Insights

Data Analytics as the Driver of Business Strategy

Introduction

At Blacksheep Consulting, analytics is not just a capability, it’s at the core of what we do. We believe that data analytics is the most important step in developing and implementing business strategy. Data can be thought of as the elements of a journey that when engaged with, tell us a story of where we are and how we arrived here.

To illustrate our philosophy, let’s go on a hike. We could track our hiking path on a map as an analysis of our journey, but there’s so much more to that: Was there mud along the path, hills and valleys, rocky terrain, rain or fallen trees blocking the natural path? All of these are additional data and factors that impact our journey.

Analytics is the understanding of the timing, cause and impact of these factors and the development of a narrative describing the journey. How much of the journey was travelled through mud? How far along the path was the first hill? How steep was the valley? How far away was the fallen tree when we first saw it? Analytics answers these questions, and packaging it tells the narrative. What do we do with this narrative though?

The impact of the obstacles faced on the journey is defined as insights. How much did travelling through mud slow the journey? How much additional energy was expended due to the hill? Was the valley helpful in making up time on the journey or was it too steep and required extra caution? Did the rain exacerbate the cold and slow the journey further? By how much was our travel distance increased due to travelling around the fallen tree?

Now that we know the impact of these factors along our journey, we may have the ability to predict or estimate the likelihood of it occurring in future, determining our strategy for future hikes. Assume we are planning for our next journey. We are now equipped with the knowledge of the potential elements along our path that could hinder or assist us. How do we accommodate for the mud? Should we ensure we have rubber boots for our travel? Did the hill expend a significant amount of our energy, or did we expend too much energy prior to arriving at the hill? Could carrying and consuming an energy drink or snack bar compensate for this, and when should it be consumed? Could using a rope tied to a tree at the hilltop allow us to climb down the valley slope easier? Should we ensure we pack waterproof thermal jackets for the journey to combat the rain? Essentially, through analytics and insights, we are empowered to develop a strategy that optimises our journey and write the next travel narrative in a manner that we pre-define. This is the core of the analytics value chain that we use at Blacksheep Consulting to assist businesses in optimising their performance.

The analytics value chain

Blacksheep Consulting utilises an analytics value chain approach when advising on business processes and strategies. We use this value chain as the underpinning element of all our advisory services, and build successful business, functional and operational strategies with insights at the core.

To understand how data can effectively be used to develop better business strategy, it is important to understand the analytics value chain. Just as a production value chain starts with raw material and adds value through each step until a consumable product is bought, the analytics value chain adds value to data until a business strategy is successfully executed.

The analytics value chain can be visualised as follows:

We can think of data in its most basic form as raw material. Unprocessed data that is housed in a data warehouse (or if you’re unlucky, on an Excel spreadsheet, or even worse, uncaptured hard copy) is the most basic form of data. Other than it being stored somewhere, it carries very little inherent value. Raw data can take many forms, and in any analytics value chain might come from various sources. A successful analytics value chain will typically consist of a mixture of internally generated data (financials, transactional data, procurement data, etc.) as well as external first and third party data (primary research, syndicated research, external databases, etc.).

Let’s make this practical. Imagine you own a convenience store that, among other things, sells bread. Each time you sell a loaf of bread, you generate raw data: You have a sale price, date and time of sale, brand of bread sold, shelf positioning and product expiry date. Potentially you even have data on the customer should you choose to record it, such as customer age, payment method, frequency in your store, demographics, products purchased in combination with the loaf of bread, etc. All of this constitutes raw data that can be analysed.

Data processing involves a first level understanding of what the raw data can provide, its shortcomings and data cleaning. It’s the first step in getting raw data to something that can add value and allows an analyst the opportunity to think about what inputs will be used in building the eventual business strategy. It is also at this stage that the link between multiple data sources are understood, and the plan for how to integrate them is developed.

Data analysis is the core in the analytics value chain. It can take the form of the most basic analysis of data (e.g. tabulation and summary statistics), or can employ complex statistical analysis and modelling (e.g. regression modelling, cluster analysis, data dimension reduction) to transform raw data into something meaningful. Good data analysis requires someone with an instinct for analytics – the best information in a dataset isn’t always obvious; an analyst with good instincts and experience will often uncover hidden strategic information in a dataset.

Continuing with our convenience store example, what does the data analysis step look like? Some analyses that could be generated are:

  • The impact of price on sales volume
  • Volume of sales analysed by time of day
  • Sales by brand
  • Sales by shelf position
  • Sales by product expiry date
  • What are the most popular payment methods for customers

Some of these relationships could be false signals however or could be confounded by other data points not yet measured. This is where good analytical instincts play a role. Consider: We may find through analysis that sales volume increases as price decreases, and therefore conclude that there is an inverse relationship between these. It is a logical assumption, but a good analyst ensures that all other options have been considered. What if, at the time that the bread sales volumes increased, it happened to coincide with a reduction in supply of hot-dog rolls or another substitute product?

Similarly, consider sales volumes by brand. A simple analysis might conclude brands with higher sales volumes are preferred by customers. Other confounding data might however influence the conclusion, e.g. differences in each brand’s pricing, each brand’s shelf placement or how long each brand has been on the shelf.

From these examples, it is clear that data analysis can become an intricate, time consuming and potentially never-ending process of delving deeper and deeper into the data. It is therefore important to ask the critical question before starting any analysis: “If we found the answer to this, what could we do differently, and what impact could that have on our business?” Even though a deep analysis could provide a conclusive answer, it might not change anything the business does, and therefore resources may be better used to answer a different business question.

Once data has been analysed, it has been transformed from a raw material to a consumable product. The next step from having analysed data is to package the analytics in a way that can be easily consumed by stakeholders that may not be analytically inclined. This usually takes the form of charting, dashboarding and presentation building – the underlying analytics as well as the strategic intent of the business will guide the kind of packaging required. A good strategic analyst will be able to package analytics in a way that is engaging to any audience and sets up the analytics in a way that allows for a strategic insights narrative to be created.

Together with analytics, insights generation is likely the most important step in the analytics value chain. Many analysts are good at analysing and even packaging analytics but cannot generate the required strategic insights or narrative from the analytics to transform it into something that is actionable. This step in the value chain requires a link between analytical capability, strategic and critical thinking and excellent business acumen.

What could insights generation look like in our convenience store example? Given the analysis performed, we might be able to come up with the following narrative:

Price has the greatest influence on sales volumes and keeping everything else the same, the lower the price, the higher the sales volumes. However, we identified that a segment of customers who aren’t as price sensitive, will purchase bread at our store regardless of price and will not purchase a higher volume even at a lower price. This customer segment is predominantly customers aged between 23 to 26 who purchased both a loaf of bread and a litre of milk consistently every second day. These customers also tend to complement their purchase with a snack of sorts, between chocolates, chips, or both. However, these products are at a lower frequency of purchase.

Furthermore, hot dog rolls and bread loaves are interchangeable for customers. Meaning that if there isn’t stock of one, the customer will easily purchase the other and the consolidation of bread loaves and hot dog rolls show stable volumes over time.

The expiry date of the bread however is also significant in determining sales. We found that the later expiry dated bread loaves sell first and on occasion, bread with next day expiry will not be purchased by customers.

Once a strategic insights narrative has been created, a business strategy can be formulated. While the strategy formulation process can justify an article of its own, it’s important to note that a data driven strategy formulation process requires the full analytics value chain to be successful. A strategic insights narrative is the key input into this strategy formulation step, which is followed by the identification of external forces, market opportunities and alignment of business operations to opportunities.

Strategy development in our convenience store example would involve an examination of the macroeconomic environment our store operates in, as well as incorporate the customer behaviours we found in the insights phase. This could then involve the generation of a high-low pricing strategy that ensures customers who will buy at higher prices will be charged a higher price to increase margins, while more price sensitive customers can make use of promotional pricing. It would align our demand planning to avoid having products that are close to expiry being on the shelf and incorporate a marketing strategy that speaks to the customer behaviours identified. Key strategic goals could include:

  • Increase sales volumes in strategic customer segments
  • Improve profit margins in strategic product categories

The analytics value chain does not stop with the formulation of a strategy, however. A well-executed data analysis step will have identified strategic performance gaps as well as business metrics to be monitored to continuously evaluate the successful execution of the strategy. Strategy execution and continuous improvement therefore complete the analytics value chain through ongoing monitoring of key data points that indicate the success of the business strategy, and suggest strategic redirections as and when required.

Strategy execution in our convenience store would involve operationalising against the strategic goals that have been set. For example, we could target those customers between ages 23 – 26 whose purchasing behaviour we already know, but could be influenced to increase the frequency of purchase. We may do this by something as simple as repositioning the snacks closer to the bread or at the point of sale, or offer discounts on bread with the combination of a snack.

Profit margins might be increased by increasing sales of hot dog rolls as they have higher margin than bread. One option could be to reposition hot dog rolls at a more influential shelf level, while moving bread to a less popular level to increase the sale of hot dog rolls.

Continuous improvement usually takes the form of test-and-learn exercises. For simplicity, let’s assume there are just three factors that impact the sale of bread: price, substitute products (hot dog rolls) and expiry date.

In period one, we could ensure there are enough substitute products, all expiry dates are the same and only price differs. In period two, we could stabilize price and ensure availability of substitute products but vary stock expiry dates. In period three we stabilize price, keep only stock that expires on the same date but vary availability of substitute products. At the end of this test period, you have fully controlled for the three variables influencing sales and can make an informed decision on how to improve sales further.

Conclusion

Data has become one of the most valued commodities in business today. Consumers and businesses trade their personal data for services, and it has become a universal currency accepted by many businesses, predominantly within big tech.

It is clear that consumer data has become this valuable as it can be used to tailor advertising and generate sales in a meaningful way. The rise of using big data and AI to more effectively advertise to consumers has perhaps overshadowed a more fundamental use of data as a whole: the development of business strategy grounded in sound data analytics.

By considering the analytics value chain as a key input into business strategy formulation, the process of developing strategy becomes an evidence based approach that elevates a strategy to something that is achievable, measurable and ultimately generates more business value.