In an earlier post we summarised the basic tenets of three major decision-making models namely RAPID, Digital Insight Management and the Lean Analytics Cycle. To further understand the main ideas and implications of each one of them, we started a series devoted to their thorough exploration. We first examined RAPID (by Bain & Company) in a previous blog post, and now would like to focus on the Lean Analytics Cycle.
As claimed by the creators of the model, Alistair Croll and Benjamin Yoskovitz, the Lean Analytics Cycle is primarily oriented at start-ups, who might experience an initial budget shortage and a slight disorientation in the immense marketplace. Their aim, therefore, is to come up with an innovative solution in the shortest amount of time, with the lowest expenditure possible and leveraging their existing resources. However, the creators also emphasise that their method is not exclusive to smaller companies and can be successfully implemented in larger enterprises.
The Lean Analytics Cycle promises to assist entrepreneurs in more efficiently achieving their aims by following four simple steps, always relying on a solid learning process and constant experimentation and innovation. While these four steps may indeed seem logical and easy to follow, they nevertheless present certain challenges that we would like to analyse.
WHAT: This first step involves choosing one aspect of your business you would like to improve (or, in the creators’ words, One Metric That Matters) and tying it to a certain KPI, in order to ensure a correct measuring process.
PROBLEM 1: While this step might work perfectly well with small start-ups, it can be problematic for middle-sized and large companies, where there are too many aspects to tackle. Moreover, these enterprises tend to have great amounts of employees and capital at their disposal, which enables simultaneous dedication to more than one issue at a time.
SOLUTION 1: Of course, it is important for larger organisations to have a full picture of their business objectives and devote time and resources to all of them. However, prioritising one problem for a short amount of time can actually lead to a quicker and more long-term resolution. This will then enable the company to efficiently approach the next objective.
PROBLEM 2: Another problem at this stage of the decision-making process is the lack of solid knowledge based on facts. In other words, I might believe that the aspect I am choosing is key to improving my performance, but how do I know for sure?
SOLUTION 2: In order to clarify and pinpoint our most significant business challenges and identify what is impeding us from strong performance, we should set clear business objectives and analyse data to see where we are not meeting them. The times of relying on “gut instinct” are long past now that we have access to great amounts of data and analytic strategies for understanding this data. For example, we can turn to our target audience and examine their complaints to decide which should be addressed first. Or, we can use our analytics tools to discover our weak points. This can be done in two ways: firstly, we can rely on our past performance to see where we are not achieving our aims. Secondly, we can create “what-if” scenarios to explore what metrics are crucial for our set objectives, to eventually discover that One Metric That Matters.
WHAT: Inspiration and creativity come into play now: you have to provide a viable solution for the business objective identified in the first step.
PROBLEM 1: The creators tackle two possibilities for this step: when the company has enough data and when it does not. Both imply certain problems: while the first case scenario may rely too heavily on data and forget about the human factor, the second seems rather unreliable and somewhat risky.
SOLUTION 1: While creativity and innovation are undoubtedly crucial in business management, we would still recommend relying on solid data in any decision-making process. Therefore, if your company does not have enough, dig for it! Turn to your clients (via surveys, questionnaires or direct interviews), your competitors (analysing their successful strategies), or even your colleagues, who can share their experiences with you. You never know where valuable insights might come from!
If data shortage is not your problem, you shouldn’t forget that raw numbers and statistics won’t guarantee strong performance. Even the most detailed data is worth nothing if it is not appropriately analysed, so if you want to form a clever hypothesis and gain valuable insights, it’s time to add the human factor into the equation. The more creative and innovative thinking you apply to your data, the better and smarter your hypothesis will be!
PROBLEM 2: As outlined in the previous post, other decision-making models clearly define the ownership of and responsibility for the process. The Lean Analytics Cycle seems to leave this issue unaddressed: who is the person to form a hypothesis? Either everyone would pass the buck to someone else, or different people would form different hypotheses based on what they believe to be the most important business problem (most likely the one that is most relevant to their department and not necessarily the whole company). Both scenarios would result in a chaotic mismanagement and eventually no proper decision would be made.
SOLUTION 2: It becomes apparent that clear role definition should be outlined before every decision-making process. Each person has to be fully aware of what is expected of them at each stage of the process, so that no time is wasted in unnecessary and lengthy debates. This also applies to the next step.
WHAT: Decide who you are appealing to, what exactly you want them to do, and why they would listen to you. When these three variables are clear, design your experiment to test your hypothesis.
PROBLEM: Running tests might turn into a risky and costly procedure, especially in case of start-ups, which the model is supposedly addressed to. Implementing a daring change has high chances of not meeting one’s expectations, as proved by the case studies that the creators provide, which might be unaffordable for a small company running short on budget. Or even for a bigger company that doesn’t want to ruin its reputation for the sake of a risky experiment.
SOLUTION: Much talked-about segmentation may be the cost-effective answer. Instead of running a big experiment involving major changes in your company’s brand message or business activity, you may choose to focus on a small sample within your target audience. This initial sample can always be enlarged in the case of success.
WHAT: An expected final step consists of evaluating the results of your experiment. The process is simple: if it proved successful, implement the changes, and if not, learn from your mistakes.
PROBLEM: How exactly will you know whether your efforts have proved successful? The answer seems pretty straightforward: by measuring the outcome of your experiment. However, while it is widely accepted that measuring one’s performance is a must for any enterprise, unfortunately, not all companies have yet embraced a data-informed approach to their business goals.
SOLUTION: If they want to survive in the marketplace, companies cannot pass on the inclusion of analytics tools. Although initially it may seem like a costly investment, it will definitely pay off in the long run. Measuring is key to success in the contemporary data-driven business environment, and implementing analytics will not only help you evaluate your experiments, but the whole performance of your organisation as well. This will help you make well-weighed and data-backed business decisions.
Our analysis has attempted to illustrate some of the problems that might result from the Lean Analytics Cycle, and how smaller start-ups and bigger enterprises can approach solving them. While it is impossible to have a flawless decision-making model due to the infinite complexity of the market and the decisions it requires, the Lean Analytics Cycle proves to be a quick and effective option that focuses on scalable results and continuous improvement.
A definite takeaway is the One Metric That Matters. It proves to be efficient both in terms of time and money, since it allows you to adopt a step-by-step approach tackling single problem at a time, which means thorough analysis and dedication and leads to more scalable results. Some final comments on this decision-making model address the correct combination of your analytics and human assets. Firstly, as mentioned before, analytics tools implementation is a must, and your business cannot advance without them. Secondly, even when you are fully equipped with those tools, you cannot forget about the human factor. Creative and motivated employees who have access to valuable data is what you should aim at. And, finally, follow this model’s advice for experimentation: do not be afraid to try out new ideas, always segmenting your targeted audience to avoid unnecessary costs.
What are your thoughts on this model? Has your company experienced similar problems, and if so, how did you tackle them?
Not Another Dashboard.