Are you overwhelmed by all of the data related to COVID? You’re not alone. “Analytics” is part of my title, and I’m struggling to make sense of it all.
Is the fatality rate 0.2%? 0.84%? Both are cited as possibilities in this article from UC Berkeley, yet this New York Times article suggests that we don’t have the right data to know. Similarly, some local Georgia media are reporting that we’ve reached our peak in COVID-19 cases, while other sources say we’re not there yet.
All of this conflicting and confusing information illustrates a core challenge that has always faced marketing and public relations teams: identifying, analyzing and explaining marketing and PR data.
The four primary mistakes that we’re seeing right now in regard to COVID are the same we often see in marketing and PR measurement:
- Cherry-picking data – The most egregious mistake when it comes to data analytics is citing only the numbers that support your case while ignoring the rest. Scroll through Facebook or Twitter for more than 10 seconds, and you’re likely to see a COVID-related example of this. In marketing, cherry-picked data often takes the shape of citing campaign successes, while ignoring red flags. For example, if your agency cites that website visits increased 20% during a recent campaign without also noting that the bounce rate increased, time on site decreased and conversions plateaued, then they’ve painted a false picture of success.
- Mixing and matching data – With so many data sources available, it can be hard not to mix and match data from various sources, but doing so creates challenges in verifying and interpreting insights. When it comes to COVID, Johns Hopkins emerged as an early leader in the U.S. for its dashboard, but that data doesn’t quite match what’s being reported by the CDC. While differences in how and when this data is reported explain the discrepancies, flip flopping between two sources leads to confusion. Similar challenges arise in marketing analytics when differing data sources are used. For example, reporting conversions out of HubSpot but clicks out of Google Ads to report on results from a search campaign can leave marketers with unactionable insights.
- Using raw numbers – Not putting data into context can lead to inaccurate assumptions. For example, 117 people have died from COVID-19 in Albany, Georgia. This number may not set off alarm bells given the tens of thousands of deaths nationwide, but when you add context to the number, the numbers become more sinister. The infection rate based on population is among the highest in Georgia and the fatality rate based on confirmed cases is more than double Georgia’s largest county. The same challenges arise with marketing and PR data. Is 10 articles about a new product launch good? If a company only got 2 articles on its previous launch, then yes! But if it got 20, then this launch may not have been a media success.
- The data simply isn’t there – The lack of COVID data in the U.S. has been much discussed. Take a close look at Georgia’s race data – it’s missing in more than 30% of cases, making it hard to draw definitive conclusions. And as in marketing, it’s often impossible to collect accurate data after the fact. For example, want to know if social media posts with images of people gain more engagement than those without? This is easy to measure with the right tracking on the front end, but would take a lot of manual labor to determine after posts have gone up.
While the influx of data on COVID-19 has put data analytics issues center stage, such challenges have plagued marketers and PR pros for years and continue to worsen as we collect more and more information. While it’s an ongoing process to keep your data clean, consistent and insightful, start with these best practices:
- Find a source of truth and stick to it. Ensuring that your team is using the same data source(s) will provide replicable data that can be used to find patterns and draw conclusions. Document your source of truth and refer to it often to reinforce this behavior.
- Define what you want to track from the start. Knowing what you want to measure – and what data you’ll be able to act on – will help you collect the necessary metrics from the outset of an initiative. Before any campaign, make a list of what you want to measure and how data will be collected to bring measurement and analysis to the forefront of your strategy.
- Create reporting templates – Templatizing reporting for marketing and PR campaigns will help to ensure that all pertinent data is reported, minimizing the opportunity to cherry-pick only the data that tells a success story. Templates also help to train your team and stakeholders what information to expect, making them more likely to raise concerns or ask questions when key data is missing.
- Use data to paint a picture – Even if your data is accurate and consistent, getting stakeholders to understand its meaning is still a challenge. Putting data into context and using it to tell a story can help you get your message across. For example, hearing that COVID-19 has now resulted in more deaths than the Vietnam War is more impactful than hearing a raw number of deaths – helping people put the gravity of this situation into context. Similarly, an executive learning that 100 media placements in a single quarter outpaced coverage the past three years combined is more impactful than just reporting on the sheer number of articles.
As the full impacts of coronavirus continue to unfold, we’re monitoring for other lessons learned and impacts for marketers. Follow our COVID marketing resource center here.