Analytics-to-Value: Digital analytics optimizing products and portfolios

| Article

Leading companies have made great strides over the past several years in managing the costs of components and raw materials—which can account for up to 50 percent of a manufacturers’ cost base—using design-to-value (DtV) programs. But with the pandemic’s economic impact likely to be deep and painful in many sectors, companies are under intense pressure to drive a new level of material-cost efficiency.

One of the most promising new strategies is analytics-to-value (AtV). This set of digital- and analytics-based practices enables dramatic increases in product cost and portfolio-optimization efficiency. It does so by exploiting product and procurement data (enriched with a wide range of internal and external data sources) and applying a digital version of traditional material-cost-management practices.

AtV uses a broader range of cost-optimization tools and approaches than DtV, covering everything from portfolio, product, and component design to sourcing. By leveraging the power of data and analytics, AtV generates more insights in a shorter timeframe (often within a few days or weeks), at scale and across the entire portfolio.

Focus on use cases rather than technology

The AtV approach helps companies focus on solving for specific, business-generated use cases rather than becoming seduced by the shiniest, newest technologies. This use-case orientation might sound elementary, but too often companies lead with the technology—starting with an answer in search of a problem instead of a problem in search of an answer. For example, companies often choose a new IT solution, such as an add-on to an existing ERP system that promises to unlock siloed data across the enterprise. But by itself, “unlocking data” may not achieve much without an understanding of what the data are to be used for—in other words, without understanding the specific business use case. Broadly defined IT projects often disappoint because in the absence of a well-thought-through, use-case orientation, they become inordinately complex, take far too long to complete, and have no clearly defined or measurable business value.

But well-defined use cases are only one requirement. Another frequent problem occurs when employees create smart, tailored analyses that end up being one-off efforts because they lack the resources to continue these analyses and refine and scale them into a replicable solution. Analysts might take a few days to cleanse and link multiple data sets in a spreadsheet, deriving meaningful insights for the next management presentation. But since they do not have the time to repeat the effort, the spreadsheet sits on their computers and is not used again.

The pillars of analytics-to-value

Analytics-to-value comprises four pillars that together combine a use-case orientation with greater sustainability and rigor.

Customer value and product design. Product optimization starts with a solid understanding of customer needs, which means companies must collect the right data in order to make quantified trade-offs between adding customer value and reducing cost.

Portfolio and modularity. Companies can optimize their product portfolios to increase margins, while using modules, platforms, and standardization to reduce internal complexity.

Technical specs and solution engineering. Because excessive technical specifications so often drive product costs, an optimization effort—such as through benchmarking with competitors’ offerings—can have a significant cost impact.

Components and supply. Identifying and securing the best cost point for product and service components—and from a set of reliable and resilient suppliers—requires continual monitoring of changes in the prices for both the components themselves and their raw materials. A company’s suppliers should be equally focused on optimizing component costs.

The four pillars of AtV can be applied either individually or in combination to create impact at scale. For example, running an AtV transformation along pillars 1 and 2 will fundamentally align a company’s product portfolio with customer preferences while at the same time finding the best trade-off between portfolio complexity and internal efficiency. On the other hand, focusing on pillars 3 and 4 in a joint, cross-functional project for a given product portfolio can result in a large-scale commercial and technical optimization of direct spend. Several companies have linked all four AtV pillars in a holistic AtV transformation, establishing an entirely new way of working for product optimization, including building AtV capabilities in the organization and establishing the necessary tools and systems, as well as governance. The first article of this series will focus on pillars 1 and 2, while a future article will be a deep-dive on pillars 3 and 4, as well as the holistic AtV transformation (Exhibit 1).

1
Analytics-to-value offers a holistic view of product optimization.

Understanding customer value for product design

The pace of innovation and the timeline to bring new products and services to market continue to accelerate. As a result, more companies have shifted to an iterative, agile design process centered on the creation of a minimum viable product (MVP) as a way to bring digital, user-friendly products to market quickly.

The ability to identify customer needs, and design products and services accordingly, has always been important—but one that historically has tended to be a qualitative exercise based on the experience of marketing and sales executives. Today, by contrast, there are methods to exploit the vast amount of customer information available online and maximize returns on customer research. Leaders across all industries looking for robust methodologies and powerful analytic tools in improving customer value and product design face two fundamental questions:

  • How can I get targeted and actionable insights from a dispersed customer base quickly?
  • How can I balance customer value and product cost?

Extracting insights from a dispersed customer base

Traditionally, companies have gathered customer feedback through a limited number of formal channels, such as surveys, interviews, and focus groups. Today, companies can tap into a treasure trove of customer sentiment from a range of sources, ranging from social media and web reviews to emails and call-center transcripts. The challenge has been that it is freeform comments in unstructured text that typically contain the most spontaneous and unbiased feedback from customers—and, until recently, interpreting the vast amount of available comments would require such an intensive manual effort that most of this valid information still remains untapped.

New digital tools that can analyze unstructured text at scale could dramatically expand companies’ ability to interpret customer sentiment, accurately identifying what drives customer satisfaction and dissatisfaction. These analyses provide crucial insights about the strengths and weaknesses of brands, product and services, how to improve existing features, and how to make the customer journey and user experience better.

For instance, customer sentiment towards different brands can be compared across different categories by automatically analyzing and classifying freeform comments into categories that define the performance of a product or service (Exhibit 2). In this case, brand 2’s purchase process is a major source of positive customer feedback, while brand 1’s customers are dissatisfied with servicing and repairs. Diving deeper into the categories allows to identify the drivers between customer sentiments: While our brand’s customers are mainly discontented with turnaround time and repairs price when it comes to Servicing and Repairs, it is the purchasing experience and ease of purchase that gives Brand 2 an edge in their great perception of Purchase performance.

2
Social-media posts and other unstructured text can reveal what matters most in product and service design.

Balancing customer value and product cost

With a clear picture of customer needs and wishes, the design process moves to concept creation. From a business point of view, the relative value of customer wishes must be quantified and weighed against their cost, with a view to maximizing product profitability.

Typically, the relative importance of product or service features is inferred from MaxDiff1 or conjoint analysis of consumer surveys. But this process can take several months and results are strongly dependent on the quality of the analysis methodology. Now, however, there are user-friendly digital AtV tools that in less than a week can prioritize features based on customer feedback and deliver solid answers about customers’ willingness to pay for certain features.

Being able to link the relative importance of certain features to their cost generates a powerful perspective on important tradeoffs. Armed with this knowledge, companies can identify product features with marginal consumer importance, which allow them to set priorities better when creating or redesigning products—and create a single source of truth in aligning stakeholders (such as sales, engineering, procurement and quality), who tend not to share a uniform understanding of customer needs and preferences. At a shoe company, for example, this analysis enables the product development team to identify several features that could safely be discarded, saving costs without risking consumer dissatisfaction and revenue loss (Exhibit 3).

3
For ballet shoes, 6 features represented only 10% of customer value but 25% of product cost.

Besides identifying improvement opportunities in existing products, companies can use these tools during the concept phase to allocate material-cost spending on features that maximize customer value. Several automakers have used this approach to allocate spend on trim in their new car models by deciding which seat features, multimedia systems, driver assistance systems, and the like should be offered in each trim level (such as sport, comfort, and luxury). In addition, some automakers use the methodology to prioritize deeper design-development decisions, as in whether to prepare the wiring harnesses of a new car generation to support a certain option, even though not all models will be equipped with it.

Now is the time for procurement to lead value capture

Now is the time for procurement to lead value capture

Cutting complexity with a modular product portfolio

Managing complexity and scaling efficiently is a major challenge, but the good news is that it is possible and, when done well, it can become a competitive differentiator. It means addressing complexity’s two distinct halves. “External” complexity, or the number of product lines and configurations customers can order, can be countered by optimizing the product portfolio, reducing its cost of complexity (CoC), and increasing margins. Internal complexity, on the other hand, centers on the sheer number of different components the company must manage, which it often can reduce by introducing platforms and modules.

In practice, successfully attacking complexity boils down to answering four essential questions:

  • What drives complexity at our company, and what does it cost?
  • How can we optimize our product portfolio for cost and sales?
  • How can we further reduce our internal complexity?
  • How can we keep complexity from reemerging?

Finding the causes of complexity

Product-portfolio complexity is not necessarily bad. A company may choose to offer a wide variety of products to differentiate itself. Often, however, at least some of that complexity does not create additional customer value: products designed to reflect narrow, regional product preferences, for example, may fail to generate adequate returns, while inadequate product-lifecycle controls can mean supporting outdated products after their profit potential is exhausted—and can contribute unnecessarily to portfolio complexity.

The first step to optimize complexity cost is to identify the main complexity drivers, which depend strongly on the individual company’s profile and offering. These drivers can, for example, be the number of countries in which a company operates, sales channels, product groups, products and components. A company may find that a small share of its large product portfolio generates 80 percent of gross margins. At the same time, the company may learn that its product complexity is very costly and a good share of these costs could be recovered—indicating a potential to simplify the portfolio and, with limited margin loss, significantly reduce cost.

Optimizing a product portfolio for cost and sales

Optimizing a product portfolio is difficult because simply pruning a few products often won’t reduce CoC—but will very likely reduce revenues: if two products share the same component but only one is pruned, the company still needs to hold the component in stock and so the associated CoC is not eliminated.

Because it’s often difficult to understand the immediate impact of pruning efforts, companies are often skeptical of the approach. The key to lowering CoC is to prune product clusters that share drivers of complexity (Exhibit 4).

4
Advanced analytics identifies optimal combinations of pruning candidates. Advanced analytics maximizes complexity reduction while minimizing sales impact.

Advanced analytics such as portfolio visualization can help improve portfolio optimization. Exhibit 5 shows a portfolio network in which the overlap of complexity cost drivers (components) is represented by the distance between dots (products). Using this portfolio mapping, a company can identify very distinct products that use many unique components, and clusters of products that share components and might be candidates for pruning in their entirety.

5
Portfolio visualizations make complexity transparent.

It is possible, though tedious, to analyze these visualizations with traditional tools. But powerful AtV algorithms can test the impact of pruning any combination of products and identify those products that, when pruned, will reduce CoC at minimum sales loss. These calculations are possible because the algorithms can identify specific components, factory lines, and other inputs that will no longer be necessary once a product or group of products has been pruned. This is much more efficient than the classic approach in which the lowest sales products are pruned first, and it helps to ensure tangible results with a clear business case.

Effectively reducing internal complexity

Once the product portfolio is optimized, companies can reduce internal complexity by using platforms, modules and standardized components.

Product platforms help companies capture synergies between similar products. Rather than defining each product from scratch, the company defines a standard outline for a family of products to capture operational synergies. However, even when a platform has standardized elements, the platform still needs the flexibility to tailor products to specific customer segments, regions, or channels. In the automotive industry, for example, a platform can allow changes in drivetrains and related components to support everything from a utility vehicle to a performance coupe. A consumer-goods company extended the concept to develop platforms to support common flavor profiles for different products.

Modules can capture synergies across platforms. Platforms that seem quite different often share certain functions (such as the need for a power supply). By standardizing these functions in a module with clearly defined interfaces, companies can simplify internal complexity.

Finally, the company can reduce internal complexity by standardizing components across products—a goal that is now more achievable than ever thanks to new technologies. Spend data and technical specifications can be merged so that machine-learning algorithms can quickly help identify parts similarities, based on factors such as geometric shapes (extracted from computer-aided design drawings) and data extracted from spec sheets. Further efficiency comes from setting clear guidelines about the components the company prefers, with measures limiting the use of unapproved alternatives. Continuously updated digital component catalogs, together with governance practices to limit exceptions, reinforce the needed discipline.

Keeping complexity from reemerging

Reducing complexity is a transformational exercise that must be continually practiced or complexity will reemerge. First, top management must actively promote the cause, which is inherently cross-functional. Without clear direction from the top, different incentives across functions could undermine the anticomplexity effort.

Second, governance must be designed with the strength to limit new complexity from emerging. This includes clear portfolio, platform, and module ownership to ensure commercial and operational tradeoffs are made diligently. Cross-functional steering boards can then help make important portfolio decisions based on a high-level, holistic business-case perspective.

Lastly, a company needs clear KPIs to measure complexity and monitor progress, and these KPIs need to be linked in a way that helps companies make the necessary trade-offs between cost and value.


AtV tools will have a profound impact on material cost reduction in the future. These can significantly improve product design and customer value, as well as reduce complexity and boost margins through portfolio optimization and modularity. Going forward, companies that adopt these approaches most rapidly will be poised for a competitive advantage.

But their work won’t be done. Companies can use AtV to unlock even more value by improving technical specifications, optimizing components and supply costs, and running an AtV transformation at scale. These topics will be addressed in the second article in our two-part series on AtV.

Explore a career with us