Last year, humans created more data than in the past 5,000 years. Yet, very little of that data is actually analyzed and used for strategic and operational decision making or to generate new revenue.
A survey conducted by McKinsey & Company in 2017 suggests that efforts to monetize data are fairly new and a majority of companies have only started doing so in the past 2 years. Yet, more than half of the respondents in that study agree that the most important change in their competitive environment is new entrants launching data-focused business and disrupting traditional businesses.
There is urgency for companies to make effective use of their data if they want to remain competitive. With data changing the way that business is done, how can you put your data to good use to increase your sales and profitability, and stay ahead of the competition?
Ecommerce Departments in a Data Driven World
Recruiting the best team members or implementing complicated data mining and analysis tools might seem like an insurmountable barrier for small and midsize companies. Simply observe what large companies are experiencing. Implementing a data-driven culture and conducting profound changes in a company’s organizational structure, leadership and culture is a monumental challenge and takes years.
However, companies with an online presence might just have unexpected resources within their organization to help maximize the value of the data that they are collecting, their Ecommerce departments!
According to the 2017 Global Ecommerce Report executed by the Ecommerce foundation and Mazars, 78% of internet users in the United States have made purchases online in the last 12 months. The e-shopper population continues to increase in the United States, despite already having the world’s highest internet inclusivity rate, and, more importantly, average e-shopper spending is increasing every year. We can expect these figures to continue increasing as millennials, now the largest and most influential consumer group, are moving into their prime spending years.
As a result, ecommerce departments within small and midsize businesses will be exposed to mountains of data collected from increased online traffic and sales. They might just be the best players within their organization to exploit this immense potential for growth if they are given the right tools: a structured database to choose data metrics from and a clear understanding of the company’s objectives.
Finding the Right Data Metrics for Your Business
Cecile Peters, senior e-commerce director at Clos19, Moët-Hennessy’s ecommerce platform dedicated to luxury champagnes and spirits of the LVMH group, sees the volumes of data and metrics as one of the main challenges faced by ecommerce departments. Data is collected through a wide variety of channels and has to be prioritized in order to extract useful analysis and derive insights that are actionable.
Vincent Van de Maele, a data scientist at Zettafox, a data science and prescriptive analytics firm acquired by Mazars in 2017, explains that the first challenge for companies is to optimally gather and organize their data for analysis. For instance, data created by different departments within a company, or through different systems, is often stored in silos.
Data silos are isolated from each other and the rest of the organization and, if they are not integrated, cannot be used to their full potential. The first step that an Ecommerce or Marketing executive should take is to inventory those internal systems collecting data and identify the type of data that is being collected in order to integrate them in the most efficient way.
If a Company lacks the resources internally to perform this task, data analytics firms such as Zettafox can provide these data architecture services to assist their clients in “cleaning up” their databases. “We do this on a daily basis” assures Vincent.
Once your data has been cleaned, the next step is to identify the right data metrics to provide the best insight for your business operations. The right data metrics should be quantifiable and closely tied to your company’s objectives. They should track both the company’s main areas of business (for an ecommerce department, website traffic) and critical areas of performance (for instance, the contribution of ecommerce sales to incremental sales).
This implies that management should ask the right questions on how to best achieve its goals and know how to measure such achievement. For instance, do you know what fuels your company’s growth? Is it fueled rather by new customers or returning existing customers? Can you identify your customers’ buying cycles and triggers? Data metrics should measure the company’s progress in achieving a goal (i.e. be trackable) and inspire actions (i.e. explainable).
For an ecommerce department, identifying the right data metrics will help monitor and understand customer purchasing behavior and improve client knowledge. As a consequence, thanks to an understanding of the customer’s expectations, coupled with an anticipation of the customer’s needs, the company will gain a competitive advantage.
Challenges to Data Analytics
Any company can go online to any given marketing website, use the traditional data metrics that are listed and follow the different approaches to selecting the metrics. However, because choosing the right data metrics will be an iterative process, they will not all be relevant for your business. Data metrics will evolve and will need to be tweaked and tailored to the company’s needs in order to achieve the best measure of performance, progress and goal achievement.
Zettafox’s data driven solutions can greatly streamline this iterative process. Zettafox uses artificial intelligence, more precisely machine learning algorithms, to develop its models. Machine learning is defined as a method of data analysis that automates analytical model building. The tailored prescriptive model developed by Zettafox can learn from data, identify patterns and suggest decisions, with minimal effort from the operational teams. The company will simply have to ask the right questions and an unlimited number of questions can be asked!
For a Swiss flavor and fragrance company, Zettafox helped find key ingredient combinations related to fragrance performance that optimize an existing formula, based on both customer surveys and technical properties. For a New York-based cosmetic and beauty company, Zettafox helped increased by 280% the conversion rate of samples for direct online orders by better characterizing customer perception of samples, more efficient bundling and understanding of buying frequency.
Traditional Data Metrics Used by E-Commerce Departments
Let’s discuss some of the metrics that can be used by companies based on data that is readily available to Ecommerce departments.
Cecile, from Clos19, explains that with the growing importance of online retail, companies have shifted in recent years to omni-channel sales and marketing strategies to provide clients with a continuous and integrated customer experience. She points out that online stores are the new flagships.
In that respect, the first metrics that an Ecommerce department should explore are related to online traffic, such as first visits and returning visitor metrics, web traffic sources to understand what is driving visitors to your website, or brand awareness metrics to track how people hear about your brand (social media, search engines).
Once the website has attracted and spurred a customer in interacting with the company, Ecommerce departments are interested in predicting how much revenue a customer will contribute throughout their lifetime. This is achieved by computing a Customer Lifetime Value (CLV), which is calculated by analyzing a customer’s transaction history, along with various behavioral indicators. There are no plug-and-play solutions; this data metric can be computed in different ways depending on how a customer can generate revenue (direct purchases, referrals, indirect marketing, etc).
Ecommerce departments should also track the cost of acquiring customers (CAC) because the gross profit of an initial purchase should be compared to the costs incurred to generate that sale. In fact, companies should be trying to maximize their CLV in relation to their CAC.
An adage in business says that it is 5 times more expensive to acquire customers than it is to retain an existing one. We are not here to argue over whether it is a universal truth or a myth. What is certain is that retention requires constant engagement with your customers and a strong commitment to deliver relevant content and value.
As such, in addition to monitoring the CLV, an Ecommerce department could monitor data metrics such as customer attrition, the customer response rate to e-marketing campaigns (newsletters, social media or targeted emails) or the purchase frequency. Such data will give the company insight in terms of targeting the right customers and addressing customer churn.
Evaluating Your Data
In a data focused world, being able to accurately predict the lifetime value of a customer and implement a data-driven strategy to retain customers are paramount. We went over some of the data metrics that could be monitored to provide insight on these challenges. These techniques are only as effective as the reliability of the data being gathered.
Implementing a Data Metric Rating System
As we discussed above, data is found in a variety of states and is often stored in silos. It can be a familiar database populated through “marketplace” solutions and from which a team member can easily download the required data. Or it could originate from a proprietary system that few people know how to exploit.
Caitlin Hudon, author of data science blog Haystacks and senior data scientist at Shelfbucks, an in-store merchandising optimization and mobile media platform, points out that “not all of the data is equally relevant to the questions we are asking of it, not all of the data is trustworthy and not all of the analyses are neatly reproducible.” She put together a data meta-metric rating system that she used to “convey the quality of the data and its collection process” to her team based on:
- Relevance – (1) data is perfectly relevant, (2) data is useful to add color to arguments or decisions, and (3) important decisions should not be made based on this data;
- Trustworthiness – (1) Data is trustworthy, including its source and ways it is captured, (2) there are reservations on the data trustworthiness, and (3) data is not trustworthy due to the way it is collected or stored;
- Repeatability – (1) process to retrieve data is fully automated or automatable, (2) the process is standardized but involves manual operations, and (3) the data does not live in a data based and extraction is very manual.
This system could be a source of inspiration for companies when inventorying they type of data that there are collecting. It will help Companies diagnose their strengths and weaknesses in collecting and storing data and could lead to improve the collection and storing processes.
The increased importance of data and the interest shown by new entrants in using this data to disrupt the market place and fuel growth have pushed companies of all size to reorganize their business model to adapt to these new factors. Potentially implying profound internal reorganization, these changes can be driven by the Ecommerce department, as they already collect and deal with data on a daily basis.
The main challenges to overcome are the ability to inventory and structure the data, often originating from various systems and stored in as many databases, and use this data efficiently and in alignment with the company’s objectives. Many solutions exist in the market to assist businesses in that process. However, because becoming a more data focused firm is an iterative process, none are plug-and-play. Companies would be well-advised to seek the expertise of specialized data analytics firms.
 Access the e-commerce study here.
 The Ecommerce Foundation and Mazars ranked countries across four categories: availability (quality and breadth of available infrastructure), affordability (cost of access relative to income and competition on the market place), relevance (existence and relevance of local language content) and readiness (capacity to access the internet including skills and cultural acceptance).
 Haystacks, a data science blog: https://caitlinhudon.com
 Haystacks, November 2017, https://caitlinhudon.com/2017/11/14/data-meta-metrics/