6 AM Pacific

The Cost of Onsite Hiring Just Went Down

The US Southwest Border Bill imposes an increased application fee of an additional $2000 for every H1B and L1 employee on a few Indian Offshore Services companies. Much has already been said about it by politicians, Indian companies and NASSCOM. [NY Times, Hindu]

In early July I had written a piece on the challenges of Local Hiring or onsite hiring. In it I wrote

Also, there is the issue of optics. Can you do X billion dollars of business in a country, with tens of thousands of employees in the country and hire just a handful from the local market? The law may allow it, but you have to do much more in the court of public opinion.

Little did I know that Senator Schumer was actually planning to change the law itself.

But perhaps this will be the turning point for onsite hiring. There are other good reasons to increase onsite hiring. Again, from my post

Going forward, local hiring will become more important. Winning business is now less about technical expertise, which is assumed. It is about domain expertise – do you understand my business process well enough? It is going to be pretty tough to build domain expertise organically – through the work the company does for clients. Hiring industry expertise into the company either from industry or from other consulting companies will become necessary to compete.

Indian Balance of Payments and Offshore Services

The RBI is concerned about India’s Balance of Payments. One of the reasons ascribed to the growing problem is the slow down in the Indian Offshore Services industry. From today’s FT

One reason, the central bank said, for the deterioration in the balance of payments was a decline in an “invisibles surplus”, caused in part by falling revenues to India’s prized outsourcing sector.

For my book I crunched some data from the RBI website. As you can see, on the Current Account, India depends a whole lot on the Offshore industry.

Without the Offshore industry being where it is today, the import regime could not have been as easy as it is today.

But I am a little confused on the timing of this. Given that last quarter was very good for most of the top companies, isn’t this now no longer a concern? Or is it that RBI data gathering and analysis lags public companies announcing their results by a quarter?

The Quest for Higher Bill Rates

Thanks for all your comments on last week’s post on Local Hiring. Here’s the outline of the key issues we want to tackle in another chapter of the book called The Quest for Higher Bill Rates.

The need for higher bill rates, or rather services/solutions that will yield higher bill rates, is a consequence of two things.

One, with commoditization of major services like Offshore outsourcing, Application Maintenance, Customer service call centers and others (the list gets longer all the time), there is downward pressure on rates. Customers understand the service well and feel that they can manage the vendor to reduce risk. Deal consultants like TPI drive the commoditization even further.

Secondly, with strong growth comes competition for talent. Which leads to wage increases well above the rate increases that can be reasonably expected from clients.

Put together, the pressure on margins is tremendous. It helps that vendor switching costs are high, which mitigates the downward pressure somewhat with existing clients. But the need to do something to protect margins is going to become crucial in the future.

Increasing bill rates can be achieved in two ways. One, by differentiating core services to be able to command a premium over competition. Some services like large scale Offshore Outsourcing deals come with a fair amount of risk. Clients are often willing to pay a premium to a vendor who has a great track record and a solution that meets their needs. Within limits this can work well.

But what really pays dividends is investing in new services. New services are less competitive and the ability to make really high margins is great. For example, in the early days of ERP related services, we weren’t doing end-to-end implementations at Infosys, but revenue growth was strong and the margins were terrific. Why? because there weren’t too many service providers who could do offshore ERP work. The only alternative for clients was to have one of the Big 5 consultants do it for them at exorbitant rates, which left a lot of headroom for offshore service providers.

New services have always made good money. Until of course commoditization sets in. The trick therefore is to be a service innovator. Start early and stay ahead of competition. As the market expands, the early pain pays off in the form of high margin, high growth revenues.

But we are now at a stage in the industry where service innovation by itself is not good enough. The addressable space in new services that are being launched now is much smaller than the markets already being addressed. The impact that the new services can make on the company’s margin is therefore smaller. We may still not be there yet, but within five years, services that are new today will also have seen commoditization and there won’t be enough potential in introducing new services at that time.

This is not to say that the potential for innovation will go away. Quite the contrary. The action will shift from services to solutions – bundled services and IP that tackle a specific business problem. The IP included could be a platform that accompanies the service, or just a deep understanding of the business problem and the experience of having solved similar ones many times before.

By its very nature, each solution will have a smaller addressable space, but in aggregate could offer great potential. Pricing could be traditional time based pricing or it could be by transaction. The key thing is that competition will low and so there will be more freedom in pricing. The client is likely to look at the cost of the solution vis a vis the value it brings to his business.

This is a very profound change for the Indian offshore industry. Their approach to business thus far has been ‘build it and they will come’ – which works fine for services. But for solutions, you have to be out there, in the flow of what’s happening in your clients’ businesses. Identify pain points that can be solved by the combination of technology and services. And then invest the time and money to build a solution. You go from being an expert in delivering a service to becoming an expert on a business problem or process.

It also means that services companies will have to make many ‘bets’. Solutions require investment. How do you decide where to put your money? Or how much. Because of this difficult transition, I think that services companies will largely acquire technology or IP for their first solutions, instead of building it themselves.

How else will this transition from services to solutions affect companies?

What’s the Size of the Offshore Industry?

I thought I knew the answer to this one. Turns out, the correct answer is “it depends”.

The GoI puts out a number which is called “Miscellaneous- of which Software Services” or Software Services in short. This data is available on the RBI website. The number for 2008-09 is $46.3 B (yes, the RBI does give data in both Rs. and USD).

This number comes from the Balance of Payments report. Software Services is actually not clubbed with Exports. It is part of Invisibles, along with Travel, Financial Services etc. – services that can be sold to parties outside the country but are never “shipped”.

This Software Services number includes both IT Services and BPO (and all other variants of XXS and XXO) but includes only revenue from billing done by Indian companies or Indian subsidiaries to their foreign clients or foreign parent companies. The problem with this number is that for an Indian company like Mindtree with no subsidiaries it will be pretty close to actual revenue, but for a company like Cognizant it will be what the Indian sub bills the parent for services provided. In all likelihood this number will be cost plus something like 15%, which is well below the typical mark up on offshore services of more than 100%. Additionally, for Cognizant the entire onsite revenue will be excluded.

I actually didn’t really check if Mindtree has foreign subs or not, but if it did then the RBI number would exclude the subsidiaries’ revenues, including only the Indian parents billing on the subs.

So, as you can see, the $46.3 B number is well under what you would consider the true revenue of the Offshore industry, which includes onsite operations related to the offshore business.

NASSCOM tries to correct for this revenue by including the revenue of foreign subs, but I don’t believe it does anything about the Cognizant situation where a foreign company has a sub in India. In truth, that is a slippery slope and it shouldn’t even attempt it. To figure out the onsite revenue associated purely with the offshore business is impossible to orchestrate over all companies. Imagine trying to do that at Accenture or Keane.

Also, included in the RBI number as also in the NASSCOM number are exports from Captives where onsite operations may not even be meaningful.

Anyway, that answers the question that used to puzzle me for the longest time – why are all the charts for software exports and not the Industry revenue? Sometimes you measure what’s measurable and leave the rest as an exercise for the reader.

By the way, I had been tearing my hair out trying to get the net Exports data for the IT-BPO industry. Finally found it when I came upon the RBI database dbie.rbi.org.in/ which is quite comprehensive and easy to use, though difficult to find. I sent some feedback. They fixed the data and sent me an email in one day! On the other hand, I sent an email to NASSCOM’s research department. No reply after a week. Go figure.

Why is Local Hiring in Offshore Services so Sparse?

For some parts of the book that I feel the need to reach out to my blog readers. Your feedback will make a better product.

One of the chapters in the book is on Local Hiring – why Indian services companies haven’t hired in greater numbers and why it might become more important to do so in the future.

The percentage of employees in Indian IT Services companies who are Indian (born and brought up in India) is definitely over 95%. If you leave out the employees acquired through acquisitions, local hires (Americans in the US, Germans in Germany) are a rounding off error. Why is this so low?

Over 95% of the revenues of Indian services companies comes from outside India. 20 to 30% of billable employees are onsite. Shouldn’t you expect a higher proportion of employees to be hired locally? After all, IT Services is a high-touch business. If you are onsite, you are constantly working with client personnel. Cultural proximity, common backgrounds, networks from college and prior jobs must be of great value. Yet, local hiring is still very slim.

There are some good reasons why this might be so.

One, Offshore services is a very unstructured business. Every project is different – client needs, business processes, size of project – can all differ greatly from one project to another. The use of templates is limited to a few project processes only. Most of the project requires “new work” every time which in turn requires a fair bit of collaboration among the project team members – both offshore and onsite. This need for intense collaboration makes it much easier if the onsite team members are actually offshore team members who have moved onsite temporarily.

Then there are cost reasons as well. Not so much compensation which for most companies is within range of what it would cost them to hire locally. But there are other reasons why the costs are higher to hire locally. Like the fact that a local employee would likely need travel and living since the project can’t always be in the metro area in which she lives. Also, for local hires, bench costs are at onsite wages.

Outside of Delivery – in Sales and Consulting – the cost reasons above don’t apply. At one time, there was a salary difference. But now, Enterprise IT Sales and Consulting are both in decline and Indian Services companies have upped compensation substantially. So that isn’t a reason either.

But even here, I feel that companies feel torn between the need to hire folks who could be more effective in forming relationships with clients and the need for them to work effectively with offshore teams. Anyone who has worked in the industry knows that a salesperson’s effectiveness can often depend upon how good his access is to expertise and resources in the company.

The only place where there really are no barriers to hire are in senior positions in Sales and Consulting. Where you get the value of their experience in a certain industry and their networks but on the ground the proposals and staffing of projects is handled by other people. But here too there is a mismatch in expectations.

While Indian companies haven’t shown much urgency in hiring locally, it’s not as if potential employees are trying to break down the door either. Perceptions in the marketplace about Indian services companies are that you are generally “thrown into the pool to sink or swim” with very little support. Also, that policies aren’t as employee friendly as their current employers. You don’t have a life any more and work takes up every waking hour. Although, I think this is the nature of the offshore business, and if they are still working regular hours, it will change soon.

Going forward, local hiring will become more important. Winning business is now less about technical expertise, which is assumed. It is about domain expertise – do you understand my business process well enough? It is going to be pretty tough to build domain expertise organically – through the work the company does for clients. Hiring industry expertise into the company either from industry or from other consulting companies will become necessary to compete.

Also, there is the issue of optics. Can you do X billion dollars of business in a country, with tens of thousands of employees in the country and hire just a handful from the local market? The law may allow it, but you have to do much more in the court of public opinion.

Take Toyota for example. When they started putting up manufacturing plants in the US in the 80s, it was probably much easier to just expand in Japan. But they knew that if they were going to be a significant player in the huge US market, they had to do some local manufacturing. Today, Toyota has manufacturing facilities in many states in the US. When they ran into trouble earlier this year because of their quality problems, their American employees and politicians from the states that have the factories, supported them.

In a few years US or Europe based competitors who are rapidly adopting the offshore model will have a truly global workforce with a (largely) India delivery model. If nothing changes, Indian companies will be almost entirely Indian. The lack of diversity will lead to a lack of diversity of thinking. And exposed to political games.

IT Services Companies Should Empower Middle Management

Last week was quite eventful in the “Workshop of the World”. Foxconn, a Chinese manufacturer for companies such as Apple, Dell and HP, gave a 20% ad hoc raise to its workers after as many as ten suicides which called into question the working conditions at its plants. Fox Conn employs 800,000 workers in 20 plants across China.

Then it was Honda’s turn. Workers in Honda’s four factories in China struck work bringing all production in China to a halt. Yesterday Honda gave a 24% raise to workers, who were still not happy.

The strike in of itself was quite surprising. From FT

The right to strike was excised from the Chinese constitution in 1982, and attempts by workers to organise outside the official All China Federation of Trade Unions are frowned on by Beijing.

Amenities at Chinese factories like Foxconn are actually considered to be good. From Guardian

Foxconn is proud of the fact that it provides a swimming pool and other facilities to its staff, as well as organising chess, calligraphy, mountain climbing and fishing.

At both companies pay is not great but is above the legal minimum wage. Many workers make much more by working overtime.

How is this, in any way connected with the Indian IT employees’ angst? They are both about employees having nowhere to go with their problems.

Even though their demands and managements’ responses have been focused on pay increases, Chinese employees taking on their managements, is not really about salaries. How can it be when most employees come from the hinterland where wages are poor, that is, when there are jobs available? It is because the workers feel powerless. They are lost in these huge organizations. Foxconn has 800,000 employees. Honda also employs a similar number across its four plants. The “official” trade union is not elected and is really part of the establishment – a proxy for management. The workers have nowhere to go with their problems.

Cut to the Indian IT Services industry. Most of the successful companies of today are very centralized in how they operate. Historically, this was necessary. To scale up they had to build an organization which ran efficient, repeatable processes. For that it was necessary to have strong central control. But the very thing that helped them scale engineering and business processes, made the middle management powerless and weakened the bonds between the company and employee. It failed at the most important aspect of scaling a company – in building a loyal, motivated workforce.

When a company is small, the founders or the members of the top management know everybody themselves or with one degree of separation. They infuse the whole company with their values. Employees form a relationship with the company based upon these shared values.

But as the company grows, one degree of separation becomes two, three, four and more. Pretty soon what top management gets to hear is what they hear from their direct reports. What they have to say is said to a select few or to the media. How do you continue to build trust with your employees? You can’t do it yourself. So you must have your middle managers become interlocutors for you.

The problem is that middle managers are so disenfranchised that they feel powerless. They are the ones who run the delivery teams, who make the company tick. But they don’t have the leeway to solve their own day-to-day problems. When they take their problems to their superiors they discover that they too are powerless. So they just keep punching the clock. If somebody in their team comes to them with a problem, they just say, “I can’t help it, that’s the way it is”. When the rank and file IT worker starts griping about the company, some of these managers stay silent. Others join the griping. Nobody defends the company. They can’t. They don’t believe in it themselves. That’s where the battle is being lost.

In the case of Chinese workers, establishing a union that is truly representative might actually be the solution. Empowering middle management is less effective. In manufacturing there is a “class divide” between managers and workers, often based upon education and qualifications. Very few workers will ever be promoted into management. So the “us vs them” feeling almost guarantees strife, especially if as in the case of Honda, the Japanese managers earn much, much higher salaries.

But in an IT Services company every engineer has the qualifications to become the CEO of the company. It should therefore be easier to bridge the divide. Also, regardless of the quantity of outpouring on the internet, it still does not represent widespread disaffection.

Winning back middle management is the first step in turning things around with employee relations. Stuffing their mouths with money just pushes out the day of reckoning.

Indians and Unpredictability

A notice at a neighbourhood Postal Annex. Truly, a picture is worth a thousand words. The proprietor is, you guessed it, Desi.

The notice is a microcosm of Indianness. If you get past the English (hey, its a foreign tongue, so stop being so fussy) it really is a reflection of our relationship with predictability.

There are three kinds of Indians:

- Those that are unpredictable and don’t care.
- Those that are unpredictable, but would like you to expect, and perhaps accept, their unpredictability. The store owner who put this notice up belongs to this set.
- The predictable kind. Within India, this is a very small set.

Indians are not brought up to be predictable in their behavior. The environment (Bangalore traffic for instance) doesn’t allow us to be. But this is certainly a phenotype issue not a genotype one. Because somehow, when Indians leave India they leave their unpredictability behind.

The great achievement of the IT Services industry has been to extract predictable outcomes for clients out of this morass of unpredictability. They erect these boundary walls around the company. Within these figurative walls there are no power cuts and meetings start on time. Defects are measured and deadlines are met. Because that’s what clients in the developed world expect. It’s gotten easier and easier over time, but in the early days the pioneers did the equivalent of moulding square pegs to fit into round holes.

Growth and the IT Services Pyramid

This appraisal cycle has been rough for Infosys. Unfortunately, at over $5B in revenues, everything it does becomes a media circus. Or in this case a social media circus.

Economic Times published a few stories in February and March of this year on a new appraisal and promotion policy at Infosys. The stories were their usual unremarkable fare, but what was truly remarkable was the sheer volume of comments (452 on this article as I post this) and the apparent rage of the commentors.

There are many interesting lessons one can learn from this episode. For one thing, this must have been quite an education on social media to many company managements in the industry. But the thing I find most interesting is that, unfortunate as the situation is, there is very little that Infosys or any other company can do about it. As growth slows down in the IT Services industry, career growth must also slow down. More »

Is Offshore Services a Scale Business

Earlier this week I wondered why Indian Offshore Services companies weren’t “bulking up” – basically acquiring companies just like themselves – just smaller and with perhaps lower P/Es.

A reader left a thoughtful comment :

I guess one of the common reasons for M&A (to add bulk) is to achieve economies of scale. Indian IT outsourcing is services business. In services industry, (compared to manufacturing industry) the economies of scale flatten out relatively fast. The profit/revenue per employee for a 10,000 employee company may not necessarily be better than a company with 5000 employees. After all its all about humans, not machines.
So a large company (Infy,Wipro etc) doesn’t gain much by acquiring a smaller size company. Their own brand name is already so powerful, that they don’t need much help in that department.

The comment makes a lot of sense and it would be hard to disagree with the fact that the IT Services does not have the economies of scale that a steel manufacturer does (ergo, Mittal acquiring every company in sight). But there are important economies of scale at the top of the range. Let me explain.

If you look at the financial metrics of an IT Services company, it is almost entirely a variable cost business. Both capital intensity (capital investment required to produce a unit of revenue) and fixed costs are low.

Also, IT Services revenue tends to be sticky. Especially, if the revenue is of the kind that “runs the company” rather than “changes the company”.

So you have an industry where you can get into business with very little capital. You breakeven fast because the fixed costs are low. And the revenue you have is relatively immune to competition. Which explains why you have so many small companies in the industry.

But you also have giants in the industry – IBM, Accenture, Infosys, TCS. There must be something that makes the giants grow faster than the guys at the bottom of the range.

There are indeed. Some things like Overheads and Sales costs could both be more efficient (lower % of revenue) and more effective (cover more industries, more countries in the world) with size. But these aren’t that major factors. The important one is the economies of scale related to branding.

Outsourcing large scale IT or BPO, more often than not, comes down to a matter of trust. A C level executive at a Fortune 1000 company is making a huge bet that can make or break his career and perhaps even a few quarters for the company. And while a part of that trust is derived from the service provider’s representatives working on the deal. A big part of it comes from that intangible – the brand.

And when it comes to branding, size matters. It’s not just that you can spend much more on advertising if you are bigger (like IBM or Accenture). Your brand automatically benefits from your larger presence. You have more client references (who have hopefully had good experiences with you). The analysts know you, the deal consultants know you. You are public. You are in the news. You are automatically on every short list.

Branding also matters at the other end of the value chain – in hiring. Employees like to join large established companies and will often forgo a higher paying job at a smaller company. (Silicon Valley may be the one exception to the rule and even here, things have changed). So Infosys is basically getting the cream of the crop at a lower cost than their competitors. Is that worth a few points on the margin? You bet.

Here’s the nub. There are several companies around $1B that are well run companies, but they run the risk of becoming marginalized, of losing the mindshare game to the big guys. In a market where growth may be harder to come by, they will have to compete harder against bigger brands, if they want to keep shareholders happy. Inorganic growth (acquisitions) can be one plank of your strategy in such companies. Obviously, it can’t be the only thing you’re doing.

That said, gaining economies of scale is not the only reason to bulk up. It has to make financial sense. Relative stock valuations, cash on the balance sheet – those are the table stakes.

Offshore Services: Whither Consolidation?

In the last 18 months the markets took a deep dive and then recovered. Valuations in the Offshore Services industry yo-yoed, business prospects sank and then recovered. Yet amidst all this, the pace of acquisitions didn’t go up much. Apart from the Satyam acquisition by Tech Mahindra, deals have been small and infrequent.

The question is why? From the industry structure it would seem like the Offshore Services industry is ripe for consolidation. It has a few major companies that have scale, geographical and service breadth and brand names both in the market and the talent pool – companies like Infosys, TCS, Wipro, Genpact, Cognizant. And then you have companies distributed across the revenue spectrum. There are mid-sized companies, small companies and tiny companies. Shouldn’t the market be consolidating?

Not necessarily. First there is the timing issue. Valuations are back to being on the high side – which makes the buyer skittish. More importantly, growth is back. When the industry is growing, sellers don’t see a reason to sell. The companies that are growing don’t want to sell because they think they’ll get a better price next year.

But is growth evenly distributed in this recovery? It doesn’t seem so. I haven’t looked at the data, but my sense is that recovery of growth has been spotty for smaller companies. Which should mean that larger companies with a high P/E should be happy to bulk up with, but that isn’t happening.

I think the reason is the quality of revenue. Acquiring companies are looking at quality of revenue – acquisitions that bring them geographic penetration, domain expertise or a wider service footprint. And because these acquisitions are complementary the companies being acquired can be quite different from the acquiring company. Company culture, business operations, core skills may be different enough that the acquiring company is afraid of biting off too much. Which is why the acquisitions are small. “Pearls in a necklace” as Wipro has called it in the past.

But a scalar acquisition strategy – basically bulking up – can make a lot of financial sense. If you are a well managed company with a P/E that is higher than the industry, an active acquisition strategy can make sense.

First, if you can negotiate a price for the target that is somewhere at or below your P/E, you are already looking good.

Second, acquiring a company with a similar business, has fewer integration risks.

Third, this industry, even though it has low entry barriers, is a business of scale. Visibility, brand, lower sales costs and overheads – all come from scale.

I think the Satyam acquisition is going to be a big test case. If the merged operations of Tech Mahindra and Satyam are successful as a single company, it will prove that large acquisitions can work and can work financially, if the price is right. As growth in the industry slows down, especially for undifferentiated mid-size companies, acquisitions should pick up pace.