Sections

Case Of The Month December 2013: Success And Failure In Transformation

John Ward

business transformation dataIn developing the BTM methodology, we carried out 13 case studies of different types of business transformation in large European corporations. Of these 30% were successful, 40% partly so and 30% were unsuccessful. Each case was assessed against the BTM methodology disciplines to understand why they were more or less successful. Many of the failures were due to lack of alignment with the business strategy, lack of clarity of the expected benefits and inadequate risk assessment. In implementation, the IT and process changes were often performed more successfully than the organizational changes, resulting in some benefits being delivered, even in some of the less successful cases. But this rarely was enough to enable the transformation to achieve its strategic objectives and the majority of the benefits. Overall the organizations whose approach to managing transformations included attention to the majority of the BTM component disciplines were more successful than those that did not.

This article reports the results of an analysis of 13 business transformation case studies. Some were successful, some failed and the rest were partly successful. It shows how the BTM2 disciplines influence the outcomes and explains why some are more successful than the others.

To remain competitive in increasingly global markets, many businesses need to transform either what they do or how they do it or both. Economic turbulence and uncertainty can also make the need to change more urgent but, at the same time, make it more difficult to accomplish successfully. Given these challenges, it is not surprising that studies show that only about 30% of transformation programs are completely successful, while 30% fail completely. Our study of 13 business transformation cases of different types in large European corporations is consistent with this pattern of success and failure:

  • 4 of the transformations were very successful achieving all the main objectives
  • 5 were partially successful as some expected benefits were achieved, but not all
  • 4 were unsuccessful, achieving none of the transformation objectives, or they were not completed

Most incurred substantial costs. Every business transformation is different but not unique and lessons can be learned from the experiences of others. Because these cases showed the same pattern of transformation success and failure as other studies, they were valuable in developing and testing the BTM methodology.

The cases were developed through interviews with those involved in the transformation and reviews of relevant documentation. They were carried out, analyzed and written up by teams consisting of experienced academics, consultants and senior company managers. Some have already been published in the BTA “360° – the Business Transformation Journal” and others will be in future. Each case was analyzed in terms of how extensively and how well the BTM component disciplines were performed. The results were compared to identify significant aspects which appeared to affect the level of success achieved. Analyzing these transformations of varying degrees of success shows that those organizations whose approach to managing transformations paid attention to the majority of the BTM disciplines were more successful than those that did not.

The case studies were in the following industries: automotive, pharmaceuticals, construction (see case study in issue 1 of this journal, page 38), food, oil (issue 2, page 46) and chemicals, financial services (issue 1, page 52), telecommunications (issue 2, page 54) and IT. The cases included transformations to develop new products and services as well as restructuring and reorganizing core business functions and introducing global processes and systems. All involved changes in organization structures and individuals’ roles, responsibilities and behaviors, including, in a few cases, large scale staff relocations and redeployments. All the cases included new and significant investments in IT to enable the business changes, but, in all except one of the cases, IT benefits were not the main rationale for the transformations.

The BTM methodology disciplines

As would be expected, the successful cases were largely green with some amber and even a few The direction and enablement disciplines (source: BTA)red. In the unsuccessful cases, the boxes were mostly red and amber, but there were also always a few that were green!

Findings for each of the eight methodology disciplines are now discussed, starting with the three direction disciplines, before considering findings regarding the enablement disciplines and the ‘Meta Management’ aspects.

1) Strategy Management

A transformation needs to be driven by a clear strategic rationale – a rationale which should be easy for every employee to understand, otherwise there will be little motivation to change. All the successful ones had imperatives to transform the business, not just one function. It was also clear that in all the unsuccessful cases the need for transformation was relatively low; either there was no pressing strategic need or it was not seen as a business priority at a senior level.

In three of the four successful transformations the need for change was endorsed at executive level and then time and effort was spent to gain the buy-in of the rest of the organization and develop the ability to undertake the changes. In most of those that were partially successful, the readiness to transform appeared to be ‘high’, as well as the strategic need. They were not entirely successful mainly due to over ambition, or even over enthusiasm; too many ‘positive’ assumptions were made with little assessment of the potential risks.

Fig. 2: Example pattern for a partially successful case (source: SAP) (see also journal issue 1, page 25)

 

Having a clear vision of the intended future business and organizational models and then allowing compromises and trade-offs in the detail of how they are implemented, is most likely to achieve stakeholder commitment. However, in some cases, when the drivers demand urgent action, a top down, mandated approach to implementation can also work, but it tends to achieve stakeholder acceptance rather than positive commitment. Most transformations involve at least two distinct phases – to create a new capability and then to deploy it. In most of the cases the capability was created, but not (yet) always exploited; hence the benefits achieved were often less than those originally envisaged. Creating a new capability can be done separately from business as usual, but deploying it usually competes with other operational priorities.

2) Value Management

In the unsuccessful transformations the objectives and business cases were often vague, based on a ‘benefits vision’ rather than evidence based benefits and an understanding of how to realize them. This made it difficult for some stakeholders to believe the transformation was worthwhile and commit the required time and resources.

There was also often confusion between ‘changes’ and ‘benefits’: for example introducing
common global processes is a change, not a benefit, although it may create the potential for benefits, such as reducing costs or higher service levels. Too often business benefits were overestimated, while the risks and the problems in making the changes were underestimated – perhaps deliberately, otherwise it would be difficult to get funds and resources?

3) Risk Management

Risk management was often glossed over, but given the high failure rate it makes obvious sense to identify and anticipate what could go wrong, before it happens! As a result many risks only became apparent during implementation, leading to increased costs, delays, scope reductions and even abandonment. This reluctance to explore the risks earlier may have been influenced by executive instigation of the transformation, which can discourage negative feedback, making it inadvisable, even career limiting, to point out the potential risks!

To maximize the probability of delivering the intended benefits, the transformation should be planned in short deliverable stages, if possible. This also reduces vulnerability to changing business conditions and makes it easier to adjust the transformation to retain strategic alignment.
In essence, the outcome of the transformation could be predicted from the predominant ‘color’ in the assessment of the directional disciplines. How clearly and comprehensively the transformation strategy, value and risks have been understood and communicated provides a strong indication of likely success. Had the organizations undertaken this analysis early in the transformation, some failures and the significant resulting waste of money and resources could have been avoided.

Having considered how the direction disciplines affect the level of success of a transformation, our attention turns to the enablement disciplines and how well they were performed in the cases.

4) Process Management

The IT and process changes are usually performed more successfully than organizational changes, resulting in some benefits being delivered, even in some of the less successful cases. But this was not enough to enable the transformation to achieve its objectives and the majority of the benefits. In some of the cases IT or process methodologies dominated the overall transformation approach, making the implementation of other changes more difficult. In two of the unsuccessful cases the IT function tried to satisfy all the expressed user needs, which increased the scope and consequently the costs considerably outweighed the benefits.

5) Program and Project Management

Transformations cannot be fully planned in advance and have to adapt to both changing business conditions and program achievements. This is not necessarily a comfortable position for senior management and requires an empowered governance group to oversee and, as necessary, adapt the program. Effective management of the change content and benefits delivery is more important than the efficiency of the process. In some unsuccessful cases the organization relied heavily on the knowledge and capabilities of a third party supplier throughout, which changed aspects of the transformation towards what the supplier could do, rather than what was required. The transformation manager should have expert knowledge in the area that is being changed and also how to manage change in the organization. A key skill is being able to reconcile the differing views of the change and resource implications between senior managers and operational line management. The priority early in the program should be to gain agreement between senior and line management as to what changes the transformation involves, before ‘negotiating’ for the funds and resources required. In some of the less successful cases the ‘contract’ between the program team and senior management was agreed before the views of line managers had been taken into account.

6) IT Management

The transformations whose main benefits were seen as IT cost reduction or rationalization or were led by IT were not successful. Some business transformations become ‘IT replacement projects’, as the first phase is about replacing old technology and systems and IT methodologies and approaches are used with little business involvement. It then usually proves very difficult to regain business interest when the IT part is completed. IT is often in a weak position in the context of a business transformation due to a lack of real business knowledge, but with a perception that they know how it works, but they only know how the IT systems work. These notions created conflict in some of the transformations. When IT ‘won’ the argument the transformation was unsuccessful, but when it was ‘business- led’, any potential conflict was more easily resolved.

7) Organizational Change Management

These cases suggest that organizations should manage business transformations as orchestrated, continuous, incremental sets of changes – co-evolving and coexisting with business as usual priorities. The successful transformations usually addressed the organizational, people and capability aspects first, then the process and IT components. The less successful tried to do the reverse. Understanding and addressing stakeholder issues and having a strategy for accommodating or dealing with them as early as possible in the transformation is vital. The longer the time available to transform, the more the stakeholder views can and should be included in how the transformation is conducted. The methodologies used should enable all the main stakeholders to directly contribute their knowledge and plan their involvement, instead of relying on experts to interpret the stakeholders’ ‘needs’. Stakeholder engagement is a critical success factor in almost every transformation, and early alignment or reconciliation of multi-stakeholder interests is very important in order to avoid, for example, dominance by a minority of stakeholders or destructive negotiations between dissenting groups. The ‘transition curve’ (see figure 3), describing how people and organizations experience major change should be respected. A comprehensive and sustained approach is needed to minimize the period that people spend in the ‘valley of tears’, which is characterized by uncertainty and even disillusionment. Figure 4 shows that different groups reach this point at different times in the transformation. Senior management interests may have moved on, just when many line managers and staff are under stress, usually due to change and business as usual pressures colliding.

Fig. 3: The transformation experience curve (source: SAP)

 

Fig. 4: Employees experience the effects of the transformation at different times (source: SAP)

 

8) Competence and Training Management

Assessing existing competences as part of the ’Readiness’ is important in order to determine the strategy, because what can be achieved is a function of two factors: first, the amount of work required to make the changes and second, the knowledge and skills that can be made available at the required times. If some essential competences skills are limited is needed early in the transformation. In the successful transformations people were informed and educated about what the intended future business should look like. This helped them apply their existing knowledge to determining how the vision could be achieved, but it also exposed where knowledge was inadequate. Where suppliers are providing essential competences, those also need to be appraised and managed – in case they have over-estimated their capabilities. Organizational and individual experience cannot always be transferred from transformation programs in other organizations. In addition to the eight direction and enablement disciplines discussed so far, ‘Meta Management’ considers themes which influence the performance of any type of organizational transformation. From the case studies a number of lessons about leadership, communication, culture and values can be learned.

Leadership

The successful transformations had ‘CEO’ sponsorship and a C level executive leading the transformation. Involvement should be real and visible or other executives will not see it as important. The evidence from these cases suggests that continuous personal involvement in the governance of the transformation is what is needed, but it is not always easy for a busy executive to sustain this over the extended period of most transformations. But the cases also show that the early transfer of ‘ownership’ to a coalition of business managers, who will actually deliver the changes and benefits, is the best way to develop the capability to change. One key decision that needs to be taken is the mode of ‘change agency’ to be adopted. Either an ‘expert task force’ or devolving change responsibilities to operational managers can work, but a lack of role clarity is likely to cause fragmentation and even disintegration of the initiative.

Communication

A common lesson from many of the cases – even the successful ones – is that no amount of communication is ever enough! Informing everyone in the organization why change is necessary and about the consequences of not changing usually needs regular repetition. Equally important is being open about what the changes are going to mean, even if they will be unpopular with some stakeholders. Evasiveness builds distrust or suggests ignorance, both of which reduce credibility and hence commitment.

The communication must explain what is going on and what the intentions are, and it must be conveyed in the ‘language’ of the different stakeholder groups. Delivering it at the appropriate times when it is relevant to the working context of the recipients is also critical if it is to be effective. Also communication is a twoway process; this is sometimes forgotten – and in some of the less successful cases little attention was paid to questions, concerns or feedback which the transformation team (wrongly) felt to be distracting or unimportant.

Culture and values

All the transformations included significant changes in organizational roles, responsibilities, and behaviors, and in many cases the changes were counter to the prevailing culture. The successful transformations recognized this was either desired or inevitable and addressed the organizational issues first to create a new context within which to bring about further changes.

In some cases the transformation also demanded a change in the organization’s values: for example, loss of autonomy and reduced discretion for local investment, consolidation to achieve corporate control of resources or standardization to achieve corporate rather local business advantages. Inevitably these changes created tensions and exposed cultural and value differences across the business units and functions, which had to be either reconciled or over-ridden to succeed with the transformation. In the less successful transformations these tensions were not addressed and existing power structures prevented or subverted the changes.

The structure and mode adopted to bring about the transformation should normally reflect the organization’s overall management style. The ‘task force’ approach, which exercises the use of power, worked well in a situation when the need to transform was urgent, the objectives were very clear and the means of achieving them were known. In the opposite situations, a more devolved approach enabled at least one successful organization to increase the scope and ambition of the transformation, through knowledge sharing across the organization and individual managers learning from experience as the program evolved.

As the transformation proceeds, it may be necessary to change modes and in turn the governance of the program. In particular the creation of a new capability can be carried out by a task force largely separated from day to day operations. However, deploying the new capability usually competes with other business as usual pressures, which can cause unexpected problems, delays or even unsuccessful deployment.

Comments

About John Ward

John Ward is an Integrated Marketing Expert at SAP. He has over 30 years of professional writing experience that includes marketing material, sales support, technical documentation, video scripting, and magazine articles.

Tags:

awareness

The Future Of Supplier Collaboration: 9 Things CPOs Want Their Managers To Know Now

Sundar Kamak

As a sourcing or procurement manager, you may think there’s nothing new about supplier collaboration. Your chief procurement officer (CPO) most likely disagrees.
Forward-thinking CPOs acknowledge the benefit of supplier partnerships. They not only value collaboration, but require a revolution in how their buying organization conducts its business and operations. “Procurement must start looking to suppliers for inspiration and new capability, stop prescribing specifications and start tapping into the expertise of suppliers,” writes David Rae in Procurement Leaders. The CEO expects it of your CPO, and your CPO expects it of you. For sourcing managers, this can be a lot of pressure.

Here are nine things your CPO wants you to know about how supplier collaboration is changing – and why it matters to your company’s future and your own future.

1. The need for supplier collaboration in procurement is greater than ever

Over half (65%) of procurement practitioners say procurement at their company is becoming more collaborative with suppliers, according to The Future of Procurement, Making Collaboration Pay Off, by Oxford Economics. Why? Because the pace of business has increased exponentially, and businesses must be able to respond to new market demands with agility and innovation. In this climate, buyers are relying on suppliers more than ever before. And buyers aren’t collaborating with suppliers merely as providers of materials and goods, but as strategic partners that can help create products that are competitive differentiators.

Supplier collaboration itself isn’t new. What’s new is that it’s taken on a much greater urgency and importance.

2. You’re probably not realizing the full collective power of your supplier relationships

Supplier collaboration has always been a function of maintaining a delicate balance between demand and supply. For the most part, the primary focus of the supplier relationship is ensuring the right materials are available at the right time and location. However, sourcing managers with a narrow focus on delivery are missing out on one of the greatest advantages of forging collaborative supplier partnerships: an opportunity to drive synergies that are otherwise perceived as impossible within the confines of the business. The game-changer is when you drive those synergies with thousands, not hundreds of suppliers. Look at the Apple Store as a prime example of collaboration en masse. Without the apps, the iPhone is just another ordinary phone!

3. Collaboration comes in more than one flavor

Suppliers don’t just collaborate with you to provide a critical component or service. They also work with your engineers to help ensure costs are optimized from the buyer’s perspective as well as the supplier’s side. They may even take over the provisioning of an entire end-to-end solution. Or co-design with your R&D team through joint research and development. These forms of collaboration aren’t new, but they are becoming more common and more critical. And they are becoming more impactful, because once you start extending any of these collaboration models to more and more suppliers, your capabilities as a business increase by orders of magnitude. If one good supplier can enable your company to build its brand, expand its reach, and establish its position as a market leader – imagine what’s possible when you work collaboratively with hundreds or thousands of suppliers.

4. Keeping product sustainability top of mind pays off

Facing increasing demand for sustainable products and production, companies are relying on suppliers to answer this new market requirement.

As a sourcing manager, you may need to go outside your comfort zone to think about new, innovative ways to collaborate for achieving sustainability. Recently, I heard from an acquaintance who is a CPO of a leading services company. His organization is currently collaborating with one of the largest suppliers in the world to adhere to regulatory mandates and consumer demand for “lean and green” lightbulbs. Although this approach was interesting to me, what really struck me was his observation on how this co-innovation with the supplier is spawning cost and resource optimization and the delivery of competitive products. As reported by Andrew Winston in The Harvard Business Review, Target and Walmart partnered to launch the Personal Care Sustainability Summit last year. So even competitors are collaborating with each other and with their suppliers in the name of sustainability.

5. Co-marketing is a win-win

Look at your list of suppliers. Does anyone have a brand that is bigger than your company’s? Believe it or not, almost all of us do. So why not seize the opportunity to raise your and your supplier’s brand profile in the marketplace?

Take Intel, for example. The laptop you’re working on right now may very well have an “Intel inside” sticker on it. That’s co-marketing at work. Consistently ranked as one of the world’s top 100 most valuable brands by Millward Brown Optimor, this largest supplier of microprocessors is world-renowned for its technology and innovation. For many companies that buy supplies from Intel, the decision to co-market is a strategic approach to convey that the product is reliable and provides real value for their computing needs.

6. Suppliers get to choose their customers, too

Increased competition for high-performing suppliers is changing the way procurement operates, say 58% of procurement executives in the Oxford Economics study. Buyers have a responsibility to the supplier – and to their CEO – to be a customer of choice. When the economy is going well, you might be able to dictate the supplier’s goods and services – and sometimes even the service delivery model. When times get tough (and they can very quickly), suppliers will typically reevaluate your organization’s needs to see whether they can continue service in a fiscally responsible manner. To secure suppliers’ attention in favorable and challenging economic conditions, your organization should establish collaborative and mutually productive partnerships with them.

7. Suppliers can help simplify operations

Cost optimization will always be one of your performance metrics; however, that is only one small part of the entire puzzle. What will help your organization get noticed is leveraging the supplier relationship to innovate new and better ways of managing the product line and operating the business while balancing risk and cost optimization. Ask yourself: Which functions are no longer needed? Can they be outsourced to a supplier that can perform them better? What can be automated?

8. Suppliers have a better grasp of your sourcing categories than you do

Understand your category like never before so that your organization can realize the full potential of its supplier investments while delivering products that are consistent and of high quality. How? By leveraging the wisdom of your suppliers. To be blunt: they know more than you do. Tap into that knowledge to gain a solid understanding of the product, market category, suppliers’ capabilities, and shifting dynamics in the industry, If a buyer does not understand these areas deeply, no amount of collaboration will empower a supplier to help your company innovate as well as optimize costs and resources.

9. Remember that there’s something in it for you as well

All of us want to do strategic, impactful work. Sourcing managers with aspirations of becoming CPOs should move beyond writing contracts and pushing PO requests by building strategic procurement skill sets. For example, a working knowledge in analytics allows you to choose suppliers that can shape the market and help a product succeed – and can catch the eye of the senior leadership team.

Sundar Kamak is global vice president of solutions marketing at Ariba, an SAP company.

For more on supplier collaboration, read Making Collaboration Pay Off, part of a series on the Future of Procurement, by Oxford Economics.

Comments

Transform Or Die: What Will You Do In The Digital Economy?

Scott Feldman and Puneet Suppal

By now, most executives are keenly aware that the digital economy can be either an opportunity or a threat. The question is not whether they should engage their business in it. Rather, it’s how to unleash the power of digital technology while maintaining a healthy business, leveraging existing IT investments, and innovating without disrupting themselves.

Yet most of those executives are shying away Businesspeople in a Meeting --- Image by © Monalyn Gracia/Corbisfrom such a challenge. According to a recent study by MIT Sloan and Capgemini, only 15% of CEOs are executing a digital strategy, even though 90% agree that the digital economy will impact their industry. As these businesses ignore this reality, early adopters of digital transformation are achieving 9% higher revenue creation, 26% greater impact on profitability, and 12% more market valuation.

Why aren’t more leaders willing to transform their business and seize the opportunity of our hyperconnected world? The answer is as simple as human nature. Innately, humans are uncomfortable with the notion of change. We even find comfort in stability and predictability. Unfortunately, the digital economy is none of these – it’s fast and always evolving.

Digital transformation is no longer an option – it’s the imperative

At this moment, we are witnessing an explosion of connections, data, and innovations. And even though this hyperconnectivity has changed the game, customers are radically changing the rules – demanding simple, seamless, and personalized experiences at every touch point.

Billions of people are using social and digital communities to provide services, share insights, and engage in commerce. All the while, new channels for engaging with customers are created, and new ways for making better use of resources are emerging. It is these communities that allow companies to not only give customers what they want, but also align efforts across the business network to maximize value potential.

To seize the opportunities ahead, businesses must go beyond sensors, Big Data, analytics, and social media. More important, they need to reinvent themselves in a manner that is compatible with an increasingly digital world and its inhabitants (a.k.a. your consumers).

Here are a few companies that understand the importance of digital transformation – and are reaping the rewards:

  1. Under Armour:  No longer is this widely popular athletic brand just selling shoes and apparel. They are connecting 38 million people on a digital platform. By focusing on this services side of the business, Under Armour is poised to become a lifestyle advisor and health consultant, using his product side as the enabler.
  1. Port of Hamburg: Europe’s second-largest port is keeping carrier trucks and ships productive around the clock. By fusing facility, weather, and traffic conditions with vehicle availability and shipment schedules, the Port increased container handling capacity by 178% without expanding its physical space.
  1. Haier Asia: This top-ranking multinational consumer electronics and home appliances company decided to disrupt itself before someone else did. The company used a two-prong approach to digital transformation to create a service-based model to seize the potential of changing consumer behaviors and accelerate product development. 
  1. Uber: This startup darling is more than just a taxi service. It is transforming how urban logistics operates through a technology trifecta: Big Data, cloud, and mobile.
  1. American Society of Clinical Oncologists (ASCO): Even nonprofits can benefit from digital transformation. ASCO is transforming care for cancer patients worldwide by consolidating patient information with its CancerLinQ. By unlocking knowledge and value from the 97% of cancer patients who are not involved in clinical trials, healthcare providers can drive better, more data-driven decision making and outcomes.

It’s time to take action 

During the SAP Executive Technology Summit at SAP TechEd on October 19–20, an elite group of CIOs, CTOs, and corporate executives will gather to discuss the challenges of digital transformation and how they can solve them. With the freedom of open, candid, and interactive discussions led by SAP Board Members and senior technology leadership, delegates will exchange ideas on how to get on the right path while leveraging their existing technology infrastructure.

Stay tuned for exclusive insights from this invitation-only event in our next blog!
Scott Feldman is Global Head of the SAP HANA Customer Community at SAP. Connect with him on Twitter @sfeldman0.

Puneet Suppal drives Solution Strategy and Adoption (Customer Innovation & IoT) at SAP Labs. Connect with him on Twitter @puneetsuppal.

 

Comments

About Scott Feldman and Puneet Suppal

Scott Feldman is the Head of SAP HANA International Customer Community. Puneet Suppal is the Customer Co-Innovation & Solution Adoption Executive at SAP.

Unlock Your Digital Super Powers: How Digitization Helps Companies Be Live Businesses

Erik Marcade and Fawn Fitter

The Port of Hamburg handles 9 million cargo containers a year, making it one of the world’s busiest container ports. According to the Hamburg Port Authority (HPA), that volume doubled in the last decade, and it’s expected to at least double again in the next decade—but there’s no room to build new roads in the center of Hamburg, one of Germany’s historic cities. The port needed a way to move more freight more efficiently with the physical infrastructure it already has.

sap_Q216_digital_double_feature1_images1The answer, according to an article on ZDNet, was to digitize the processes of managing traffic into, within, and back out of the port. By deploying a combination of sensors, telematics systems, smart algorithms, and cloud data processing, the Port of Hamburg now collects and analyzes a vast amount of data about ship arrivals and delays, parking availability, ground traffic, active roadwork, and more. It generates a continuously updated model of current port conditions, then pushes the results through mobile apps to truck drivers, letting them know exactly when ships are ready to drop off or receive containers and optimizing their routes. According to the HPA, they are now on track to handle 25 million cargo containers a year by 2025 without further congestion or construction, helping shipping companies bring more goods and raw materials in less time to businesses and consumers all across Europe.

In the past, the port could only have solved its problem with backhoes and building permits—which, given the physical constraints, means the problem would have been unsolvable. Today, though, software and sensors are allowing it to improve processes and operations to a previously impossible extent. Big Data analysis, data mining, machine learning, artificial intelligence (AI), and other technologies have finally become sophisticated enough to identify patterns not just in terabytes but in petabytes of data, make decisions accordingly, and learn from the results, all in seconds. These technologies make it possible to digitize all kinds of business processes, helping organizations become more responsive to changing market conditions and more able to customize interactions to individual customer needs. Digitization also streamlines and automates these processes, freeing employees to focus on tasks that require a human touch, like developing innovative strategies or navigating office politics.

In short, digitizing business processes is key to ensuring that the business can deliver relevant, personalized responses to the market in real time. And that, in turn, is the foundation of the Live Business—a business able to coordinate multiple functions in order to respond to and even anticipate customer demand at any moment.

Some industries and organizations are on the verge of discovering how business process digitization can help them go live. Others have already started putting it into action: fine-tuning operations to an unprecedented level across departments and at every point in the supply chain, cutting costs while turbocharging productivity, and spotting trends and making decisions at speeds that can only be called superhuman.

Balancing Insight and Action

sap_Q216_digital_double_feature1_images2Two kinds of algorithms drive process digitization, says Chandran Saravana, senior director of advanced analytics at SAP. Edge algorithms operate at the point where customers or other end users interact directly with a sensor, application, or Internet-enabled device. These algorithms, such as speech or image recognition, focus on simplicity and accuracy. They make decisions based primarily on their ability to interpret input with precision and then deliver a result in real time.

Edge algorithms work in tandem with, and sometimes mature into, server-level algorithms, which report on both the results of data analysis and the analytical process itself. For example, the complex systems that generate credit scores assess how creditworthy an individual is, but they also explain to both the lender and the credit applicant why a score is low or high, what factors went into calculating it, and what an applicant can do to raise the score in the future. These server-based algorithms gather data from edge algorithms, learn from their own results, and become more accurate through continuous feedback. The business can then track the results over time to understand how well the digitized process is performing and how to improve it.

sap_Q216_digital_double_feature1_images5From Data Scarcity to a Glut

To operate in real time, businesses need an accurate data model that compares what’s already known about a situation to what’s happened in similar situations in the past to reach a lightning-fast conclusion about what’s most likely to happen next. The greatest barrier to this level of responsiveness used to be a lack of data, but the exponential growth of data volumes in the last decade has flipped this problem on its head. Today, the big challenge for companies is having too much data and not enough time or power to process it, says Saravana.

Even the smartest human is incapable of gathering all the data about a given situation, never mind considering all the possible outcomes. Nor can a human mind reach conclusions at the speed necessary to drive Live Business. On the other hand, carefully crafted algorithms can process terabytes or even petabytes of data, analyze patterns and detect outliers, arrive at a decision in seconds or less—and even learn from their mistakes (see How to Train Your Algorithm).

How to Train Your Algorithm 

The data that feeds process digitization can’t just simmer.
It needs constant stirring.

Successfully digitizing a business process requires you to build a model of the business process based on existing data. For example, a bank creates a customer record that includes not just the customer’s name, address, and date of birth but also the amount and date of the first deposit, the type of account, and so forth. Over time, as the customer develops a history with the bank and the bank introduces new products and services, customer records expand to include more data. Predictive analytics can then extrapolate from these records to reach conclusions about new customers, such as calculating the likelihood that someone who just opened a money market account with a large balance will apply for a mortgage in the next year.

Germany --- Germany, Lower Bavaria, Man training English Springer Spaniel in grass field --- Image by © Roman M‰rzinger/Westend61/CorbisTo keep data models accurate, you have to have enough data to ensure that your models are complete—that is, that they account for every possible predictable outcome. The model also has to push outlying data and exceptions, which create unpredictable outcomes, to human beings who can address their special circumstances. For example, an algorithm may be able to determine that a delivery will fail to show up as scheduled and can point to the most likely reasons why, but it can only do that based on the data it can access. It may take a human to start the process of locating the misdirected shipment, expediting a replacement, and establishing what went wrong by using business knowledge not yet included in the data model.

Indeed, data models need to be monitored for relevance. Whenever the results of a predictive model start to drift significantly from expectations, it’s time to examine the model to determine whether you need to dump old data that no longer reflects your customer base, add a new product or subtract a defunct one, or include a new variable, such as marital status or length of customer relationship that further refines your results.

It’s also important to remember that data doesn’t need to be perfect—and, in fact, probably shouldn’t be, no matter what you might have heard about the difficulty of starting predictive analytics with lower-quality data. To train an optical character recognition system to recognize and read handwriting in real time, for example, your samples of block printing and cursive writing data stores also have to include a few sloppy scrawls so the system can learn to decode them.

On the other hand, in a fast-changing marketplace, all the products and services in your database need consistent and unchanging references, even though outside the database, names, SKUs, and other identifiers for a single item may vary from one month or one order to the next. Without consistency, your business process model won’t be accurate, nor will the results.

Finally, when you’re using algorithms to generate recommendations to drive your business process, the process needs to include opportunities to test new messages and products against existing successful ones as well as against random offerings, Saravana says. Otherwise, instead of responding to your customers’ needs, your automated system will actually control their choices by presenting them with only a limited group of options drawn from those that have already received the most
positive results.

Any process is only as good as it’s been designed to be. Digitizing business processes doesn’t eliminate the possibility of mistakes and problems; but it does ensure that the mistakes and problems that arise are easy to spot and fix.

From Waste to Gold

Organizations moving to digitize and streamline core processes are even discovering new business opportunities and building new digitized models around them. That’s what happened at Hopper, an airfare prediction app firm in Cambridge, Massachusetts, which discovered in 2013 that it could mine its archives of billions of itineraries to spot historical trends in airfare pricing—data that was previously considered “waste product,” according to Hopper’s chief data scientist, Patrick Surry.

Hopper developed AI algorithms to correlate those past trends with current fares and to predict whether and when the price of any given flight was likely to rise or fall. The results were so accurate that Hopper jettisoned its previous business model. “We check up to 3 billion itineraries live, in real time, each day, then compare them to the last three to four years of historical airfare data,” Surry says. “When consumers ask our smartphone app whether they should buy now or wait, we can tell them, ‘yes, that’s a good deal, buy it now,’ or ‘no, we think that fare is too expensive, we predict it will drop, and we’ll alert you when it does.’ And we can give them that answer in less than one second.”

When consumers ask our smartphone app whether they should buy now or wait, we can tell them, ‘yes, that’s a good deal, buy it now’.

— Patrick Surry, chief data scientist, Hopper

While trying to predict airfare trends is nothing new, Hopper has told TechCrunch that it can not only save users up to 40% on airfares but it can also find them the lowest possible price 95% of the time. Surry says that’s all due to Hopper’s algorithms and data models.

The Hopper app launched on iOS in January 2015 and on Android eight months later. The company also switched in September 2015 from directing customers to external travel agencies to taking bookings directly through the app for a small fee. The Hopper app has already been downloaded to more than 2 million phones worldwide.

Surry predicts that we’ll soon see sophisticated chatbots that can start with vague requests from customers like “I want to go somewhere warm in February for less than $500,” proceed to ask questions that help users narrow their options, and finally book a trip that meets all their desired parameters. Eventually, he says, these chatbots will be able to handle millions of interactions simultaneously, allowing a wide variety of companies to reassign human call center agents to the handling of high-value transactions and exceptions to the rules built into the digitized booking process.

Port of Hamburg Lets the Machines Untangle Complexity

In early 2015, AI experts told Wired magazine that at least another 10 years would pass before a computer could best the top human players at Go, an ancient game that’s exponentially harder than chess. Yet before the end of that same year, Wired also reported that machine learning techniques drove Google’s AlphaGo AI to win four games out of five against one of the world’s top Go players. This feat proves just how good algorithms have become at managing extremely complex situations with multiple interdependent choices, Saravana points out.

The Port of Hamburg, which has digitized traffic management for an estimated 40,000 trucks a day, is a good example. In the past, truck drivers had to show up at the port to check traffic and parking message boards. If they arrived before their ships docked, they had to drive around or park in the neighboring residential area, contributing to congestion and air pollution while they waited to load or unload. Today, the HPA’s smartPORT mobile app tracks individual trucks using telematics. It customizes the information that drivers receive based on location and optimizes truck routes and parking in real time so drivers can make more stops a day with less wasted time and fuel.

The platform that drives the smartPORT app also uses sensor data in other ways: it tracks wind speed and direction and transmits the data to ship pilots so they can navigate in and out of the port more safely. It monitors emissions and their impact on air quality in various locations in order to adjust operations in real time for better control over environmental impact. It automatically activates streetlights for vehicle and pedestrian traffic, then switches them off again to save energy when the road is empty. This ability to coordinate and optimize multiple business functions on the fly makes the Port of Hamburg a textbook example of a Live Business.

Digitization Is Not Bounded by Industry

Other retail and B2B businesses of all types will inevitably join the Port of Hamburg in further digitizing processes, both in predictable ways and in those we can only begin to imagine.

sap_Q216_digital_double_feature1_images4Customer service, for example, is likely to be in the vanguard. Automated systems already feed information about customers to online and phone-based service representatives in real time, generate cross-selling and upselling opportunities based on past transactions, and answer customers’ frequently asked questions. Saravana foresees these systems becoming even more sophisticated, powered by AI algorithms that are virtually indistinguishable from human customer service agents in their ability to handle complex live interactions in real time.

In manufacturing and IT, Sven Bauszus, global vice president and general manager for predictive analytics at SAP, forecasts that sensors and predictive analysis will further automate the process of scheduling and performing maintenance, such as monitoring equipment for signs of failure in real time, predicting when parts or entire machines will need replacement, and even ordering replacements preemptively. Similarly, combining AI, sensors, data mining, and other technologies will enable factories to optimize workforce assignments in real time based on past trends, current orders, and changing market conditions.

Public health will be able to go live with technology that spots outbreaks of infectious disease, determines where medical professionals and support personnel are needed most and how many to send, and helps ensure that they arrive quickly with the right medication and equipment to treat patients and eradicate the root cause. It will also make it easier to track communicable illnesses, find people who are symptomatic, and recommend approaches to controlling the spread of the illness, Bauszus says.

He also predicts that the insurance industry, which has already begun to digitize its claims-handling processes, will refine its ability to sort through more claims in less time with greater accuracy and higher customer satisfaction. Algorithms will be better and faster at flagging claims that have a high probability of being fraudulent and then pushing them to claims inspectors for investigation. Simultaneously, the same technology will be able to identify and resolve valid claims in real time, possibly even cutting a check or depositing money directly into the insured person’s bank account within minutes.

Financial services firms will be able to apply machine learning, data mining, and AI to accelerate the process of rating borrowers’ credit and detecting fraud. Instead of filling out a detailed application, consumers might be able to get on-the-spot approval for a credit card or loan after inputting only enough information to be identified. Similarly, banks will be able to alert customers to suspicious transactions by text message or phone call—not within a day or an hour, as is common now, but in a minute or less.

Pitfalls and Possibilities

As intelligent as business processes can be programmed to be, there will always be a point beyond which they have to be supervised. Indeed, Saravana forecasts increasing regulation around when business processes can and can’t be digitized. Especially in areas involving data security, physical security, and health and safety, it’s one thing to allow machines to parse data and arrive at decisions to drive a critical business process, but it’s another thing entirely to allow them to act on those decisions without human oversight.

Automated, impersonal decision making is fine for supply chain automation, demand forecasting, inventory management, and other processes that need faster-than-human response times. In human-facing interactions, though, Saravana insists that it’s still best to digitize the part of the process that generates decisions, but leave it to a human to finalize the decision and decide how to put it into action.

“Any time the interaction is machine-to-machine, you don’t need a human to slow the process down,” he says. “But when the interaction involves a person, it’s much more tricky, because people have preferences, tastes, the ability to try something different, the ability to get fatigued—people are only statistically predictable.”

For example, technology has made it entirely possible to build a corporate security system that can gather information from cameras, sensors, voice recognition technology, and other IP-enabled devices. The system can then feed that information in a steady stream to an algorithm designed to identify potentially suspicious activity and act in real time to prevent or stop it while alerting the authorities. But what happens when an executive stays in the office unusually late to work on a presentation and the security system misidentifies her as an unauthorized intruder? What if the algorithm decides to lock the emergency exits, shut down the executive’s network access, or disable her with a Taser instead of simply sending an alert to the head of security asking what to do while waiting for the police to come?

sap_Q216_digital_double_feature1_images6The Risk Is Doing Nothing

The greater, if less dramatic, risk associated with digitizing business processes is simply failing to pursue it. It’s true that taking advantage of new digital technologies can be costly in the short term. There’s no question that companies have to invest in hardware, software, and qualified staff in order to prepare enormous data volumes for storage and analysis. They also have to implement new data sources such as sensors or Internet-connected devices, develop data models, and create and test algorithms to drive business processes that are currently analog. But as with any new technology, Saravana advises, it’s better to start small with a key use case, rack up a quick win with high ROI, and expand gradually than to drag your heels out of a failure to grasp the long-term potential.

The economy is digitizing rapidly, but not evenly. According to the McKinsey Global Institute’s December 2015 Digital America report, “The race to keep up with technology and put it to the most effective business use is producing digital ‘haves’ and ‘have-mores’—and the large, persistent gap between them is becoming a decisive factor in competition across the economy.” Companies that want to be among the have-mores need to commit to Live Business today. Failing to explore it now will put them on the wrong side of the gap and, in the long run, rack up a high price tag in unrealized efficiencies and missed opportunities. D!

Comments

Erik Marcade

About Erik Marcade

Erik Marcade is vice president of Advanced Analytics Products at SAP.

Tags:

The New Digital Healthcare Patient Experience

Martin Kopp

Digitized healthcare has arrived. And it is only going to get better. Since the 1950s, information technology has had a growing influence on the healthcare industry. And today, more than three-quarters of all patients expect to use digital services in the future. That is, if they are not using them already. Healthcare consumers have become more informed and proactive.

Today, a pregnant woman can schedule a gynecology appointment electronically. Her insurance company probably offers a smartphone app to monitor her health. She can download the app and self-register. The app documents her ongoing health as she updates the profile data. And because her data is stored in the cloud, her gynecologist has immediate access to it.

These are a few examples of the important trends shaping the patient experience with digital innovation. The latest digital solutions are bringing the patient and the healthcare industry closer together. And this digital connectivity means more personalized patient care.

Digital technology is changing the role of the patient. Patients are better informed and more involved in their own health decisions. With greater access to information, they can sometimes self-diagnose certain health issues. Due to digitization, they have better communication with healthcare providers and easier access to their own test results.

Monitoring illness

Healthcare providers are better equipped to gather and analyze data. So, healthcare outcomes are faster and easier to realize. Providers can react earlier to conditions. And they can even sometimes predict medical conditions before any symptoms appear. Therapies are transforming to a more user-centric design. This is all possible because digital networking of data informs caregivers earlier and keeps them informed. We have moved past the patient’s chart as the most important source of information.

Improving wellness

The ability to predict medical conditions gives providers a tool to promote wellness. This is changing the healthcare value chain. Remote monitoring is possible, making trips to the clinic or doctor’s office less necessary. Wearable monitoring devices have changed the medial landscape. And the use of wearable devices is expected to grow. According to the McKinsey Global Institute (MGI), 1.3 billion people will be using fitness trackers by the year 2025. In some regions, this will account for up to 56% of the population. The millennial generation sums up the benefits in a word: convenience.

The blending of physical and digital realms into a common reality is referred to as the Internet of Things (IoT). The IoT makes many things possible that were only dreamed of a few years ago. It extends the reach of information technology. From remote locations, we can electronically monitor and control things in the physical world. Basically, it is the digitizing of the physical world.

With the IoT, MGI predicts a savings in healthcare treatment costs of up to $470 billion per year by 2025. But even more important is the improvement in healthcare. In addition to driving down treatment costs, this will extend healthy life spans and improve the quality of life for millions of people. And it will improve access to healthcare for those who are underserved in the present system. Plus, this extensive use of fitness tracking devices will create a multi-billion dollar industry.

Re-shaping the patient experience

The patients of today and tomorrow have more information and more options than ever before. Patients are already seeing increased value from the Big Data that healthcare professionals now have access to. Patients are more engaged in their own care. We are entering an age of personalized healthcare based on far-reaching knowledge bases.

Because of digital innovation, healthcare consumers can more easily seek relief when they are sick. They can be more involved in disease prevention and self-supported care. With patient-owned medical devices, they are connected to the Big Data of cloud computing. This cloud-based information provides proven treatments and outcomes for specific conditions.

Value chain improvements

The digital value network connects all aspects of the healthcare ecosystem in real time. This connectivity drives better healthcare outcomes that are specifically relevant to the patient. Digital innovation in healthcare improves interactions to provide personalized care based on Big Data. In that respect, you can think of it as Big Medicine for the little guy. A massive database gives healthcare providers a 360-degree view of the patient. Data is stored in the cloud and processed in the core platform.

Services and functions that this efficient system provides include medication reminders for patients. It tracks your health for you, your family, and friends. Remote home monitoring and emergency detection offer an increased level of safety and protection. Remote diagnostics can mean you stay at home instead of being hospitalized. Prediction of organ or other physical failures before they happen can save lives.

SAP software provides a single platform that brings together healthcare providers, patients, and value-added services. It offers a seamless digitization of the entire patient experience. And it provides results in real time, available to all parts of the healthcare ecosystem. This broad connectivity creates an omni-channel, end-to-end patient experience.

To learn more about Digital Transformation for Healthcare, click here.

Comments

Martin Kopp

About Martin Kopp

Martin Kopp is the global general manager for Healthcare at SAP. He is responsible for setting the strategy and articulating the vision and direction of SAP's healthcare-provider industry solutions, influencing product development, and fostering executive level relationships key customers, IT influencers, partners, analysts, and media.