Sometimes it feels that all we're talking about these days is data, analytics and insights. Even though we're in the industry, it does feel like it's a conversation that has spilled over into the zeitgeist. I believe there's another article required to express my personal views on where the surveillance economy is headed, but I'll leave that for another day.

With the pressure that marketers have these days to rightly base their decisions on actionable insights, we felt it valuable to have a discussion about that.

Data isn't "data". I guess, in a way, it is but I feel like there should be some more distinctions made. Not all types of data can answer all types of questions. For a marketer to make the best of what they have available, or for them to choose the correct technology, strategies and new hires - it's important start by taking a step back and realising that "throwing data at the problem" isn't really enough. It's not efficient. Many a million has been wasted on big data projects and technology and projects by trying to solve a broad data problem.

This is an opinion piece, so it's definitely not definitive, but when I think about data, or when a client speaks to me about their data needs, I like to match it against some of the questions below. The answers to these questions all tend to lead to different strategies and solutions, with varying costs, resources and implementation timeframes.

Questions to ask: 1. Firstly, do I know what my question is? 2. Do the answers to my questions need to be "in-the-moment" and dynamic? 3. Am I interested in user behaviour, consumer preferences or opinion? 4. Could there be a major difference in data originating from different personas? 5. Is this data meant for internal strategic decisions, or for targeting and marketing purposes?

Below are my thoughts on classifying the types of data mentioned in question 3.

User Behaviour

Behaviour is an analytics discussion. In the social media analytics world it's often referred to as "volumetrics", specifically relating to things like engagement, fan page growth and virality.

In other spheres it could relate to understanding:

• Website traffic
• Journey mapping and journey orchestration
• Conversion and click-through-rates
• General SEO metrics, like time-on-page, bounce rates and the like

The point about utilising User Behaviour is that it's not about the individual. Tracking an individual's movements and behaviours in a way that identifies them is actually not ethically acceptable anymore. So it's all about the collective, and it's definitely about the trends.

Consumer Preferences

Consumer preferences can be two-fold:

1. A user deliberately chooses a preference on a platform, or channel, which alters their user experience. This could be something as simple as whether they want to receive your communication with them via text message or email, or which customer service channel they prefer to use with an organisation.
2. A group of consumers have, and hopefully state, a preference to interact with - and associate themselves with - a type of organisation that appeals to their preferences. In this case think about large swathes of consumers moving towards eco-conscious travel options, or electric cars, or vegetarian diets. These are the kinds of consumer preferences that one can easily track and monitor over time.

Opinion

Opinions tend to be the fuzzier of the 3 types of data listed in this article. It's the qualitative option. But it's still quantifiable, if you utilise the right kinds of tools and methodologies to understand them.

Opinions can change in a single news cycle. While preferences tend to be a little more long-term, opinions can be swayed almost instantaneously. We're in a highly emotive time in our history, and the ability to share and amplify opinions in a second has led - not only - to polarisation in politics, but also the ability for opinions of your brand to swing in a day.

The most recent example that comes to mind is that of Clicks.

At the beginning of September 2020 they're listed as the "Most Loved" brand by Kantar's Brandz Top 30 report, and a few days later they're the most hated brand on the internet with the Economic Freedom Front allegedly burning down stores in reaction to a social media backlash, involving TRESemmé.

Source: https://www.kantar.com/campaigns/brandz/south-africa/

Opinion can be mostly measured in 2 ways these days:

1. Analysing public discourse on social media, forums, television debates and radio talk shows.
2. Targeted surveys asking pointed questions

Understanding opinions on a topic that is vital to your business is a day-to-day affair. Receiving a brand health report at the end of the week is too late. Set up the correct kinds of alerts and ensure that the right people are mandated to keep abreast of this to ensure your company can roll with the punches. **

Now for my opinion

Once you've asked the right kinds of questions, and worked out the best way to get the answers - create a culture of leveraging those answers. Does your Exco sit on a daily/weekly basis to make strategic decisions based off of the data available? Do the reports that get build and paid for actually get read and digested? It's rare, but a directive from the top, and a habit of interrogating the data is vital to any organisation's agility these days. Go out and prosper. Hope this helps.

***I say this often, but crisis monitoring should NOT BE OUTSOURCED. You should definitely have the best PR agency on hand to handle it, but don't take your eye off the game and hope that someone else will let you know when to pay attention.