Data can be important in making better decisions, but relying solely on data driven insights without context can prove costly in time and money. I’ve seen organisations suffer from data analysis paralysis when attempting to base decisions on data alone.
ore data means more analysis, and if you don’t have data models designed for a specific outcome, unstructured data is of little or no use in decision-making. Yes data is extremely valuable in marketing automation, personalisation, product innovation and gaining customer experience insights, but only when you have a future state vision.
Data insights have to be actionable, requiring a goal orientated approach. Knowing your desired outcome needs to be clearly defined, and more importantly actionable, with carefully designed feedback loops to stay on track. This is where I’ve seen most data-driven approaches fall flat.
According to Forrester “Despite big investments in big data. We found that while 74% of firms say they want to be “data-driven,” only 29% say they are good at connecting analytics to action. That is the problem”.
Sometimes what’s needed is more insights, goals and action, not more data.
A successful approach to data-driven decisions could be summarised in the 5 level learning model and applies not only to marketing and CX, but to all areas where we need to garner insights and make important decisions.
We often use this 5 level learning model to make sense of data and can be scaled for any organisation. It starts with data and ends in wisdom.
Knowledge (+ experience)
Level 1 - Data
Data is all around us and is mostly free. All organisations have good internal data, from customer purchases to website and process data. External industry data is also widely available and more specific data from governments, big business and even the CIA at no cost.
The IBM Institute for Business Value estimates we create 2.5 quintillion bytes of data globally a day.
So there’s enough data out there, you just need to find the correct data to suite your specific needs. The key as a learning organisation is understanding that data is only useful if it’s actionable.
A simple example is improving the user experience on a website. If clients are complaining about a specific website page or feature we can gather data from the website analytics, heatmap tools and feedback just to name a few. This data on it’s own doesn’t offer us any solutions, it’s just data, until we begin to structure and contextualise it.
Level 2 - Information
Information is data structured for an outcome or goal. Google search and books are examples of data being structured to provide information on a subject. Being informed by no means gives us deep insights into a specific subject, but rather forms a basis for future action.
Following with our website example, gathering quantitative data from the website isn’t enough on it’s own. Qualitative data from ethnographic surveys and customer feedback will give you a clearer picture of which pages/ processes on the website need improving.
Information compiled from these multiple data sources (internally from employees as well as externally) is then used to create a touchpoint map, highlighting the user interactions with the company in relation to the specific problem.
We now have a clearer picture of the problem which means we can start looking for solutions.
Level 3 - Knowledge (+experience)
Knowledge is organised information. All this new information gathered from data is of little value though without action. It’s like reading a book, spewing out what you’ve read, and thinking you are knowledgeable about the topic. Knowledge comes from applying the information into something actionable.
We all have tons of information at our fingertips, yet we don’t know what to do with it. This is where as mentioned previously we often fall flat because we either don’t have a clear strategy to move forward, or we choose to ignore the findings and carry on doing what we’re doing.
As per our website example, having the touchpoint map is of little value if there’s no action taken to rectify the issue. Once we’ve discovered the problem, using design thinking enables us to generate a few solutions based on past experience (or best practice) combined with our new information.
This is simply us using the information, gathered from data, to propose and test solutions with real customers. It usually takes the form of usability testing, workshops or even basic card sorting.
Level 4 - Understanding
Putting your knowledge or learnings into practise (experience) is when we gain real understanding of the information and data. This allows you to improve where necessary and create innovative products and services that customers love.
Understanding entails using your knowledge gained from experience to make better decisions. It’s about putting what you’ve learnt into practise to evaluate the data and test your assumptions. This is data put into an action. In our simplistic website example, designing even rough UX layouts on paper to get feedback from customers will give you an understanding of whether your ideas are solving the problem, or improving the experience. This allows you to better understand customer frustrations, allowing you to take action and eliminate (or reduce) obstacles to a seamless interaction.
Level 5 - Wisdom
Wisdom is using our understanding to plan for the future and use foresight and good judgement to reduce uncertainty. This is also where successful companies are proactive in solving customer issues before they happen and execute effective marketing campaigns that are personalised and targeted.
From experience, most organisations think they can plan for the future without going through levels three and four first. This applies across industries although I find is most common amongst startups who, from data insights launch products without the doing proper customer research and testing.
Not nearly enough resources are put into testing and getting early customer feedback with predictable results.
In our website example, this stage is the most crucial as you need to apply your understanding of customer frustrations and prevent future design pitfalls. The pinnacle of good experience design would be going beyond solving current problems by offering innovative solutions that customers didn’t even know they needed.
In conclusion, remember that as humans we have a natural decision-making bias. This affects how we collect, and use data to justify our decisions - correct or not. You should also evaluate data quality and compare assumptions with actual testing to understand the results is also crucial.
While some say data never lies, look for patterns that can skew results and affect your outcome. So, regardless of how you plan to use data, make sure it’s a blessing or not a curse.