Timken Rolls Out a Marketing Dashboard for Industrial Bearing Group
 |
The Timken Company makes things run smoother — on land, at sea, or in space. With over $5 billion in annual revenue, Timken manufactures a broad range of bearings, steel alloys, and other products that facilitate motion between mechanical parts. Their Industrial Division, the subject of this case study, serves the friction-management needs of key industrial markets like construction, mining, aerospace, rail, consumer, steel, power transmission, and general industrial equipment.
Operating in a Lean Six Sigma environment, Timken had already tackled many of the areas in which they had seen the biggest potential benefits from process improvement, namely operations, where the problems tended to be more easily scoped and the data was available for analysis. Yet it was becoming increasingly clear that the real benefits might not accrue until they could apply Lean Six Sigma to the customer-facing dimensions of the business like customer satisfaction, pricing, growth potential, and so on. But before they could start optimizing those elements, they had to baseline them. So, the original concept of the dashboard was to start to begin to measure some of the customer-facing activities.
 |
Choosing to tackle one division at a time to help control the scope of the project, the Timken marketing team, headed by marketing director Brian Berg and dashboard project manager Kristin Stanford, set out to build a marketing dashboard for the Industrial Division that would provide the necessary foundation for the subsequent application of Lean Six Sigma discipline.
Aligning on the Definition and Role of Marketing
The first step for the team was to meet with senior executives of the Industrial Division to get their perspective on the specific role of marketing in supporting their business objectives and how they evaluate marketing’s efforts. With that stake in the ground, the marketing team facilitated a “value stream” workshop with representatives from product management, divisional marketing, sales, finance, IT, and operations.
The first step for the team was to meet with senior executives of the Industrial Division to get their perspective on the specific role of marketing in supporting their business objectives and how they evaluate marketing’s efforts. With that stake in the ground, the marketing team facilitated a “value stream” workshop with representatives from product management, divisional marketing, sales, finance, IT, and operations.
In the workshop, they started out by clarifying the Industrial Division’s business goals to ensure alignment amongst the team. Next, they defined a specific set of marketing objectives to support that overall mission. Then the group worked to identify the performance metrics that would most insightfully measure success, looking first at what they could measure with available data, then at what they thought they should be measuring. Finally they sought to determine both the nature and the magnitude of the gaps between could and should as a means of prioritizing their learning needs.
It didn’t take long before the group discovered that referring to their project as a “marketing” dashboard caused confusion about the scope. At Timken, as with many companies, “marketing” is not solely the province of the marketing department. Many functional groups are involved in the act of marketing, and to measure the breadth and depth of marketing activity would tread into the responsibilities of many groups outside the marketing department. Accordingly, it was very important to negotiate terms of what should and shouldn’t be on the dashboard and what implications were ascribed to each element.
 |
Choosing the Metrics for Version 1
The workshop group developed a list of over 50 items they hypothesized might be the right things to measure, grouped into three buckets, loosely defined as:
The workshop group developed a list of over 50 items they hypothesized might be the right things to measure, grouped into three buckets, loosely defined as:
- business metrics — unit and dollar sales metrics normally tracked on a daily, weekly, or monthly basis;
- marketing effectiveness metrics — communications impact, marketing planning, pricing impacts, lead-to-order conversion, etc.; and
- marketing efficiency metrics — measures of the ROI associated with various marketing investments.
The senior management stakeholders, however, felt 50 metrics were too many for a useful tool. So the dashboard team further screened the options by scoring each proposed metric on the basis of the value of having the information vs. the degree of difficulty in obtaining the necessary data (see matrix below). Metrics were assigned to one of four categories, with priority logically placed first on those in the high-value/ease-of-access quadrant. The secondary prioritization was a mixture of lower-value/ease-of-access and high-value/difficult-access elements for which a clear development path to get to the necessary data was evident, even if it wasn’t in place yet.
 |
The goal was to make the dashboard as simple as possible in version 1 so that it didn’t overwhelm people with data. Knowing they could always expand metric categories in subsequent versions, the team tried to pick just the necessary few key metrics per area of measure.
But while simplicity in scope and design are important, they are not necessarily sufficient for a successful dashboard. With each selected metric, the team also found it important to test the interpretation of the results to ensure they were really measuring the things that facilitated key decisions.
One such example dealt with the ratio mix of sale of core traditional products to newer lines. When initially expressed as a percentage of the overall sales, the team found users unable to ascertain which product lines were growing and which were shrinking. But when adjusted to present sales of each line against a target percentage of overall sales, the information more clearly signaled if and where action was required. In short, the team found that metrics that measure performance against target goals vs. against each other helped improve users’ ability to derive relevancy from the dashboard. Moreover, the team observed how some initially “clunky” presentations of metrics might actually be driving behaviors contrary to the divisional goals and to the marketing objectives.
Drilling Down
Once they defined the right metric categories, the team had to determine what views into the data were required and how deep to go. Version 1 of the dashboard was designed to match the organizational structure of profit centers built around lines of business and broken down by geography. This way, the company president can look at performance from a global level by division, then drill down to regions and/or business units. Business unit leaders could look at the data by region and by market to find specifically what they need to fine-tune performance.
Once they defined the right metric categories, the team had to determine what views into the data were required and how deep to go. Version 1 of the dashboard was designed to match the organizational structure of profit centers built around lines of business and broken down by geography. This way, the company president can look at performance from a global level by division, then drill down to regions and/or business units. Business unit leaders could look at the data by region and by market to find specifically what they need to fine-tune performance.
Despite its engineered flexibility, v1 of the dashboard wasn’t constructed to be a data mining tool with which users could drill down to the nth degree of detail. Early assessments of user needs found that the first thing people wanted was overall performance reporting presented in a concise, consistent manner. The team tried to develop a balance between presenting only the most meaningful and insightful metrics and allowing some degree of depth of analysis, without dramatically expanding the scope and complexity of the project, possibly scaring off all but the most technically savvy power-users.
 |
Given its restricted scope of drill-down dimensions, the initial iteration of the dashboard was developed in Microsoft Excel, with imbedded hyperlinks to facilitate users navigating around in a non-linear manner. The read-only file is posted on a shared drive with direct user access via password protection. Users each have viewing privileges that differ by their logon identification. Data is updated continuously as new information becomes available, but not in “real time.” Importantly, the underlying data structures and graphics are tightly controlled and maintained by a single authorized manager.
Start to finish, the process to get v1 operational took approximately six months, driven by one dedicated manager/analyst and partial time commitments by a variety of supporting teammates from finance, sales, operations, and individual business units.
Some Key Lessons Learned To Date
There were a few things the team learned in the process of deploying their first version of a dashboard that are worthy of mention here.
There were a few things the team learned in the process of deploying their first version of a dashboard that are worthy of mention here.
First, they had difficulties defining and staying within scope in the early meetings with users. Ultimately they discovered that by focusing on “marketing activities” instead of the “marketing function,” they were able to demonstrate the value of a dashboard across suborganizational groups and engender greater support for the project in both spirit and resources. In fact, one of the beneficial outcomes has been a more consistent definition within the company about what marketing is and who is responsible for the various component parts. In hindsight, the team felt they might have had fewer initial obstacles by referring to the project as a “business effectiveness dashboard” instead of a “marketing dashboard.”
Second, one of the things the team felt worked in its favor was starting within the auspices of a Design for Six Sigma project. That gave the project legitimacy in many corners of the company and suggested a level of discipline in process to be applied, as well as a suggested outcome.
Third, the size of the team might have been too large initially. In an effort to engage as many of the key stakeholders as much as possible, they invited representatives from many functions and suborganizations, which created the classic challenges associated with large groups. In hindsight, it might have been more effective to have a smaller operating team reporting to a broad steering committee, requesting support with prioritization and resource allocation as required.
 |
Planning for the Next Evolution
As the Timken dashboard team continues to collect user feedback and observe the impact of the dashboard on decision processes, they are beginning to define some of the priorities for subsequent phases of dashboard development. Key priorities include:
As the Timken dashboard team continues to collect user feedback and observe the impact of the dashboard on decision processes, they are beginning to define some of the priorities for subsequent phases of dashboard development. Key priorities include:
- upgrading the look and feel to create a more user-friendly navigation experience;
- continuing to fine-tune the selection of metrics to focus on only the most relevant and meaningful elements, with an eye toward reducing the overall number;
- testing various means of assessing the ROI associated with specific marketing activities on a business unit or cross-company level;
- developing the insight to understand which external factors of a macroeconomic nature are drivers and incorporating them into the dashboard;
- acquiring the data necessary to populate some of the hypothesized high-value metrics that they’ve been unable to execute thus far, such as the development of some type of share-of-customer tracking capability, predicated upon developing a method of calculating the estimated total category purchases of customers — something that the team recognizes will require some carefully stakeholdered assumptions amongst senior managers; and
- evaluating the potential for replatforming the underlying database as the company continues down the path of its SAP system installation — the dashboard is out in front of this ERP implementation, so Web enablement will likely have to wait for the larger IT initiatives to catch up.
The Benefits Accruing
Already, senior managers at Timken regularly check the dashboard for input and guidance into critical business decisions. In the past, to get a true picture of marketing activity performance, they typically had to search through several source documents, each with their own lexicon, which weren’t designed to align or provide a consistent interpretation. Now, with just a few mouse clicks, they can get to just about any metric they need and as a result are operating with greater speed and certainty in making critical decisions.
Already, senior managers at Timken regularly check the dashboard for input and guidance into critical business decisions. In the past, to get a true picture of marketing activity performance, they typically had to search through several source documents, each with their own lexicon, which weren’t designed to align or provide a consistent interpretation. Now, with just a few mouse clicks, they can get to just about any metric they need and as a result are operating with greater speed and certainty in making critical decisions.
The Timken marketing dashboard has only been “live” for a few months now, but it’s showing all the signs of a well-planned project with great potential for improving marketing planning and execution. Moreover, it’s a textbook case study for a strong dashboard development process in any type of industry.
