High-end professional services firms that cater to corporate clients have a clear upside: Because they provide specialized expertise, their offerings can be very lucrative. But there’s a less obvious downside: If a consulting firm, say, or a law practice wants to double its revenue, it has to double its staff of consultants or attorneys. Consultancies, law firms, ad agencies, and other professional services firms struggle to nudge their gross margins above 40% as they achieve scale. Contrast that with product companies like Google and Adobe, which don’t have to deal with the same cost structure and which enjoy gross margins of 60% to 90%.
Technology offers professional services firms a way out of their predicament. By leveraging the power of algorithm-driven automation and data analytics to “productize” aspects of their work, a number of innovative firms are finding that, like Google and Adobe, they can increase margins as they grow, while giving clients better service at prices that competitors can’t match. Productivity rises, efficiencies increase, and nonlinear scale becomes feasible as productized services take over high-volume tasks and aid judgment-driven processes. That frees up well-paid professionals to focus on jobs that require more sophistication—and generate greater value for the company.
There are distinct challenges, however, in developing products to embed in services. The nature of a product and its role in a business’s value proposition are not the same for a services firm as they are for a company that manufactures goods. This means that services firms must take a different approach to creating, managing, and monetizing products.
In the following pages I present a guide to product development for professional services firms. I describe the three key stages of the process: discovering potential products by identifying opportunities for automation; developing the products and enabling them to process, analyze, and learn from data; and monetizing them by building a revenue model that captures benefits from automation and the application of analytics.
Embedded Products in Service Offerings
In a professional services firm, a product is created when some aspect of a service is automated, infused with analytics, and monetized differently. This involves systematizing the service, leveraging data to improve it automatically, and then changing the method of payment for the resulting improvements.
The product, therefore, is embedded in the service offering and sold as an element of it. Services remain the center of gravity, and customers continue to buy the service offering, not the product per se. From the customer’s perspective, little changes other than the pricing of the service. That drops because the value created by the new product is shared between the firm and its customers.
As an illustration of a service provider with embedded products, consider Littler, a global employment and labor law practice. Littler does legal work for companies in more than a dozen countries. To improve the quality and efficiency of its services, it has “unbundled” the tasks involved in their delivery and assigned them either to people with specialized knowledge or to products with automation and analytics capabilities, depending on the level of sophistication involved. Essentially, the firm has reengineered its legal services by developing offerings that are powered by technology and humans.
One example is Littler CaseSmart–Charges. This offering helps HR professionals and in-house attorneys better manage employee discrimination claims and complaints by combining software, project management tools, and the skills of flextime attorneys (FTAs) and data analysts. FTAs focus on specific tasks in the litigation process and have deep subject-matter expertise, which makes them highly efficient and effective at performing particular services. (They also work out of home offices on a flexible schedule, which reduces the company’s overhead.) Data analysts, meanwhile, focus on reviewing, interpreting, and translating data on behalf of lawyers and work at a lower price point.
Littler uses a dashboard that enables clients to track discrimination charges filed with the Equal Employment Opportunity Commission. The dashboard provides data-driven insights to proactively address business risks, which in turn lowers legal costs and speeds up the process of managing pending cases. In some instances, this can help prevent the cases from escalating to litigation.
Similarly, Littler CaseSmart–Litigation provides a streamlined method for HR clients to manage the litigation process in cases where they are being sued by individual plaintiffs. A dashboard interface provides insights on employment issues while tracking the progress of legal cases, and that technology is coupled with attorney services. Again, the offering improves the speed and quality of Littler’s work while lowering costs for both Littler and the client. It also allows clients to look across their portfolios of litigation and identify recurring factors that may be contributing to those cases (for example, they can determine whether there’s a pattern involving a particular jurisdiction, decision maker, or policy and then proactively manage that issue).
To share the benefits of these innovations, Littler has entered into alternative fee arrangements (AFAs) with clients that save them money while boosting the firm’s revenue. Instead of billing for the hours its attorneys spend on claims, Littler uses a fixed-fee model in which charges are based on productivity (per grievance or complaint). This change has resulted in lower legal costs for clients—they’ve reported drops ranging from 10% to 35%—which has enabled the CaseSmart team to win new business. Revenue doubled from 2014 to 2015, and in spring 2016, Legaltech News heralded Littler as a Client Service Innovator of the Year, and BTI Consulting Group named it one of the 22 law firms that were best at AFAs.
Whereas product manufacturers’ ideas for new offerings are driven by an external focus on customer needs, professional services firms identify product opportunities inside their businesses. They’re looking not for unmet needs but for untapped potential to automate the services they’re already delivering successfully.
Consider EXL, an operations management and analytics company I advised as a board member for a decade. One service that EXL provides to its health insurance clients is medical claims management, specifically as it relates to overpayment caused by fraud or abuse. Years ago that service was manual: EXL employees would examine medical claims for incorrect coding, subrogation, payment errors, nonbeneficial services, and other causes of overpayment. They’d investigate claims that seemed questionable and then focus on recovering undue outlays.
After processing millions of claims, EXL began to recognize patterns in the circumstances that surrounded instances of overpayment. It discovered that certain procedure codes, diagnosis codes, providers, patients, locations, and other variables were systematically associated with fraudulent or erroneous activity. With those insights, EXL was able to develop a tool that could scan and analyze claims for the relevant attributes. Each claim earned a score that predicted the likelihood of abuse or fraud, and the ones flagged as suspect went up for review.
By productizing this service, EXL was able to significantly increase the number of claims it processed, reduce the costs of handling them, increase the amount of money recovered, and prevent overpayment on new claims. In fact, for one client, EXL’s payment integrity tool recovered $50 million in three years and prevented an estimated $20 million in further overpayments.
Once you’ve identified patterns in your services, you’ll want to evaluate which tasks are best suited for productization via automation. To do this, you need to sort them according to two variables: the frequency with which they’re performed and the level of sophistication (meaning knowledge or intelligence) required to perform them. (A high-sophistication task in an advertising agency, for example, might involve developing creative assets for a new marketing campaign. A low-sophistication task might involve optimizing search engine marketing for a brand.)
The tasks that meet two criteria—they’re performed frequently and they require little sophistication—are the low-hanging fruit for productization. That’s because the algorithms that drive automation are very good at performing high-volume, repetitive tasks. Volume is also important for improving the algorithm over time; the more input the algorithm receives, the more it will learn and the better it will perform.
To get a better sense of opportunities that fall into this category, consider this analogy: When you drive long distances on the highway, you repeatedly perform certain tasks that require very little intelligence, such as maintaining a steady speed and keeping an eye on the lanes to your left and right. These high-volume, low-skill tasks are ideal for automation—and, in fact, the technology already exists (think cruise control and blind-spot monitors).
By contrast, low-volume tasks don’t provide enough data on which to base automation, while high-sophistication tasks are not easily automated because they require strategic decision making. For professional services companies, these opportunities simply aren’t worth the investment.
Professional services firms have the advantage of already knowing what they’re marketing and whom it’s for. These companies aren’t creating something out of nothing; they’re converting something (a service) into something else (a service with embedded products).
This changes the process of developing and improving an offering in profound ways. In early-stage development, a product company will design various prototypes and try them out on sample customers, with a view to determining the key components in a value proposition. Smart professional services firms, however, aren’t trying to identify desired features. Instead, they use prototypes merely as a foundation on which to build precision, sophistication, and complexity. These improvements are typically driven by the ability of the product to gather and analyze data automatically, thus harnessing technology to create a “smart” product that improves itself.
High-volume, low-skill tasks are ideal for automation.
Deloitte, a leading audit, consulting, tax, and advisory services firm, provides a good example. Its Argus tool makes use of machine-learning techniques and natural-language processing to analyze electronic documents for auditing purposes. Argus can “learn” from every interaction it has with humans and every document it processes, so it gets better at identifying and extracting key accounting information over time. Within a few months of its release, Argus had already been used by more than 1,000 auditors to analyze more than 30,000 documents.
When products that are embedded in a service are basically software, improvements are more frequent than with stand-alone products, whose improvement usually involves the launch of a new generation or model. As I’ve pointed out, the tool is always learning from and adapting to its users, and it’s arguably misleading to draw strict boundaries between prototypes, finished products, and generations of finished products.
These incremental product improvements have broader business implications. As the basic functionality of a product grows more sophisticated, the enabling technology can be expanded to other uses. For example, Deloitte is now applying the platform behind Argus to its consulting business.
Note, however, that embedded products do not replace service offerings; instead they strengthen the value proposition that service offerings present. Argus amplifies Deloitte’s auditing services but does not serve as a substitute for them. For instance, if a client requested the development of a maturity model for cybersecurity readiness, an auditor would need to have strategic discussions with the company to devise guidelines, policies, and tools. That’s because such work involves complex analysis and decision making that exceed the capabilities of an embedded product like Argus.
For similar reasons, self-service products (such as the basic legal and accounting tools offered by LegalZoom and TurboTax) are rare in the context of high-touch professional services firms. Specialized knowledge, strategic thinking, and sophisticated decision making are integral to the delivery of high-value services, so people at those firms must play a bigger role than products do. It’s also preferable for professional services firms to do some hand-holding with clients, because that’s how they usually make their money. And it’s usually best to keep products on company premises, where they can remain proprietary and protected as a source of competitive advantage.
A professional services firm may sometimes find it advantageous to turn a tool into a stand-alone product and then spin it off and sell it. However, after creating such a product, the company will almost always return to the business of providing a service. This observation brings us to the final stage in the product-creation process.
For an embedded product to be worth developing, you have to figure out how to capture its value. If your firm’s services have become more efficient or effective, it doesn’t make sense to continue with a pricing model that’s based on time and materials. Indeed, if the goal behind productizing services is to push beyond a linear growth rate, you must change your monetization model—or risk getting paid less for your work.
Two monetization levers—transaction-based pricing and outcome-based pricing—correspond to the productivity gains and intelligence gains that automation and analytics respectively deliver. Once your company adds automation to a service, it must shift to transaction-based pricing to capitalize on the increased quantity of the offering (because automation improves productivity). And once your company adds analytics to a service, it must shift to outcome-based pricing to capitalize on the increased quality of the offering (because analytics enables smarter decision making). In other words, this is a sequential process in which you transition from getting paid for inputs (time and materials) to getting paid for throughputs (transactions) to getting paid for outputs (outcomes). Note that this progression requires both time and trust. You need maturity and experience with the process to establish the correct pricing structure at each of these stages. And you need to build trust with your clients before attempting to convert them to a new pricing model. In practice, this process can take several years.
When segueing from billable hours to transaction-based pricing, it’s important to do your math. Consider your revenue under a time-and-materials-based model, calculate how your costs and margins will change as a result of automation, and adjust your fees accordingly. Doing these computations will prevent you from pricing your service too high and creating a dissatisfied client, or going too low and ending up with subpar margins.
Here’s an example: Let’s say that your company reviews legal agreements at a rate of $200 an hour and each agreement takes about 10 hours, resulting in a fee of $2,000 per agreement. Now suppose you automate that process so it takes only two hours per agreement, which translates to a fivefold productivity gain. Since your client won’t be happy about paying an hourly rate that’s five times higher ($1,000), a better approach is to propose a per-agreement price with a discount thrown in for good measure. Thus you might charge the client $3,500 for two contract reviews—less than the previous cost of $4,000. Your client will be pleased with the reduced fee, and you’ll both come out ahead.
Reaping the monetary value of analytics, however, requires moving from transaction-based pricing to outcome-based pricing. Consider this example from EXL: While managing collection calls for a utility company, EXL developed an algorithm that scored each delinquent customer on the likelihood that he or she would pay the bill following a phone call. EXL used that information to prioritize the calls to make, and the efficiency of the collections process increased dramatically as a result. However, to be compensated for that increased value, EXL would have to get paid for results it delivered (recovering money from overdue bills) rather than for each transaction (each call). EXL is investigating that model for the future.
Pricing outcomes is more difficult than pricing transactions, because it requires qualitative judgments as well as quantitative assessments. A professional services firm has to figure out how to define value, measure it, and attribute value creation to the proper source. To negotiate outcome-based contracts with clients, therefore, you may need relatively high-level salespeople or product specialists with consulting or creative strengths. In addition, you may have to elevate the conversation to top decision makers at the client company, because the negotiation may be too strategic to be left to employees accustomed to buying your services on a time-and-materials basis. Finally, it pays to pilot your product and new billing model with customers with whom you’ve built a trusted relationship and who are prepared to participate in the experiment. Make sure they understand that you’ll be rolling out the product and new billing model with other clients.
Changing the structure of your contractual agreements may influence the types of clients you pursue in the future. For example, you may want to focus on companies that have highly repeatable problems. Or you may decide to concentrate on opportunities where you can clearly measure and determine the source of the strategic value you’ve created. This is one of the reasons that EXL primarily provides collection services to companies, as opposed to, say, helping companies improve their customer satisfaction rates. It’s a lot easier to measure efficiencies or effectiveness generated by the former.
People and Processes
Successfully developing products to embed in a service requires more than just a sound process. A firm’s culture and people’s mindsets have to change. So does the organizational structure. Here are three things that are necessary for success:
A unit dedicated to product development.
In the same way that product companies build innovation units to incubate ideas, services companies should set up teams devoted to developing products internally. It’s important to make such a team somewhat autonomous; it needs its own budget, people, goals, and metrics. But keep it connected to the business units, since that’s where product ideas will arise. Create a two-way exchange in which business units can come to the product team with ideas—and vice versa—while you empower the team to incubate those ideas.
A cross-functional approach.
The product development team should include people with expertise in three areas: the business domain, IT, and pricing. You need domain experts to provide firsthand knowledge about clients, work processes, and business patterns. You need IT experts to add automation and intelligence to your services and ensure that the product can integrate with existing systems. And you need business analysts who can appropriately price your services.
A different dashboard.
The client-facing units within services firms have a tendency to examine and evaluate their performance and budgets almost daily. Product-management organizations can’t work this way, and it’s important to get the organization to really value long-term goals, because the benefits of product-enabled services may take time to blossom. To measure an embedded product’s performance, therefore, professional services companies have to change how they define success. Instead of focusing on classic service-based metrics (such as client satisfaction or process efficiency), use product-based metrics (such as ideas generated, prototypes created, or level of automation achieved).
Beyond the organizational changes, all services firms contemplating embedding products in their offering need to recognize that doing so comes with a hefty price tag. Accepting this reality can be uncomfortable. Although product companies understand that costs come long before sales, and although entrepreneurs can rely on funding from venture capitalists with that same understanding, investing ahead of revenues is an alien concept for firms that provide services. It’s important to accept that you have to spend money without knowing exactly how you’re going to get paid.
Productization is also a source of fear for many employees. The flip side to the benefits of intelligent automation is that firms will need fewer people to manage a process. So when robots take over manual tasks, companies generally move to a model in which they offer fewer but more-demanding jobs. Employees with the best skills and knowledge will keep their jobs, while those tied to repetitive manual tasks will find themselves at risk. In theory, you could even remove people altogether.
It’s therefore easy to conclude that intelligent automation pits humans against robots. I’d argue that’s not the case. Algorithms are created and improved by humans, and technology is nothing without people to guide it. Thus the future workplace will not be about you versus robot; it will be about youand robot. It’s also worth noting that intelligent automation will ultimately leave employees with more-meaningful jobs and companies with more-profitable business models.
The world of professional services stands ready to be transformed by analytics and automation. That’s good news for services firms; they can leverage the power of embedded products to break free from the linear-growth trap. But there’s another, perhaps more pressing, reason why they should put products into their offerings: Customers are increasingly demanding it. By following the steps outlined in this article, professional services firms can increase their profitability and gain an advantage over their competitors.
Mohanbir Sawhney is the McCormick Foundation Chair of Technology at the Kellogg School of Management, where he also directs the Center for Research in Technology & Innovation.
IMAGE CREDITS: http://s1.ibtimes.com/