Will SAS' star continue to shine in the increasingly crowded BI space?
With five billion gigabytes of information generated worldwide every day there has never been more need of ways to help make sense of this gigantic data deluge.
That's where software company SAS comes in - aiming to allow businesses to analyse data and glean insights such as how to outperform competitors or better serve customers.
Not that SAS is some upstart hoping to exploit the modern information explosion - the company has been around since 1976 and has seen its revenues rise every year since its creation. In 2009 SAS' annual turnover reached $2.31bn, generated by a global customer base that includes 93 of the top 100 companies on the 2009 Fortune Global 500.
SAS is unusual among major software companies in that it has never been floated on the stock market, and today is the world's largest privately owned software company.
The lack of investors pulling the strings behind the scenes has allowed the singular vision of its CEO and co-founder Dr Jim Goodnight to come to the fore.
SAS headquarters in Cary, North Carolina
(Photo credit: Jo Best/silicon.com)
Goodnight, with his doctorate in statistics, has driven SAS' heavy R&D spend, which swallows up more than 20 per cent of its revenue each year.
Outside of the lab, another hallmark of SAS is the focus it has on investing in its staff, offering facilities to rival any Silicon Valley HQ at its campus in North Carolina. As a result, SAS topped the 2010 list of the 100 best companies to work for compiled by Fortune magazine, its 13th successive appearance on the list.
Yet the road ahead for SAS is not clear of threats: the shadow of IBM looms large, as it attempts to muscle in on SAS' business analytics turf with its purchase of SAS' main competitor SPSS.
At the same time, Base SAS, the programming language for SAS tools that for decades has been the language of choice for carrying out statistical analysis is being challenged by an open-source alternative known simply as R.
Interesting times then - a period of growing opportunity but also of new challenges to SAS' statistical analysis crown.
Back to the beginning - how a private empire was built
The company was born out of a programme to analyse data collected by the US Department of Agriculture at the University of North Carolina.
At the time of its founding in 1976...
...SAS software ran on hardware very different to the server rooms of today, with SAS software stored on punch cards and processed by room-sized computers. But SAS quickly expanded its customers to include pharmaceutical companies, banks, new government bodies and other organisations. The company broke the $1bn revenue mark in 1999 - establishing a global customer base it has continued to grow to this day.
Throughout the company's uninterrupted growth, and in contrast to its rivals, Goodnight has resisted the temptation to go public with SAS, giving the company greater freedom to set its own business strategy than its publicly owned peers.
For Dan Sommer, principal research analyst for software markets with analyst house Gartner, SAS' privately owned status has given the company a competitive advantage.
"They can operate in a whole different way than companies that are chasing the next quarterly close," he said.
"It allows for a longer-term focus - for example, in 2009, when a lot of its competitors downsized to meet challenging market conditions, SAS chose to keep all of its talent."
Freed from investor oversight, SAS can also afford to invest in capital projects with an eye on long-term strategy, rather than short-term payback.
"They have invested in their campus - other companies would not be able to dream of making such a large capital investment," said Sommer.
"They also have large datacentre farms that could reap benefits over coming years, for example allowing them to build up their cloud and grid computing strategy," he said.
There is also less external pressure on SAS to release software products in order to bump up quarterly financials, allowing the company to focus on maintaining a high quality bar.
Thanks to its ability to look beyond short-term returns, SAS has also been able to sell its products to parts of the world that are not well served by its competitors.
"SAS has gone in and built a footprint, whereas other haven't done that because the return on investment wasn't clear short term," said Sommer. He cited SAS' strong presence in the Middle East and Eastern Europe - SAS revenue in 2009 was split so that 45 per cent came from Europe, Middle East and Africa, 44 per cent from the Americas and 11 per cent from Asia Pacfic.
What SAS offers
SAS' software and services broadly cover three areas: organising and integrating data, analysing that data and reporting the results.
SAS' data-integration products...
...and that of its partners, such as data-warehousing specialist Teradata, bring together useful data from inside and outside of the business so it can be analysed.
SAS products can organise and bring together the two types of information that will be handled by an organisation: structured and unstructured data.
Structured data is information that has been tagged or linked to other data to make it machine-readable - for example, data that has been categorised and organised into data types in a database.
Unstructured data is any information that is not tagged and therefore is not machine-readable, and can be anything from text in a document, email or blog post to picture or video files in an email attachment.
This unstructured data accounts for 95 per cent of all information that enters an organisation and is more difficult to categorise, and therefore analyse. SAS is adding new products designed to analyse unstructured data product portfolio, which includes SAS Text Analytics and SAS Social Media Analytics.
SAS helps businesses get to grips with the variety of data a modern organisation has to deal with
(Photo credit: Shutterstock)
SAS focus on the high end
When it comes to data analysis, SAS' strength lies in its core business of high-level statistical analysis, such as predictive modelling and data mining, an area where SAS has been a market leader for decades.
This type of analysis allows companies to pull data from many different sources inside and outside the business, and run it through sophisticated analytical software to discover customer or business trends. For example, feeding in data on past customer behaviour might reveal which group of customers are most likely to switch to an organisation's rival or in which geographies a certain product is likely to be successful. The software will flag up these trends itself, rather than simply providing information in response to a query provided by the user.
Gartner's Sommer said SAS continues to lead in the high-end analytics market: "They have a whole load of products - everything from SAS Enterprise Miner to a programming language called Base SAS, which is pretty much the lingua franca when it comes to statistical analytics.
"That's clearly their sweet spot...
...and what they have been able to monetise to a very large extent because there hasn't been much competition that has been able to reach parity with them in that space."
Lower end business-intelligence services differ from these more sophisticated analytical techniques in that they generally return information based on specific queries made against historic data. For example, a retailer may ask the BI service to highlight the month in which the business sold the highest numbers of a certain soft drink.
The battle for the BI platform market
SAS, with its traditional focus on high-end statistical analysis, is not the market leader in providing these lower end BI platform services, such as ad-hoc query and Online Analytical Processing (OLAP), according to analysts.
Madan Sheina, lead analyst for information management at analyst house Ovum, said: "SAS used to thumb its nose at traditional BI, it turned out that this is really what companies want, they don't want rocket science.
"A lot of BI vendors start off doing the simpler things and then move onto the more complicated stuff; SAS started at the top and is having to change parts of its strategy to address the not-so-complicated needs."
He estimates that SAS is about the fifth largest provider of BI platform services behind SAP, IBM, Oracle and Microsoft - which have swallowed up many of the independent BI specialists such as BusinessObjects and Cognos.
Gartner's Sommer agrees: "[SAS] went into the BI platform space five to 10 years ago but in that space they are more of a me-too vendor," he said.
If Sommer is correct, the company's focus on the high-end analytics market could impact its future revenue growth due to limited demand for such services. A survey by Gartner in 2009 found that high-end analytics services, such as predictive modelling and data mining, account for less than 10 per cent of overall business intelligence use by companies.
In contrast, ad-hoc query accounted for about 50 per cent of business intelligence use.
Counting in SAS' favour is the fact that demand for the type of high-end analysis that SAS specialises in is highest among major corporates, whose operations are large enough to continue to provide healthy returns for SAS.
And while SAS may offer more analytics than most businesses currently use, Gartner's Sommer said...
...demand for advanced predictive capabilities is likely to grow in future.
"We certainly can see that shift towards more predictive functionality; the question is when and how strongly it will happen," he said.
Who uses SAS products
SAS provides business intelligence and analytics to every major industry: helping retailers decide what stock to put in stores and at what prices, allowing insurance companies to better assess policyholder risk to determine premiums, helping banks to detect fraud more quickly and letting manufacturers match their output to customer demand - to name but a few uses of its products and services.
As well as producing products tailored to individual industries, SAS also provides analytics products that target common business demands - such as supply-chain management, CRM, financial analysis and IT infrastructure oversight.
Overall, SAS provides software and services to more than 45,000 customer sites in 127 countries.
Financial services account for close to half of SAS' revenue, 42 per cent in 2009, with SAS analytics helping financial institutions judge the risk of investments, check regulatory compliance and detect fraud.
Providing business intelligence to the financial services industry will remain a focus for SAS going forward, according to Goodnight, who told silicon.com that "banks and insurance companies... especially in Europe" were the industries in which there was currently greatest demand for SAS products.
As well as helping banks to assess financial risk, SAS is bringing its real-time credit and debit card fraud detection, which it provides to HSBC to other banks.
Fraud detection is also a major part of the work SAS does with its next largest customer base by spend: the public sector. SAS works with many governments, including...
...the US, helping deal with healthcare fraud, which is a growing business for SAS and in the UK.
Why leaving SAS is hard to do
Staff at SAS' HQ in Cary, North Carolina, seem loathe to leave, perhaps because the landscaped campus boasts facilities that would put a town to shame.
While SAS runs a 35-hour working week and offers staff flexible hours, there is little reason to stray from the campus site, which boasts a healthcare centre, subsidised cafeterias and day-care centres, dry-cleaning and even a meditation garden.
If that wasn't enough, employees can also relax in the billiard hall, gym, hair or manicurist salons, or in the Olympic-sized swimming pool that can be found in the campus' 66,000-square-foot recreation and fitness centre.
Another perk of working at the Cary campus is an office for almost every member of staff, not to mention the free M&Ms campus-wide, with staff having been known to munch through 22 tonnes of the sweets in a year.
A shot of the landscaped exterior of the SAS campus
(Photo credit: Jo Best/silicon.com)
SAS' benevolence towards its staff does translate into a tangible business benefit beyond keeping staff productive, with Goodnight claiming employee churn at SAS remains far lower than the average turnover at its competitors.
"The fact that we try to make it a great place to work has enabled us to maintain a very low employee turnover rate," he said.
"We don't see much over four per cent, whereas other software companies are seeing turnover in the teens or as much as 20 per cent."
In Goodnight's eye, the cost of maintaining a happy and, more importantly, stable workforce is money well spent.
Sizing up the competition and the threat from IBM
Because of the breadth of industries served by SAS, the company has...
...many competitors, ranging from small specialist vendors to large corporates such as SAP and IBM.
Increasingly, SAS has found itself in competition with the major tech players as they swallow up independent BI and data analytics pure-plays in order to build their own software portfolio - with SAP acquiring BusinessObjects, Oracle buying Hyperion and IBM picking up Cognos and SPSS, to name a few examples.
It is IBM's 2009 acquisition of SPSS, SAS' main rival in delivering high-end data analytics, that Gartner's Sommer believes will pose a challenge to SAS.
With SPSS under its wing, IBM will offer customers advanced data analytics as part of the extensive suite of IT software, hardware and services in its stable. "SAS is facing a competitor that has everything, the whole stack, and also significant consulting and skills," he said.
IBM's commitment to competing in the data analytics market is also reflected in its investment in delivering analytics as an outsourced service. "They will have 8,000 consultants doing this type of stuff by the end of 2010 - so it's a big bet from IBM," said Sommer.
The New York headquarters of IBM, which is challenging SAS in the high-end analytics market following its acquisition of SPSS
(Photo credit: IBM)
Goodnight himself is not fazed about the threat from big tech players buying up specialists such as SPSS. He maintains that SAS' smaller competitors rarely become more effective competitors after being swallowed up by a larger entity.
"We have about 400 listed competitors that we follow, because we are involved in so many industries we find there are specialist companies that cover just that one industry," he recently told silicon.com.
"We have been seeing companies like Oracle or IBM buying those competitors up, although they are still pretty much the same competitor. Sometimes they tend to disappear because the acquisition didn't work well.
"We reckon there's only a 50 per cent chance of those companies being acquired surviving."
Why SAS is choosy with its chequebook
As a company with a history of building technology in its own R&D labs...
...SAS has been less acquisitive than some of its competitors, such as Oracle.
"SAS hasn't been a particularly aggressive acquirer historically and I think that has been a conscious stand for the company to build everything in-house," Ovum's Sheina said.
SAS generally acquires companies in order to pick up technology to enhance its existing products and give SAS the ability to enter a particular market, rather than as a way of growing SAS revenues inorganically or capturing market share.
For SAS, some acquisitions are about improving its existing analytical tools, such as its purchase of language-processing specialist Teragram in 2008, which has allowed SAS to build natural language-processing into its tools for text mining and analysing unstructured data.
Other acquisitions are focused on picking up tech that will give SAS a foothold in a specific industry; for instance, SAS' decision last year to buy Memex - a company specialising in technology that manages criminal intelligence - which will help SAS to strengthen its law-enforcement products.
SAS also uses acquisitions to better serve particular business needs; for example, its 2008 purchase of IdeaS Revenue Optimization, a company focused on enterprise revenue optimization technology, which is strengthening SAS' revenue forecasting, optimisation and pricing offerings.
SAS' many partners
While SAS may not be the most acquisitive company, it does have partnerships with some of the biggest hardware manufacturers and technology consulting firms in the world.
The partnerships allow it to bundle its products into packages that are more attractive to its customers - for example, allowing SAS customers to buy SAS products with the computing hardware needed to run them or to get help in integrating SAS offerings with existing enterprise software.
SAS' competitor, IBM, is also one of its major partners, working with the firm's technology arm to sell the hardware and software needed to run SAS products and with IBM's consulting arm on integrating SAS software into existing ERP, CRM and supply-chain management technologies.
Other important technology partners include...
...HP, Intel, Oracle and Teradata.
On the consulting and outsourcing side, partners provide advice and support on using SAS products, as well as providing hosted SAS services, with major partners including Accenture, Capgemini, CSC, Deloitte and Wipro.
A new language for analytics
Base SAS is the programming language used to write programs to carry out statistical analysis using SAS tools.
For decades, it has been the main programming language taught to university students learning how to carry out statistical analysis.
The upshot had been that most statistics graduates' skills lay in using SAS products, leading to those products being the de facto choice for statistical analysis after those graduates enter the workplace.
"SAS were very early going into universities with their Base SAS programming language," said Gartner's Sommer. "When people graduated they went out with that language and when they were hired that led them immediately to source SAS software."
However a challenger has emerged to Base SAS as the statistical analysis programming language of choice, with the rise of a new language called R.
Today there are more routes to choose from when it comes to programming languages for statistical analysis
(Photo credit: sacks08 via Flickr.com under the following Creative Commons licence)
R is an open-source language that is allowing university students to learn how to carry out statistical analysis using software tools not linked to SAS.
Sommer said: "Now there is this open-source programming language that is spreading and a lot of [statistics] graduates program largely in that language.
"[It] obviously opens up the market in a whole different way than it used to be, when all the skills were driven by SAS tools."
R is also allowing competing BI vendors to challenge SAS in its core market of statistical analysis, with vendors, such as Information Builders, building statistical analysis tools into their existing BI software that are controlled by R.
"All of these small vendors are coming out with all kinds of statistical analytics tools that weren't there before," said Sommer. "There's going to be more competition going forward of people embedding analytics into their tools that they weren't able to do so to the same extent before. That's a major threat to SAS."
Cracking the SME market
SAS' business model may also pose an obstacle to the company...
...achieving the stellar revenue growth it has in the past, according to analysts.
A challenge faced by SAS is how to bring more smaller organisations - those with fewer than 1,000 employees - into its customer base, according to Sommer.
"SAS has always targeted larger organisations, in particular in financial services. We estimate that one-tenth of the revenues are companies with fewer than 1,000 employees," he said.
Sommer added that one of the barriers to getting smaller companies to use SAS was that smaller companies struggled to employ staff with the necessary skills to get the greatest benefits from using SAS software.
"There's a skills gap, it's very difficult to source those PhDs, those guys that have this knowledge to use SAS to its largest extent," he said. "Those types of people have generally been hoovered up by those larger organisations."
A further difficulty for SAS has been finding the right organisations to sell on its products to SMEs. Sommer said SAS' focus on corporate customers could prove a problem for the company, as the larger companies are already heavily invested in BI and analytics tools, limiting the potential for future growth.
"I think the Fortune 2000 is largely sold into already, all of the big companies have five different BI tools already so there's more greenfield opportunity in mid-market accounts," he said.
Ovum's Sheina believes that SAS is in a better position to target SMEs today than it was in the past, by offering them prepackaged versions of its analytical tools targeted at SME needs, which don't require much expertise to set up and use.
SAS claims that it is winning over SME customers, with the company announcing that 80 per cent of new commercial accounts in 2009 were SMEs with annual sales under $500m.
Winning over new customers
The wider question of how many new customers SAS is able to attract is also a concern for Sommer.
He said SAS' products suffered from a reputation for being expensive...
... that discouraged some businesses from starting to use SAS.
"We are hearing that people discard it immediately; because they have that reputation of being expensive, they don't really evaluate the product.
"A very small proportion of their revenue pie is sales to new accounts. Their model is rather to go in small and to expand from there in existing accounts," said Sommer. "They are more likely to expand into existing accounts than to find new customers.
"We estimate that roughly one-fourth of SAS' annual software revenue - not including consulting - is new software sales. We expect that probably less than 10 per cent of software revenue is sales into new accounts outside install base."
What the future holds for SAS and Goodnight
So much of the values that are quintessentially SAS - from the high investment in R&D to the enviable employee benefits - stem from the philosophy of the company co-founder James Goodnight.
Given Goodnight's level of influence on SAS, it is unsurprising that there is a marked level of interest in who will succeed the 68-year-old as company CEO.
In 2006, Goodnight said he saw himself remaining in the role for at least another 10 years, and more recently stated that publicly naming a successor for CEO can cause a company to lose its top staff, saying: "if you say it, two or three of your best people may decide to leave because they're not going to make it [to CEO]".
Gartner's Sommer said: "As long as there is not an openly communicated succession strategy then it is a little bit of a concern because the whole company is so much built around Goodnight's strategy in terms of the product itself and SAS' culture as an employer."
Goodnight's departure "will leave a bit of a hole," Sommer added, but in spite of Goodnight's refusal to comment on the issue in public, SAS had mapped out the company's direction after Goodnight's departure.
Whatever the future holds for a post-Goodnight SAS, the man himself told silicon.com he is confident that the SAS of today still embodies the values of the small firm he co-founded more than three decades ago.
"I don't see a whole lot of difference between the company now and when we first started," said Goodnight.
"The culture of SAS is a culture of trust, where people trust each other, where they are not stabbing each other in the back all the time trying to get ahead.
"That culture makes it possible for us to continue moving forward with the same principles and beliefs."