When you first think about innovation, statistics may not be at the forefront of your mind. We are often drawn to the creative side of the process, thinking of wacky ways to generate ideas.
We may stop to think about the ‘why’ and ‘how’ those ideas could be implemented but how often do we really evaluate our process and our success? What are the Key Performance Indicators and how do we know if we are achieving our goals?
Statistics play a vital role in maintaining the optimum performance of a process. Someone who lives and breathes statistics is Maitha Al Junaibi, the Acting Director of the Department of Methodology & Quality Assurance at the Statistics Centre – Abu Dhabi (SCAD). Maitha also manages their innovation lab ensuring ideas and applications are explored from inception through to successful conclusion. Maitha is perfectly placed to understand the need and benefit of statistics in innovation so read on for her take on why statistics should be part of your innovation process…
‘Innovation is widely accepted as an important contributor to increased production for organizations. Employees are encouraged to generate new ideas, promising ideas are selected and a chosen few are implemented. Successful innovation promises benefits but change involves risk and the potential for failure. Larger innovation projects may take time to implement; with new practices requiring time to settle and rewards being delivered over the longer term. With all the potential uncertainty, how can project managers and senior management judge whether the project is on course or has taken a downward dive? How can we fine-tune the project and prepare for deviations from the expected course?
The solution to this question is the application of statistics to proper project management, including innovation management. Statistics fundamentally deal with uncertainty. Statistics provide us with a way of summarizing raw data to produce meaningful measures that can help us to understand whether or not we are achieving our goals. Setting the appropriate goals and objectives for each stage of the project is central to charting progress. In part, this requires a good understanding of what the innovation project is intending to deliver at each stage and for what effort and cost.
Defining the appropriate success indicators is an important part of the task. Ensuring that the indicators are measurable is also vital. Unmeasurable objectives, based on intuition and gut feeling, will be of limited value and possibly detrimental. People are subject to confirmation bias and may interpret evidence either favorably (if they are supporters of the project) or negatively (if pessimistic). For this reason, success must be defined in advance so that it can be monitored appropriately, i.e. we need a clear understanding of what outcomes are expected and by when. We also need a plan for collecting the appropriate data and a plan to produce statistics that show us whether or not we are meeting our success criteria.
Statistics require data. Business intelligence data may already exist in the form of sales and profits for commercial organizations or take-up for non-commercial organizations. As the aim of innovation is to increase efficiency, market share, sales, and/or service take-up our key measures of success will be measures of these objectives; and targets may be gradually staged over time to allow customers to adapt and behavior change to occur. In addition to these hard measures of success there will often be an interest in how customers react to change. Are current customers happy with the innovation? What features do they value and what is not liked? This information will help to fine-tune change. We may also be interested in what is motivating potential customers to hold back from using our services. How can we encourage them through understanding their needs? All of these questions form a part of a company’s marketing strategy and largely should be answered using statistical procedures through market research.
What we want to achieve is good quality data that enable us to make the best decisions in our monitoring and evaluation of the innovation project. If we only took the opinions of satisfied people our data would not accurately reflect the views of the total population, we would not recognize the areas where we could be doing better and not maximize our sales or take-up. Similarly we want to avoid selecting only negative people. As such we want to ensure that our sample selection mechanism includes a random component. It is this randomness that helps protect us from accidentally oversampling groups of people who may be easy to contact and persuade to respond, but who also may be more or less positive or negative than people in general.
We must summarize our mass of data to make sense of it. Measures such as totals and averages provide easy to interpret summary statistics. We can ask if differences in an outcome measure between different groups of people are likely to be real or to have occurred by chance when using samples through applying statistical hypothesis tests. Statistical models can be used to identify which features are more or less important in understanding differences between the characteristics of people when making a choice of one outcome option over an alternative option (to buy or not to buy). Recent advances in data visualization also offer a wide range of tools to summarize data pictorially and to make it easy to see important features.
Statistics are concerned with ensuring high quality data collection through appropriate questionnaire design and sampling techniques. Statistics help us to select appropriate sample sizes, keeping data collection costs down; and provide a suite of appropriate estimation, analysis and presentation tools. All of these tools are vital for proper project evaluation, especially when used alongside clear objective setting and well defined success criteria.’
So if you are involved in an innovation process and you have yet to take a serious look at statistics, maybe now is the time to start.
