bad thing? Different people in your business may have different priorities, or be privy to different strategic decision making processes. All this will affect their interpretation of data points like this.
INTERNAL EVENTS OR CHANGES: At any given time, a website could be undergoing development updates or user experience experiments. Your paid media team could be testing different campaign strategies. Does everyone who’s looking at a dashboard know about all website changes, experiments or other factors before they draw insights from the data they’re seeing?
5. Perceived ROI is inflated when opportunity costs are ignored
People often go into the dashboard creation process with the belief that they can invest a relatively small amount of time building a dashboard, which will then save huge amounts of time for their organisation in the long run. In this line of thinking, your usual monthly reporting process is replaced with a one-time dashboard set up, followed by extremely efficient reporting at the click of a button, forevermore.
While a dashboard can offer a really accessible and handy reporting solution for a variety of stakeholders, you’ve got to account for the cost of maintenance and upkeep of the dashboard, as well as training people on using the dashboard correctly.
But perhaps most importantly, you have to account for the high cost of poor interpretation of data (or accurate interpretation of incorrect data!), if the dashboard isn’t used or maintained correctly. This is why I say a dashboard enables a data analyst but doesn’t replace them.
6. Data security and access management
While the shareability of dashboards is a real benefit, it’s very important to consider how you limit access to sensitive data – both internally at your organization, and externally too.
The rise of cloud computing has made data accessible almost everywhere, with relatively basic levels of authentication. Access to digital dashboards can be vulnerable to email phishing scams and the like. And scams aside, a Google Data Studio dashboard can be made public on the web very easily, or made accessible via private links.
This means sharing and editing capabilities need to be carefully controlled and monitored to avoid any commercially sensitive data being exposed.
Ok… so how do I build digital dashboards that actually add value?
Like I said, I love dashboards and I’ve used them frequently to help clients get more from their data. So here are my five steps to make sure you use dashboards in the right way.
- Treat each dashboard like a new software product. Whichever data design methodology you adopt, make sure that you collect commercial requirements, build a business case, and design and deliver features iteratively.
- Have a central team that designs, delivers and maintains dashboards and reporting for the organization. Their work is to empower other teams, but they also need to be the guardians of data consumption in the organization.
- Training should always be provided on how to interpret the data, before granting access to dashboards. Plus, the dashboard itself should include caveats or notes in any places where metrics might be misunderstood.
- Add usage analytics to your dashboards. You should review the stats at regular intervals, and dispose of or update any areas of the dashboard that aren’t getting used.
- Make data validation a constant priority. From the name of a paid social campaign in Facebook, right through to the way that campaign is mapped to revenue data in Google Analytics via Data Import, you need to ensure you’ve validated your data “end to end”. And you must keep validating and monitoring data over time. Alert your team to any anomalies immediately – don’t wait two weeks before realizing that conversion tracking has broken. The integrity of your data is paramount to making the right decisions.