As part of our efforts to profile the rapidly evolving service delivery automation (SDA) landscape, I am speaking with the leaders of many of the technology players who are helping stimulate innovation in this space. This second of a series of blogs on SDA technologies, is based on observations and learnings from a recent briefing with Hans Christian (Chris) Boos, CEO of Arago.
The company was founded in 1995 but its intelligent automation software for enterprise IT, in its current form, became generally available only 2-3 years ago. Arago has since experienced rapid growth, more than trebling revenue since 2011.
Arago’s flagship product is AutoPilot. This uses an inference engine with, what essentially sounds like, a neural network to speed up processing, although the term was not used by Boos during the briefing. Instead, he refers to human brain like activity to learn and apply learning (knowledge items) to new or changing environments to infer how to process requirements automatically. The machine gets more useful the more knowledge it gains but it also has to manage the knowledge, for example, deal with rules that contradict each other. It does this in a mathematical way and uses analytics. According to Boos, it can apply this approach to different areas such as database management, incident management and also to more architectural processes and business logic.
Arago figures show that AutoPilot processed nearly 2 million tickets (as produced by infrastructure management tools such as BMC) for clients in 2013. Circa 87% of these were fully automated. Processes automated at the middleware layer, AutoPilot’s sweet spot, had the highest level of automation at 98%.
Clients are typically large organizations or IT service providers. These include two major global IT service providers.
The software is available as a service, as well as on premise but interestingly the majority of clients want it on premise.
Two licensing models are available from Arago:
AutoPilot comes with built connectivity to infrastructure management tools such as BMC and IBM Tivoli and with APIs for integration with other packages.
Arago’s proposition comes with an estimated cost saving of between 30% and 50%.
If AutoPilot can successfully tap into its acquired knowledge to handle non-standard environments or changing conditions, then it could minimize the need for predefined scripts, to automate parts of IT that are more challenging to automate. I believe this can complement other automation tools that are highly scripted and which are used in other parts of IT infrastructure. The potential benefits in large and highly heterogeneous IT environments, could soon accumulate.
This is advanced technology and could also increase complexity, potentially leading to tickets itself, at least initially while the knowledge-base is being developed.
In terms of Arago’s target market, the company is selling to a converted crowd – IT service providers and IT departments of large organizations that have automated parts of their IT infrastructure already. Its challenge is its size which is not big enough for the demand that it is seeing. Arago is enhancing its partnership network. It is also expanding geographically. At the moment Arago operates out of Germany with all its 92 staff currently based there. It is looking to open an office in the United States soon but it has no physical presence in other countries such as India.
Other measures include creating a community where clients can share automations/knowledge items for free or buy or sell them.
These plans will start to pay off but for now demand is likely to remain choked by lack of scale, I believe, particularly, in initial consultancy and client training services.
AutoPilot is still a relatively new product and I expect some functionality enhancements to be on the cards. More work on the UI is already underway.
Growth opportunities include selling to smaller companies. Arago has released a community edition that can give smaller organizations a fully functioning AutoPilot that is only limited in the size of the IT that it can automate. This is a clever bit of marketing that prepares the ground for attracting large companies of the future.
Arago’s core technology is application agnostic. The company chose to apply it to IT first but the core product can also learn to handle business logic, potentially leaving Arago with opportunities to expand into business process automation in the future.
As a part of our efforts to profile the rapidly evolving service delivery automation landscape, I am speaking with the leaders of many of the technology players who are helping stimulate innovation in this space. It is an exciting time for automation and the following observations and learnings come from one of my recent conversations, a briefing with Andrew Anderson, founder and CEO of Celaton. Stay tuned to learn more as I speak with other leading players.
UK-based Celaton was born out of the management buyout of Redrock software from Netstore plc and the acquisition of DG Tech in 2004. Today it has revenue of circa £2.5m and the same amount of investment by Business Growth Fund to enhance its sales and marketing capabilities.
Celaton’s artificial intelligence software, inSTREAM is designed to handle labor intensive administrative tasks. It takes unstructured content, such as correspondence, complaints, letters, faxes, e-mails, and attachments, learns to understand the content and context and then processes the information. inSTREAM is a self-learning system. When it is first deployed it will need human guidance on what to do. It learns from experience. The more it does the faster it becomes as it learns how to handle different requirements according to the rules and knowledge that it accumulates.
inSTREAM reads unstructured content and applies rules to it to identify and understand key information such as context, sentiment, importance and urgency. It then structures the content and feeds it into the appropriate line of business application (LOB) for processing. The objective is to get guaranteed perfect structured data that can be fed into a line of business system such as ERP, CRM, and workflow, so that the data can continue along its corporate journey. In many instances, the data that goes into the LOB system has to be connected to its source e.g. route to source in the insurance industry where there is the need to go back to the original document for audit purposes. InSTREAM retains the data that it has processed including the original source document. It delivers the data to the LOB system and the document to a document management system.
InSTREAM is a non-invasive system. Integration is done via web services or the data can be delivered to a holding area for the target system to pick up.
inSTREAM runs on Microsoft platforms utilizing SQL Server, Internet Information Server and .NET technologies. It can process all types of incoming documents, and it is platform agnostic. It is provided on a hosted basis. Subscription rates are based on volumes, complexity of the processes in question, and the levels of benefits that it is expected to generate. Pricing starts from £1500 per month and can go up to £60k or more per month.
Celaton’s typical customers are retail, travel and insurance companies. Benefits are realized through increased productivity and improved customer management. One client, a UK loss adjuster, has reportedly reduced its head count by 85% while managing a fivefold increase peak in demand in insurance claims.
Celaton has a reselling partnership with Agilisys, the UK technology and outsourcing services company, was the first to sign up. Agilisys Automate, based on inSTREAM, is targeted at UK local government sector and has its first customer, a London borough council, signed up already. There are more council deals in the pipeline.
Celaton is carving a niche for itself in textual and document processing automation. It is in the right place at the right time to grow with rising demand in the market.
My take on the company’s proposition, benefits and challenges:
Competition: There are not many competitors in this field with AI-based standalone tools, but some capabilities are on offer as part of other offerings. Examples include Oracle RightNow Email Management Cloud Service and its Email Management which is integrated with a self-learning knowledge base and across customer interaction channels. Optimized for smartphones and mobile web devices, this receives enquiries via email and web forms and automates responses. New entrants to the market are highly likely with at least one new product on its way – a new cognitive engine from a well-known IT services automation company.
Marketing: The cost advantage of automation can clearly be significant but there are challenges too. Celaton has to overcome buyer uncertainty about machines doing the job of an employee in a service-line, such as in-bound document management, which has to deal with highly unstructured content. A Good marketing of a few success stories could work wonders. Anderson is doing a good job of telling the Celaton story and the company also has a substantial investment by BGF to orchestrate a robust marketing campaign.
Go-to-market and Scaling Up: I believe one way for Celaton to find more willing clients is to target companies that have outsourced their in-bound document handling and who are looking beyond labor arbitrage and offshoring to build on efficiencies. Celaton also has to look for ways that it can scale up to respond to demand. At the moment it is the only company that can fully configure, deploy and host inSTREAM. Agilisys is coming up to speed but more deployment partners are needed to meet the two objectives of reaching the right client segment and gaining scale.
Outsourcing service providers are also looking for new ways to increase their cost competitiveness, but they need to think about alternative pricing to the FTE-based model.
Service providers will be investigating partnership opportunities with Celaton and other automation technology providers, such as Blue Prism, and asking themselves the classic question about timing new investment. Do they invest in business process automation today or wait for the opportunities to come before spending on new capabilities? Agilisys has gone halfway – with Agilisys Automate, it is focusing on technology sales for now and gaining new skills. Other service providers, such as Sutherland Global Services and Capita have already invested in automation (e.g. Blue Prism). For the undecided, there are lessons from the journey of analytics into the business process services market. What started as added value is now being built into specific offerings by some leading vendors. A similar approach to automation could lead to a significant competitive edge through automation.
Check back for more of our views on technology players in service delivery automation.
The Global In-house Center (GIC, formerly referred to as captives) market was once thriving with unprecedented statistics – 97 new GIC set-ups in 2009, 105 in 2010, and 103 in 2011. Then there was a dip, with only 75 new centers in 2012, and 69 in 2013. This, coupled with numerous acquisitions of GICs by service providers, (e.g., KBC Group’s financial arm by Cognizant, Bayer’s Indian IT operations by Capgemini, and Hutchison Whampoa’s India-based call center operations by Tech Mahindra), is likely to raise questions and concerns about the future of the in-house model.
Let us look at the ground realities of the GIC model’s growth and evolution:
Further, while buyers’ moves from an outsourced to an in-house model rarely receive considerable fanfare, they do paint a picture of the health of the GIC model. For example, HP had been General Motor’s main IT vendor per a US$2 billion contract awarded in 2010, but in 2012 the automaker decided to insource a huge amount of its services as part of its new strategy, leaving HP with only a few. AstraZeneca plans to reduce its outsourcing work, which is currently spread across multiple Indian software service providers. BT plans to have more control of its processes by taking back its outsourcing contracts from service providers, and increasing its capacity in existing shared services centers in India and Malaysia.
The bottom line is that while GIC set-up growth may be slowing, the model continues to be an integral component of organizations’ sourcing strategy. Firms continue to leverage both sourcing models (service providers and GICs) based on best fit with their sourcing needs, cost and value objectives, and services demand profile.
For more insights on the GIC model landscape, please refer to our recently released report “Global In-house Center (GIC) Landscape Annual Report 2013.” The report provides a deep-dive into the GIC landscape and a year-on-year analysis of the GIC trends in 2013, comparing them with trends in the last two years. The research also delivers key insights into the GIC market across locations, verticals, and functions, and concludes with an assessment of strategic priorities for GICs.
Leading bank – “We have grown our sourcing footprint to over 20,000 FTEs across multiple locations in the last 15 years. We definitely have a mature global sourcing model.”
Leading insurer – “We have evolved in the way we source global services, from transactional work to highly complex and niche services. Ours is a highly mature global sourcing model.”
Leading retailer – “We started our global sourcing journey only five years back by outsourcing our IT work. We are still struggling to define if there is a GIC play needed in our global sourcing model, which most of our peers have adopted. I think our global sourcing strategy is still in its nascent stages.”
Everest Group hears statements similar to the above on a regular basis from many of our enterprise/buyer clients. That’s because most organizations measure the maturity of their global sourcing model in terms of a few simple metrics, like scale, age, type of work, and models adopted.
While FTE scale is often the most commonly used yardstick to measure the global sourcing maturity, in our experience it’s the most misleading one. Why? Most organizations fail to understand global sourcing components beyond scale, scope, and model (GIC vs. third-party.) But we believe global sourcing is a much more complex play of multiple factors, and that an organization’s approach toward each of them helps determine its overall maturity.
The following diagram depicts the factors that are important to assessing the sophistication and maturity of global sourcing model of enterprises.
First, scale. Yes it is an important aspect, and to a large extent indicates the level of investments and commitment to the model. However, on an absolute basis, scale alone can be misleading. When the same number is viewed as a percentage of headcount in back-office/middle-office functions in an enterprise, it represents the true penetration of global sourcing model. For example, two organizations with similar global sourcing FTE scale can have different (%) values for global sourcing penetration.
Other important components of global sourcing that can determine its maturity are:
Scope of work: Both breadth and depth of work sourced are important to assessing maturity. However, more important is the approach through which organizations decide the scope of work to be sourced. Mature organizations have defined frameworks/toolkits to achieve this, and make decisions based on assessment of risk and benefits while avoiding any stakeholder biases.
Sourcing model strategy: Does using both sourcing models (GIC and third-party), and even their complex avatars such as virtual GIC, make the sourcing model mature? Not really! Irrespective of the model, the measure of sophistication is in the manner it is used to achieve sourcing objectives, the type of engagement with the model (strategic partners versus provider of services), and approach to model selection.
Value beyond arbitrage: Mature adopters of global sourcing have moved beyond the cost savings-only play and are looking to create wider business impact through their sourcing strategy.
Location portfolio: Leveraging India or the Philippines does not make any organization more or less mature. The hallmark of mature organizations is their ability to build a location portfolio instead of a “collection of locations.” Each location in the portfolio has a designated role, and helps meet one or more global sourcing goals (e.g., savings, access to niche talent, risk diversification, etc.)
Demand management approach: This is often the most neglected parameter, or the one in which organizations tend to have a myopic view. Immature enterprises not only lack demand-profiling skills, but also tend to take a short-term view (12-18 months). Mature enterprises take a longer-term view, and align the demand for global sourcing with the organization’s overall business and growth plans.
Risk & performance measurement: Successful and mature organizations measure risk as well as performance of the global sourcing strategy, and leverage it to influence/inform their future global sourcing initiatives.
The next time you hear someone make a comment about the maturity of an organization’s global sourcing strategy, point him or her to this blog. Remember that not all organizations have the maturity to measure maturity!
If you would like to know more about our global sourcing maturity assessment methodologies, share your experiences, or have a discussion on this with one of our analysts, please contact us.
Did you see the news earlier this week about the Avon – SAP relationship in Canada? The eye-grabbing headline, “Avon halts work on big SAP implementation, cites lack of ROI,” thrusts the deal to the attention of both service providers and enterprises planning or in the midst of business transformation deals. The services industry has decades of piercingly clear evidence that large-scale implementations can be problematic and disruptive, and Avon stated that its decision to halt the further roll-out was to stop further disruption to its business. But I believe we can conclude much more than that from this news. In short, this action by Avon suggests that a good deal of change is happening in the services market.
In uncovering what is behind the headlines in the Avon – SAP relationship, let’s start with the fact that both Avon and SAP affirmed that the products company will continue using the SAP system in Canada, the system is working as designed, and their relationship “remains solid.” And they had experienced an earlier successful pilot of the project. So what sparked this recent decision to end the full-scale implementation?
As the saying goes, often what you don’t know will cost you. Here’s what you need to know about this deal.
We see this as a shining example of a phenomenon we are observing in the marketplace. The SAP implementation was part of a broad-based business transformation Avon was trying to drive. But in today’s market companies now are skewing away from the $100 million-plus or large-scale ERP implementations and instead look to buy outcomes and functionality or capability on as as-needed basis.
Fundamentally companies don’t want to spend hundreds of millions of dollars on transformative IT implementations because it’s too expensive, takes too long, and usually the return isn’t there. However, they still need new functionality or capability for competitive advantage so they can drive their businesses forward. This is where the as-a-service models come in; these models allow them to get the new capability without replacing their entire ERP system.
We see a significant movement in this direction. It manifests itself sometimes by companies asking for outcomes. (“We need these results.” “Just give us the insight from the analytics.” “Give us this new market access.” “Allow our retail business to deal with stock-outs.”) They are willing to pay good money for these capabilities, but they want to pay for it in this manner:
We find that the as-a-service models open opportunities for a different kind of sale. It allows a new breed of vendor or new breed of offering to get the customer out of the complexity of buying the component parts of a solution and focuses purely on the outcome of what the company is trying to accomplish.
Avon is happy with its existing ERP vendor, SAP; but it no longer wants the cost and time of the large-scale transformation project and replacement of its ERP system. We see the Avon decision as important.
Buyer and providers, take note: Avon’s action heralds a new set of industry offers that allow organizations to buy parts of ERP as a service and layer it on top of existing ERP implementations. How will the as-a-service models impact your business?
Photo credit: Marlon Malabanan
Whether you are a service provider or a buyer of services, you can benefit significantly from a relatively new delivery model in the services ecosystem — outsourced in-home services.
Why is it attractive? The in-home model leverages the labor arbitrage idea but applies it in a different way for even greater benefits.
Over the last 10 years companies learned that if they moved work to the people rather than moving people to the work, they would get a much lower cost of labor. For instance, moving software programming work from New York City ($150,000/year job) to Bangalore ($25,000/year) produced a big labor arbitrage. The in-home model takes the work to people, but they work from their own homes instead of being concentrated in an office, factory or call center. This model gives providers access to a much larger talent pool at a more affordable rate.
The in-home model works in locations such as India or the Philippines, but providers are deploying this model primarily in the United States or in region. In the U.S. where people are politically sensitive about sending work offshore, this is a particularly attractive model to consider.
Three critically distinct advantages of the in-home model
The first benefit of deploying the in-home model is access to a wider, richer and more diverse talent pool at a cheaper labor rate. At the same time, this pool is willing to work at a substantially discounted rate compared to workers in a central office, factory or call center. They have strong reasons for preferring to work from home, which are a trade-off for the lower price of their labor. Typically these preferences arise from:
All of these preferences for working from home are very important choices in today’s world. They lead to two other major advantages of deploying an in-home model:
Because people in this talent pool prefer working from their houses instead of commuting to an office or factory, they tend to be more loyal to the employer that facilitates this need and more focused on ensuring they meet or exceed the employer’s expectations for productivity.
In a nutshell, the strong attraction for providers to the in-home model is access to a completely different talent pool that is cheaper and often more loyal and more productive. So it’s not surprising that the in-home delivery model is showing tremendous growth.
Impact of work-management tools
Management tools to allocate and manage work in a virtual environment are transforming this space. These tools now are inexpensive as many are SaaS apps that enable providers to pay for them on a consumption basis. The SaaS model often makes these tools the same or even a lower-cost investment than tools needed for remote management in locations such as India.
Prominent areas of adoption
The in-home services model is quickly growing in voice services, especially in the call center and customer spaces.
It also has great applicability in rare skills such as programming. One of the very significant dilemmas in today’s digital world is aged-out technology languages that still need to be supported. For instance, it is very difficult these days to find COBOL programmers as programmers see it as a dead-end career.
The in-home services model eliminates the turnover problem and increases motivation and productivity in COBOL work as the greatest number of qualified, highly productive COBOL programmers are retired. They are willing to work but not in a central office or factory and not full time. They can supplement their income while living a retired lifestyle, and the provider gets superior quality, highly productive, very reliable support at a discount.
Another skill area that is ideal for the in-home model is legal services. In fact, we use this model for Everest Group’s outsourced legal services. In today’s world the major law firms have raised their rates to an unconscionable level, making the cost of legal services very high. Many firms are starting to tap into the in-home model, outsourcing work to ex-partners who have opted out of the legal race because of lifestyle issues. Rather than giving the work to a less-experienced associate, they have access to seasoned, credentialed lawyers whose turnaround time is faster and quality of work is better. And the price point is half or a third of the price for utilizing a partner in a major law firm.
It’s a win-win-win
Clearly the in-home model for outsourced services is a win for those who want to keep jobs in America, a win for workers who want to allow for a culture of self-actualization or are handicapped, a win for the service providers and their customers that get better services at lower cost. You can’t beat it — it’s a win for everybody.
This outsourcing model is gathering a lot of consideration around the world. I’ll discuss it in more depth in my upcoming presentation on “A New Paradigm — Truths and Myths” at the CORE conference on November 5 in Toronto.
Photo credit: Matt Crawford
While buyers have typically approached, evaluated, and made third-party business process service delivery sourcing decisions at the operational level, and separated out decisions on the underlying software applications and/or technology infrastructure, they are increasingly realizing the value of looking at IT and BPO in an integrated manner.
Enter Business-Process-as-a-Service (BPaaS), a model in which buyers receive standardized business processes on a pay-as-you-go basis by accessing a shared set of resources – people, application, and infrastructure – from a single provider.
There are many potential upsides of BPaaS…but is it right for your organization? The answer lies in evaluating the model based on a holistic business case that looks at a range of factors including total cost of ownership (TCO), the nature of the process/functional area under consideration, business volume fluctuations, time-to-market, position on the technology curve, and internal culture and adaptability to change.
Let’s take a deeper look at the TCO factor, which must be analyzed in terms of both upfront and ongoing costs for all three layers of service delivery.
For our just released research report, Is BPaaS the Model for You?, we developed and used a holistic evaluation framework that compared TCO for BPaaS and the traditional IT+BPO model across three buyer sizes: small (~US$1 billion revenue/5,000 employees), medium (~US$5 billion revenue/20,000 employees), and large (~US$20 billion/100,000 employees.)
BPaaS brings big benefits. It is independent of deal duration, delivers 35-40 percent savings compared to traditional IT+BPO, enables leverage of the provider’s economies of scale, provides access to otherwise cost-prohibitive technology, and allows entry into BPO relationships that as stand-alone’s lack the necessary scale. Small buyers are also highly amenable to BPaaS’ process and technology standardization requirements.
BPaaS is pretty impressive. It delivers 25-30 percent savings over the traditional IT+BPO model, which while less than what small buyers reap is still significant in driving a successful business case. Medium buyers’ increased scale allows them to capture some of the economies of scale benefits even in the traditional IT+BPO model, thereby having a lower differential between the two models.
BPaaS is not too shoddy. While it only provides ~10 percent cost savings compared to traditional IT+BPO, the absolute differential in cumulative TCO can still be substantial given the high base. And, in certain buyer-specific situations such as technology enhancements, exploration of new BPO/IT infrastructure relationships, and expiration of legacy technology licenses, BPaaS can be a good model for large buyers to evaluate. But…their tendency to balk at following a tightly defined, standardized approach – unless significant configuration features offset a good portion of customization needs – reduces BPaaS’ appeal for them.
As you see, our evaluation framework shows an inverse relationship between buyer size and cost savings from BPaaS – i.e., the larger the buyer, the lower the percentage savings. In fact, as buyer size increases, the scale benefits of renting versus owning infrastructure and applications can dip into the negative column over ten years. Of course, the assumptions in our BPaaS to IT+BPO model analysis are ideal, and your organization’s individual reality may be quite different.
To learn more about how to evaluate BPaaS’ applicability to your company, select a BPaaS provider and solution, and implement the selected solution, please read our report, Is BPaaS the Model for You?