Over the course of a few months, we published a range of ABCs to help businesses think about generating more impactful insights.
Here they are, condensed for your reading pleasure.
The ABCs of Insights
While often overlooked, an analysis plan ensures that effort is directed toward uncovering the insights that address the business’ questions. Without one, it’s easy to spin your wheels and create mountains of charts, assuming quality = quantity. Even a simple analysis plan can up your insights game, and ensure you remain on task.
Human beings are, unfortunately, quite biased. We tend to view the world from our perspective, assuming others are like us. As marketers, it’s easy to forget you’re not your target market. Good survey design, analysis technique and insights generation is about being mindful of this bias, minimising it, and creating as objective a view of the world as possible. Think ‘To what extent would you say…?’ not ‘Do you agree that this is a good idea?’.
It’s important to remember that we are all people: people working with other people to understand people. The human race has gotten this far because of collaboration, so let’s keep it going. Find and embrace opportunities to collaborate across client, agency and research participant – you’ll be amazed at how much gold there is when coming from a place of mutual respect and interest!
When conducting pack tests, product tests, comms tests, new logo tests, messaging, etc., we need to replicate the ‘real world’ as much as possible. This means using research techniques that separate people into new states of the world, and comparing the statistical difference, or ‘delta’ between different conditions – not asking people to evaluate ‘new worlds’ side by side – this almost never happen in reality.
Ethnography is all about understanding people in natural settings to more deeply explore attitudes, behaviours, cultural practices, etc. While expensive in the past, the adoption of virtual technologies, mobile research, new diary tools and observed data collection means that ethnography is a powerful, and ever-more accessible way to deeply understand behaviour.
The humble focus group has had a tumultuous journey, in parts equal parts criticized and celebrated in pop culture and insights land. From the mainstay of agencies and businesses alike, COVID moved groups online, where many still remain. Their popularity remains: there is little substitute from talking to people directly, with all the rich non-verbal that comes with it, and the opportunities for organic conversation.
About 15-20 years ago, gamification was just taking off. Given decreasing attention spans and proliferation of surveys, capturing participant attention is more important than ever before. After all, insights are only as good as the data, and that relies on engagement. Gamification is all about making surveys as engaging and fun as possible, and can transform the respondent experience… leading to better data!
Harking back to our first post this month about analysis plan, starting with a set of hypotheses is a powerful way to supercharge your insight. It doesn’t matter if it’s conservative or provocative, having a set of hypotheses you set out to prove / disprove can help focus insights – from how you design your collection, to wading through the data to hand in search of proof.
To generate richer insights, you need to uncover the unconscious. To do so, we need to tap into System 1 thinking – the instinctive, gut-reaction, subconscious associations of consumers. One way to do so is use implicit association techniques. Across a large enough sample size, this technique is a fantastic way to uncover the unconscious and better tap into what people really think and feel – even if they won’t say it!
The most powerful insights are often the most seemingly simple. The ‘ah-hah!’ moment that makes everything fall into place, and it just seems so… obvious. The antithesis of this is a report full of jargon: measuring a JND at a 95% CI to maximise ROI on a l10n NPD PL strat… what? Great insights reveal profound truths in easy-to grasp-ways – rejecting jargon socializes and keeps insights accessible.
KPIs & Key Performance Indicator Setting
Many people fear KPIs: given how much these are tied to remuneration, manipulation is incentivised. From gaming the collection to the gaming the numbers… it’s waste of an insight opportunity. Truth is, KPIs are a powerful means to measure progress, and setting targets should be done in a sensitive and nuanced way, for instance goals relative to the market. This can then generate real insight – not box ticking.
A longitudinal study is one in which the same group of individuals are followed over-time. Typically, these are used to measure some social outcome – but they can be powerful for commercial organisations too. Organisations with communities can use these studies to carefully build out a robust ‘whole of person’ trigger, and measure changes during key lifestages, unlocking powerful consumer insights.
Monadic testing refers to participants only seeing one item at a time – think one pack, or one ad, or one product. This is a great way to approach testing, as it allows careful consideration of a future state to a today-state, and more closely reflects the reality. Doing so will get you to a deeper understanding of the impact of changes and more closely represent reality, leading to better decision making.
Words have immense power. For this reason, the language we use is important, as is the nuance we uncover. To lead to a profound understanding of something and uncover the insight, be comfortable with, and active in exploring the nuance. Hone your words. And hone them again. In nuance lies the richness of understanding, and the power of business success.
One of the greatest stumbling blocks of any research data is the ‘say-do’ gap: what people claim to do, and what they actually do, can be worlds apart. It’s been seen time and again. Wherever possible, fusion of observed datasets (e.g. actual transaction data, online behaviour, in-store data, etc.) with survey data (their attitudes and reflections) can reveal a truer picture. And from this, businesses can make better decisions.
The five Ps
Sure, most marketers stumble into this at uni and it’s quickly forgotten, but the 5Ps remains a fabulous framework for measuring the opportunities for brand success. In short, ensuring your research program accounts for measurement across the domains of Product, Place, People, Promotion and Price will quickly help reveal areas of strength, and weakness, to guide direction.
Qual / Quant
There's long been a divide between researchers along the qualitative / quantitative fault-lines. The truth is this is an artificial divide. With new technological capabilities, qual can happen on quantitative scales, and quantitative surveys can be filled with rich free-text and sophisticated prompting techniques. To generate better insights, it's vital that we abandon this idea of the divide.
Random Forest Regression
One of the more exciting developments in research analytics is using random forest as a regression technique. This is a ‘learning’ algorithm where multiple decision trees are constructed in different ways and we see what predicts the outcome best – wisdom of the crowd, in analytics form! This is an accurate and fantastic way to better model an outcome, and help focus on the metrics that matter.
Story telling insights
We’ve all sat through a dreaded 100+ slide slideshow, with the report seemingly structured in order of presentation flow. It’s the fastest way to put an audience to sleep. Instead, focus on storytelling. Rich visuals, compelling narrative, hypothesis-driven analysis, centred on the business questions ruthlessly focused on less-is-more – the ultimate elevator pitch with visuals; try this and you’ll see engagement increase!
Types of market research data
Broadly, there are two types of data you can deal in: primary, and secondary. Primary data is a powerful way of capturing unique data, just for you, tailored in a way you need, to inform competitive advantage. Secondary data comes from existing sources, and is a great way to reduce the number of ‘unknowns’ without overinvesting. Knowing when to use each, and how, is vital.
One of the most powerful things we can do as insights professionals is be unambiguous. Provide clarity where you can, and be honest where you can’t. The data to hand and consultant’s expertise should be used to create a clear set of research conclusions, and recommendations. The ‘what, so what, and now what?’s if you will. Always strive for unambiguous, jargon-free, clear insights.
Virtual Reality Research (VR Research)
New advances in virtual reality (foveal tracking) to reduce processing load for gaming, can be leveraged to provide powerful market insights. VR lets us create worlds at the touch of a button, which consumers can walk and interact with, while we track where they look. This allows for rich behavioural insights, at a fraction of the cost of making such changes in the real world, giving confidence in how best to proceed.
What Who Where When Why
When looking to brief in a research agency, the Five Ws provide a great template. Why are you conducting insights work? What will the insights be used for? Who do you want to interview? When do you need insights? Where should the research be conducted (both method and geography)?
Great insights, starts with good design. Thinking carefully about the end-objective using ‘Backward Market Research’ and developing an experimental design to produce these valuable insights is a fantastic way to ensure your research and analysis generates the right outcomes. If you haven’t, read HBR’s 1985 “Backward Market Research” – it’s incredibly powerful.
Year on Year
One of the many challenges we see with tracking studies is the emotional rollercoaster – numbers go up, numbers go down, and we jump at shadows. Taking a step back to look at the bigger picture by investigating year-on-year trends can reveal the necessary insights to course-correct. This smooths out the fluctuations and helps uncover the bigger trends at play. And from this, you can uncover the insights needed to guide decisions.
It’s important to remember brands and behaviour don’t operate in a vacuum. Everything is influenced by the wider climate in which our research occurs. Cultural understanding must be up-to-date and regularly updated, ensuring any behaviour is decoded against the backdrop of the zeitgeist. Without this understanding, real insights and trends can be missed.