One of our biggest challenges as planners can often be feeling like we’re drowning in data.
Sam Knowles came in to discuss his recently published book Narrative by Numbers: How to Tell Powerful and Purposeful Stories with Data, which had recently been nominated for Best Book at the Business Book Awards.
Sam's a strategist with 30 years’ experience helping some of the biggest brands tell more powerful stories. Through psychology, data analytics and crafting narratives, he's helped some of the biggest organisations talk and sound human.
On the night, we had a wide ranging discussion on how to come up for air, make sense of data and pick out the most compelling data to help us tell the most powerful brand stories.
How did numbers help Brexit?
Sam opens his book with a little philosophy followed by a reflection on how numbers helped create the #Brexit narrative. What was it about the that magic number - £350 million for the NHS - that worked? He explained to us that it was a round-but-not-suspiciously-round figure, which tends to work better; it was designed to become earworm that could be repeated by both sides ad nauseum; connecting the number with the NHS – the three most emotional letters in the British political lexicon – gave it emotional power; and, above all, it was simple, vague, didn’t invite analysis and the only number consistently used. It was also clever because, placed deliberately outside the context of government expenditure, it concealed the fact that this number was actually tiny. It clouded truth in emotion. It’s a very instructive lesson in the power of data in storytelling.
How do we know what numbers to choose?
‘Are we fiction writers or fact writers?’, was one question put to Sam. More to the point, how do we know what numbers are the right numbers to choose? Sam mentioned a project in which he looked into birth trend names to get a sense of wider changes in Britain. There were a lot of explanations, but trend was interesting and provided space for interpretation. These changes provided a compelling picture, and from this, he worked to ‘triangulate’ with other data points to provide substance and explanation. He said ‘the statistician never goes on a fishing expedition’ – they’d much rather start with a hypothesis and explore the data, and so should planners. And always be on the look-out for a ‘hidden third cause’ of your observation.
How do we deal with the pressure to proffer insight?
You can sometimes feel desperate to provide clients with an insight or story when you know you don’t have enough facts in your hands. Knowing what you don’t know and need to know is a basic essential skill of planners, but before commissioning data, Sam has found clients often aren’t sharing all the data they have. They may be sensitive about sharing it, or they may feel it’s not relevant. It’s always worth trying to get hold of whatever client data you can. For example, getting weather data can unlock huge insight into consumer behaviour, as we’ve seen with soup and ice-cream brands.
What tools are a data story-teller’s best friend?
Firstly, going back to the previous point, the best thing you can do is ask for the best data you can get, especially client data. Be difficult up-front and ask for it. Be clear with them what you’re looking for. Getting this will help you set sensible KPIs and make your campaign more likely to succeed. The other thing is to not get scared off by all the methods and tools out there. SPSS and TGI can be beneficial at times, but in Sam’s view, nothing is better than good quality, unfiltered, relevant data. Excel and pivot tables are a planner’s best friends.
What’s more important: correlations or anomalies?
While every case is different, anomalies are usually more useful to hunt town than correlation. Difference is always better in Sam’s view, unless that correlation is very strong, or highly unusual. What the planner should be looking for are things that have potential to be psychologically or emotionally resonant. This is especially important given what we know now about the primacy of emotional ‘System 1’ decision-making. The watch-out is, the data-driven stories planners craft for consumers still need a bit of ‘System 2’ thinking in there to justify the stories (we make sense of our decision-making rationally). We need to deftly blend emotions with choice facts. It’s our job to find the data and create the fuel to spark creative work
To show, or not to show?
Sam was asked how much we should show our information and workings to clients. Because storytelling can be a powerful bridge in making data and insight meaningful to others, to help them make informed decisions, what’s the right balance to strike in how much data to provide? While clients and agencies differ, Sam recommended making simple rules to focus yourself on powerful storytelling. He tries to limit decks to 20 slides, but he’s OK with providing ‘infinite appendices’ if needed. Every planner needs to be able to show their workings if asked, but most of the job is about convincing the right people at the right time. You want what you say to have ‘memetic potential’, so focus on the story, and have back-up. Your back story should only ever be shared “if asked …”.
How much can story depart from fact?
But what about when the stories you’re telling depart from data and fact? Things can go wrong when disagreements over strategy become disagreements of opinion because they depart from research. In the discussion, it was suggested that research agencies should focus on presenting data and offering top-line observations, but not to offer explanations or insights. It’s for others to up to analyse and interpret. One contributor said it may be that surveys offer sight on consumers, but they offer no vision, which is up to the planner to create and hone that perspective to solve client problems. The important thing in data-led storytelling, Sam said, is to know what information is important and how this underpins your narrative, so the job of the planner us underpinned by fact. On the other hand, Sam suggested, research agencies have an opportunity in the strategy space.