Author: Ben Kolada

Via Oak Hill, ViaWest expands

Ben Kolada

Contact: Ben Kolada, Jeff Paschke

In its first acquisition under new ownership, ViaWest announced today that it is buying three Salt Lake City datacenter facilities from Consonus Technologies. The transaction is the first in a line of expected deals after the company received a capital infusion from its sale to Oak Hill Capital Partners in April.

Financial terms weren’t disclosed; however, our analysts at Tier1 Research estimate that the transaction was in the range of $35-40m. The acquisition brings ViaWest’s total Salt Lake City datacenter count to six, and comes just three months after the vendor announced that it was adding 14,000 square feet to one of its existing Salt Lake City facilities.

Indeed, even at the high end of our estimated range, $40m may seem to be a bargain for three facilities with a combined total of 100,000 square feet of datacenter space. However, we would point out that ViaWest is not in the business of owning the shell building, but rather just the internal infrastructure (generators, switches and all furniture and fixtures). Its facilities are generally maintained on a long-term lease basis.

Under Oak Hill’s wing, we expect ViaWest to continue to acquire additional properties. (The firm inked three other deals under previous ownership, according to The 451 M&A KnowledgeBase.) Given its past acquisitions, ViaWest appears to prefer buying into new markets, rather than acquiring properties in markets where it already has a presence. As such, we don’t expect to see another market-consolidating play anytime soon. The company has a strong foothold in the western US – Salt Lake City follows Denver and Dallas as the firm’s third-largest market, in terms of usable datacenter space – but ViaWest executives have told us in the past that acquisitions will not necessarily be limited to the western states.

PAETEC’s risky business

Ben Kolada

Contact: Ben Kolada

As the communications industry continues to consolidate and the pool of desirable targets dries up, the remaining buyers appear to be stretching a bit in their M&A moves. But even within that, PAETEC’s recent pickup of Cavalier Telephone looks to us like the riskiest telecom acquisition we’ve seen in the past year. The reason? Roughly three-quarters of Cavalier’s business is outside PAETEC’s focus.

To be fair, other telcos have also made challenging moves. Windstream Communications took big bites in the past 12 months, acquiring four companies that set the telecom provider back $2.7bn. (That figure includes the debt at the acquired companies that Windstream will be taking on.) The vendor’s spree boosts its top line by about 50%, a substantial increase that brings a not-insignificant amount of risk. Even Cablevision Systems, which is typically a stay-at-home company, inked a deal, reaching across the country to pick up Bresnan Communications for about $1.4bn.

However, the deals by Windstream and Cablevision made sense, if just because they expanded on each company’s existing strategy. Not so with PAETEC’s purchase of Cavalier. When we look at the transaction, we suspect that PAETEC was really only interested in Cavalier’s fiber assets. Understandably, the Richmond, Virginia-based competitive local exchange carrier wouldn’t have considered selling its fastest-growing division. Since it was unable to just get the part of Cavalier’s business that it probably wanted, PAETEC was forced to shell out $460m (including assumption of debt) for the whole company.

Cavalier had $390m in sales in the year leading up to the acquisition. However, the company’s fiber division itself generated only about $98m, or 25%, of total revenue. That means that a vast majority (75%) of Cavalier’s business appears to us to be an ungainly match to the business its buyer is in. PAETEC serves enterprises, which generate an average of $2,300 in monthly revenue. On the other hand, the majority of Cavalier’s revenue comes from consumer accounts and small businesses with monthly recurring revenue of only about $500.

Rather than spend to get this odd pairing, we think PAETEC would have been better off buying one of the number of fiber operators looking for a sale. A juicy target would have been Zayo Group. The company is on a $240m run rate for 2010. Based on recent valuations for Zayo’s competitors, we believe it could be had for roughly $500m – only slightly higher than Cavalier’s price tag, but without the unwanted baggage.

Private equity goes back to the hosting table in a big way

Ben Kolada

Contact: Ben Kolada

So far this year, three private equity (PE) firms have each shelled out at least $400m for a hosting provider, making 2010 the most active year for big-ticket hosting deals for PE shops. And these firms are no novices. Welsh, Carson, Anderson & Stowe, GI Partners and Oak Hill Capital Partners have a combined $32bn in capital under management, and each has had previous experience in the hosting sector. The fact that they’re coming back to the hosting market – and paying relatively rich valuations to do so – is a hearty endorsement of the sector’s long-term growth potential.

In the most recent deal, Welsh Carson teamed up with Peak 10 management to buy the company from Seaport Capital and McCarthy Capital. Although terms of the transaction weren’t disclosed, we understand the buyout consortium paid just north of $400m for Peak 10, or about 12 times the company’s annualized 2010 EBITDA. For comparison, Savvis, in which Welsh has been invested since 1999, is currently trading at 5x annualized EBITDA.

In another management buyout, SoftLayer Technologies’ management announced in August that it was partnering with GI Partners to buy the dedicated hosting specialist from its angel investors. Again, terms weren’t disclosed, but we believe the deal valued SoftLayer at about 10x its annualized EBITDA, or about $450m. As my colleagues Philbert Shih and Aleetalynn Schenesky-Stronge noted, GI Partners is a well-known participant in the hosting and Internet infrastructure space, having invested in Digital Realty Trust and The Planet. GI Partners intends to combine The Planet and SoftLayer, with SoftLayer’s management left in charge. The combined company, which would have $270m in estimated revenue for 2010, could go public as early as next year.

SoftLayer was GI Partners’ second hosting play of the year. In April, the firm banded together with Oak Hill Capital and ViaWest’s management to buy the company from a consortium of PE investors. Oak Hill Capital was the lead investor, with GI Partners and management retaining minority stakes. We estimate the price of the deal at $420m, which works out to about 10x ViaWest’s cash flow. Oak Hill Capital isn’t new to the datacenter industry, having previously invested in TelecityGroup.

More PE moves could be in the works, as we’re aware of quite a few more properties for sale. If the flurry of M&A activity during the recent VMworld conference is any indication of what happens when a group of likeminded individuals gets together, our 2010 Hosting & Cloud Transformation Summit could lead to a number of hosting and Internet infrastructure deals. The conference opens today in Las Vegas and continues through Wednesday.

Select PE hosting deals in 2010

Date announced Acquirer Target Deal value
September 1 Welsh, Carson, Anderson & Stowe/Peak 10 management Peak 10 $400m*
August 4 GI Partners/SoftLayer Technologies management SoftLayer Technologies $450m*
April 20 Oak Hill Capital Partners/ViaWest management ViaWest $420m*

Source: The 451 M&A KnowledgeBase *451 Group estimate

Winners and losers in data warehousing

Ben Kolada2 Comments

Contact: Ben Kolada

Just a month after Greenplum was swallowed by EMC for an estimated $400m, fellow data-warehousing startup Kickfire was sold for probably one one-hundreth of that amount to Teradata. Why did the two data-warehousing vendors – both venture-backed, Silicon Valley startups targeting the same market – see divergent outcomes? The answer to that multimillion-dollar question lies in each company’s targeted markets.

The scrap sale of Kickfire was the end result of a misguided approach by the Santa Clara, California-based startup to the low end of the data-warehousing market. Basically, Kickfire was trying to sell appliances through an expensive direct-sale model. However, the economics of a high-cost business model for a low-cost product only work on big sales. Kickfire never got anywhere close to that, collecting only about a dozen customers in its four years of business. (We would contrast Kickfire’s business model with that of its closest competitor, Infobright. That company, which sells a software-only product through an indirect channel, has more than doubled the number of customers over the past year to 120.)

As Kickfire was struggling to sell to small businesses, 30 miles up the road in San Mateo, California, Greenplum was ripening nicely by selling to enterprises. The company’s high-revenue customer accounts helped it quickly grow total sales to just shy of $30m at the time of its sale to EMC. (That works out to an eye-popping valuation of 14 times trailing sales – a multiple that’s twice as high as any valuation the data-warehousing sector has seen in major acquisitions.) Part of the reason it garnered such a high price is that Greenplum counted some 140 customers at the time of its sale.

Other data-warehousing vendors have also experienced the highs of the enterprise market. Netezza and Teradata both made it to the public markets. (Although we heard a rumor that Netezza was almost erased from the market. Word is that EMC first talked to Netezza, even floating a bid earlier this year that basically would have valued Netezza at its current price on the NYSE. Needless to say, talks didn’t go too far between the two Boston-area companies.) And of course, DATAllegro was scooped up by Microsoft for an estimated 7x trailing sales.

With all of this consolidation playing out, we expect that much of the attention in the data-warehousing space is now turning to Aster Data Systems. The fast-growing vendor, which is based in San Carlos, California, has raised $27m in venture backing. If Aster Data gets snapped up in a trade sale (like many of its rivals have), we wouldn’t be surprised to see Dell as the buyer. The two companies are currently partners, and Dell has shown an increasing interest in big data following its continued attempts to buy 3PAR.

Storage sector M&A holding steady

Ben Kolada

Contact: Ben Kolada, Henry Baltazar

In its eighth storage play, IBM announced last week that it is acquiring data compression vendor Storwize. The move, which came quickly on the heels of Dell’s purchase of data de-duplication provider Ocarina Networks, brings the number of storage deals we’ve tallied in 2010 to 19. That’s roughly on par with the volume of storage transactions in the same period last year.

Of course, deal flow in the sector last year was dominated by a bidding war over Data Domain, which sold to EMC for $2.3bn after NetApp put the data de-dupe specialist in play but then got topped. We would note that EMC – the most active acquirer in the storage industry, having picked up some 15 storage companies over the past eight years – has been out of the storage market since it bought Data Domain. However, the storage giant may figure into the industry’s most recent deal. What do we mean?

Big Blue’s purchase of Storwize appears to be a reaction to EMC’s announcement in May that it was adding compression to its midrange Clariion and Celerra platforms. (The Storwize deal was first rumored in June, just after EMC’s announcement.) Storwize is unique in the storage space because it offers real-time data compression of up to 80%. Further, my colleague Henry Baltazar claims that IBM has already been working with Storwize for about a year. Storwize’s appliances run on System x servers, which Big Blue points out should ease the integration process – and help it to match the competitive moves by rival EMC.

What’s up with the Bay Area?

Ben Kolada

Contact: Ben Kolada

Bay Area buyers have roared back to life in 2010. Compared to the same period a year ago, Bay Area buyers’ deal volume has increased 46%, while at the national level M&A has risen only 21%. Year-to-date, Bay Area-based acquirers announced 230 transactions, 19% of all technology deals undertaken by US-based companies. Further, these companies represent 19% of the total declared deal amount, including four of the 18 billion dollar-plus transactions made by US-based buyers. In the same period last year, Bay Area acquirers did only 162 deals.

So, what’s up with the Bay Area? Our data suggests that 15 big serial acquirers accounted for most of the increase. In fact, the number of Bay Area buyers acquiring three or more companies increased five-fold in 2010, compared to a 50% increase at the national level. After waiting on the sidelines in 2009, these companies have resumed M&A activity in full force. As a group, they bought 52 more companies in year-to-date 2010 than they bought in 2009. (An interesting note, Internet content providers were the preferred targets across the board, representing 22% of acquired companies at both the Bay Area and national levels.)

M&A activity by Bay Area buyers

Acquirer 2010 deal volume, year-to-date 2009 year-ago period
Google 15 0
Oracle 7 5
Playdom 6 0
Apple 4 0
Facebook 4 0
Symantec 4 1
Synopsys 4 1
Trimble Navigation 4 5
Cisco Systems 3 3
Hewlett-Packard 3 2
TIBCO Software 3 0
Twitter 3 0
VMware [EMC] 3 0
Yahoo 3 0
Zynga 3 0
Totals 69 17

Source: The 451 M&A KnowledgeBase, 451 Group research

Google is the poster child for Bay Area M&A. Year-to-date, the company has been involved in 15 transactions – the most since it inked the same amount of deals in full-year 2007. However, the search giant is noticeably absent from the 2009 ranking. Even though Mountain View, California-based Google had $8.6bn in cash at the end of 2008, the vendor took nearly a year-long break from M&A activity. Google’s M&A drought began after it acquired TNC in September 2008 and ended 11 months later, when it announced its first purchase of a public company – On2 Technologies – in August 2009.

Nokia hiring by acquiring

Ben Kolada

In an unusual bit of dealmaking, Nokia bought geo-tagging vendor MetaCarta in April and then turned around and sold it three months later. The recent divestiture might appear to be a botched acquisition. However, as we look closer at the deal, it turns out that Nokia actually got what it wanted out of the purchase. It is retaining MetaCarta’s engineering team while shedding its enterprise accounts to Qbase. (Nokia didn’t really have any use for the startup’s enterprise business, which was largely oil and gas industry as well as government installations.)

Cambridge, Massachusetts-based MetaCarta employed approximately 20 development engineers, plus 15 enterprise sales and support staff. Although terms of the deal weren’t disclosed, we understand that Nokia paid about $30m for MetaCarta. If we look at the price in terms of what assets Nokia actually wanted to obtain, we pencil it out at about $1.5m per engineer. This is obviously an expensive way to recruit personnel, and underscores the increasing pressure that Nokia is seeing in the mobile-mapping space.

Nokia ‘hired’ MetaCarta’s engineers to reinforce the search feature in Ovi Maps, Nokia’s most popular application. MetaCarta is a specialist in geo-tagging unstructured text such as websites and emails. While mapping competitor Google does the same, MetaCarta’s information will be layered on NAVTEQ’s mapping data, which is arguably more detailed than Google’s maps.

The transaction is another in the long line of acquisitions that Nokia has made in its move toward mobile advertising. However, Nokia’s rivals have also been active in the mobile M&A space. Research In Motion reached for GPS vendor Dash Navigation in June 2009. In November 2009, Google outbid Apple and bought AdMob for $750m. In response, two months later, Apple picked up Quattro Wireless for an estimated $275m. Nokia hasn’t made a purchase of this magnitude, but we still believe it could be on the hunt for additional mobile providers. The company could build on its MetaCarta acquisition by buying location-based advertising vendor 1020 Placecast. The San Francisco-based firm is a major strategic partner of Nokia’s NAVTEQ, and would supplement MetaCarta’s geo-tagging capabilities.

Cablevision breaks the mold with Bresnan acquisition

Ben Kolada

Contact: Ben Kolada

Marking a significant departure from its recent practice, Cablevision Systems said last week that it would hand over almost $1.4bn in cash and stock for Bresnan Communications. The deal by the Dolan gang is their first major telecom acquisition since they picked up a portion of Tele-Communications in 1998. And they certainly paid up for Bresnan.

Cablevision’s offer values Bresnan at about 3.4 times trailing revenue and just over 8x projected 2010 cash flow, according to our understanding. On a per-subscriber basis, the acquirer is paying $4,500 a head. Across the board, that’s a far richer valuation – in some cases, twice as rich – than fellow telco RCN got in its take-private in March. Buyout shop ABRY Partners paid $1.2bn for RCN, or roughly 1.6x trailing revenue (on an enterprise value basis) and $2,800 per subscriber.

Given the size of this deal, along with the fact that Cablevision used equity in the purchase, we don’t expect the Bresnan transaction to signal the beginning of a shopping spree. Indeed, Cablevision executives maintain that they are not looking for additional properties. After all, Cablevision doesn’t need to buy more systems – the Tele-Communications acquisition gave it sufficient economies of scale. The Bresnan buy is simply a rare opportunity to obtain upgraded systems with strong growth potential.

M&A: Cable comparison

Date announced
 Acquirer Target Deal value Price-to-sales multiple
March 5, 2010 ABRY Partners RCN $1.2bn 1.6
June 14, 2010 Cablevision Systems Bresnan Communications $1.4bn 3.4*

Source: The 451 M&A KnowledgeBase *451 Group Estimate