Going vertical makes machine learning deal volume horizontal

by Scott Denne

Last year marked a banner year for machine learning M&A. Purchases of vendors building products around that emerging technology, as well as firms whose offerings enable others to build machine learning products, soared in 2019. Yet the change was more than quantitative. As that new technology pervades older industries, it’s bringing in new acquirers and the rationale behind many machine learning deals has evolved.

According to 451 Researchs M&A KnowledgeBase, acquisitions of machine learning providers rose 75% to 278 in 2019. In other words, machine learning accounted for one of every 13 tech transactions last year. Within that number, purchases of companies selling into a specific vertical produced the most prominent rise as buyers nabbed 112 vendors building machine learning offerings that address problems within a particular industry, more than double the 53 they bought earlier.

In previous years, many acquirers were content to pick up general-purpose tech and expertise to bolster their machine learning portfolios. Serial shoppers Microsoft, Google, Salesforce and Apple, for instance, each picked up more than 10 businesses in recent years to expand their products’ machine learning functionality. As we previously noted, buyers have already become more discriminating, and more likely to seek tech that fills specific gaps in the product portfolio.

But the application of machine learning to certain markets has not only kept acquirer interest in the burgeoning technology high, it has also brought new buyers into the tech M&A market as staid industries see opportunities and risks that come with the emergence of machine learning. For example, in one of the largest machine learning deals last year, Prudential Financial printed its first tech acquisition with the $2.4bn pickup of Assurance IQ to expand its online sales via software that customizes insurance products around data signals.

As the universe of buyers and the range of rationales for machine learning transactions continues to expand, this technology will likely continue to have an outsize impact on the tech M&A market. At least that’s the take from the 451 Research Tech Banking Outlook, where 91% of bankers surveyed anticipate that machine learning purchases will accelerate in 2020.

Figure 1:

Source: 451 Research’s M&A KnowledgeBase. Includes disclosed and estimated deal values.

A softer software market

by Scott Denne

The rising tide of 2018’s record software M&A market gave way to a top-heavy 2019 as fewer business application vendors fetched a 10-figure price but a record number nabbed 11 figures. And while the size of the market shrank a bit from last year’s high as both strategics and sponsors slowed their spending, the total stands well above the deal value and volume seen in a typical year. Despite the decrease in overall deal value, acquirers continued to pay premium multiples for those assets they did purchase, something they’re less likely to do in the new year.

According to 451 Research’s M&A KnowledgeBase, acquirers in 2019 bought a record 1,296 application software companies worth $83.2bn, about $12bn less than the previous year. A dramatic drop in transactions valued north of $1bn led the decline in spending. Yet the biggest software deals got bigger. Two acquisitions – Salesforce’s $15.5bn reach for Tableau and Hellman & Friedman’s $11bn Ultimate Software take-private – passed the 11-figure mark in 2019, something that only one other application vendor (Autonomy) had done since the start of the decade.

While only half as many application providers sold for $1bn or more in 2019 as did a year earlier, the 15 transactions that went off in 2019 stand as the second-highest total, according to the M&A KnowledgeBase. In other words, it may be more accurate to view 2019 as an inevitable decline that followed a sudden surge in 2018. In each year since 2010, we’ve consistently recorded about 12 application software deals valued at or above $1bn, meaning that 2019 was notably high, yet well below the 30 we saw in 2018.

And while the number of $1bn software transactions shrank, prices didn’t. The annual median for such deals finished a hair above 7x trailing revenue, the same level as the previous year, our data shows. Who was willing to pay those prices shifted in 2019, however. While private equity firms paid that same multiple in 2018, the median price paid by sponsors to acquire software companies for 10 figures declined a full turn last year. Strategic buyers moved in the other direction, paying a 9.3x median as they continued to pay higher prices in each of the past few years.

Although valuations for application software vendors held up amid a decline in overall spending last year, prices could well begin to fall in 2020. That’s the outlook of respondents to our M&A Leaders survey in October, where they projected a drop in private company valuations at a rate of five to one. Moreover, 2020 starts off with a tougher macro environment than its predecessor. In that same survey, nearly two-thirds of respondents (61%) said that ‘fear of a recession’ could weigh on deal flow in the coming 12 months.

Figure 1:

Source: 451 Research’s M&A KnowledgeBase. Includes disclosed and estimated values.

The end of Omniture’s overture

by Scott Denne

This past weekend marked 10 years since Adobe’s $1.8bn purchase of Omniture, the deal that arguably started the race among enterprise software vendors to build out customer experience software portfolios. Although that transaction marked the beginning, its decennial looks like the beginning of the end. While we tracked a record haul for customer experience software M&A in 2018, those companies are becoming increasingly harder to sell.

According to 451 Researchs M&A KnowledgeBase, buyers spent $29.6bn in 2018 on vendors developing software for advertising, marketing, customer service, e-commerce and other forms of customer engagement. So far this year, just $8.3bn has been spent on such targets, on pace for the lowest annual total since 2015. Moreover, we’ve seen just five companies in this space sell for $200m or more, while each of the six previous full years have seen at least 10 deals of that size.

And the multiples on those acquisitions have fallen dramatically, our data shows. Last year almost every vendor in the category selling for that amount blew past the 5.2x trailing revenue that Omniture commanded. This year, however, such transactions fetched a median valuation of 2.3x, which is at least a full turn lower than the median valuation of similar deals in any single year over the past decade. This year, Dynamic Yield and TrendKite nabbed north of 8x in their respective sales to McDonald’s and Cision, while none of the other $200m cohort printed above 2.5x.

There’s little doubt that Adobe has seen success with its Omniture buy. The company expects to grow its Digital Experience business, the unit that houses Omniture (now Adobe Analytics) and several other related targets, by 23% to $3bn in the soon-to-close fiscal year. But other early, marquee investments in this sector weren’t as successful and there may not be as many deep-pocketed buyers as there once were. Both IBM and Teradata, for example, shed their marketing software units. Meanwhile Oracle, which still ranks as the most prolific acquirer in the segment, only printed one deal last year and none in almost 18 months.

Figure 1:

High on low-code

by Scott Denne

The value of acquisitions in the low-code application development software market is rising. With Temenos’ $559m purchase of Kony, we’ve now recorded more deals and higher total value in this corner of the software market than all of 2018.

In reaching for Kony, Temenos, a developer of banking software, gets both a generic low-code tool and a portfolio of prebuilt digital banking applications. Although few low-code acquisitions we’ve tracked are vertically specific, applications developed with these tools often replace vertical-specific applications. That has helped bring private-equity investors, which have demonstrated an affinity for vertical software companies, into the space. Sponsors have printed three of the last five low-code vendor purchases.

Since these tools are often the foundation for multiple applications within an enterprise, they tend to have low churn rates – something that appeals to both strategic acquirers and sponsors. Kony, for example, has about a 5% attrition rate. So far this year, $1.7bn have been spent across four acquisitions in this market, according to 451 Researchs M&A KnowledgeBase. That’s up from $1.3bn in three deals in 2018.

Part of the reason for the rise is that buyers are reaching for larger targets. Kony projects its topline will grow to $120m in 2020. QuickbaseNintex and Mendix were all nearing or above $100m in their recent sales. (Subscribers to 451 Research’s M&A KnowledgeBase can access our estimates of those transactions by clicking on the links in the company names.)

Acquisitions of low-code app development vendors (includes disclosed and estimated deal values)

Middling exits in sales analytics

by Scott Denne

Companies developing predictive analytics for sales teams have done a poor job of predicting their own exit opportunities. In this corner of the sales-software market, several companies have exited, although most appear to be ‘acqui-hires,’ including the most recent deal, Anaplan’s acquisition of Mintigo.

Although Anaplan didn’t disclose the terms of its first acquisition as a public company, we expect the total came in below the $50m that Mintigo raised from investors. Anaplan only disclosed the acquisition during its earnings call, emphasizing that the purchase was done to land the target’s 50 employees, not for its B2B sales software. That would be a familiar outcome for the half dozen or so companies that launched earlier this decade to develop predictive analytics for B2B sales.

In 2015, LinkedIn acquired Fliptop to bolster the development team around its Sales Navigator product; a year later, eBay picked up the team that developed the now-defunct SalesPredict product; and in 2017 ESW Capital, a bargain-hunting PE firm, scooped up Infer. The exception, so far, is Lattice Engines, a growing business that sold to Dunn & Bradstreet at a respectable multiple (subscribers to 451 Researchs M&A KnowledgeBase can access our estimate of that deal here).

For the remaining vendors in the space, the exit potential looks a bit brighter. Most have evolved, if not outright pivoted, beyond stand-alone sales analytics. Everstring relaunched a little over a year ago as a provider of business data, 6Sense is expanding into a marketing suite on top of its intent data, and Leadspace is moving into sales analytics from its foundation of sales data management. Topline growth at these companies could compel business data providers or enterprise software companies to make more strategic acquisitions of sales analytics than we’ve seen so far.

Microsoft jumps into retail media

by Scott Denne

The sudden surge of Amazon’s advertising business has sparked acquisitions of software companies that help retailers become publishers. Microsoft is the latest entrant with the purchase of PromoteIQ, a maker of software that enables retailers to run product ads. As audience monetization becomes a feature of more e-commerce businesses, more deals could come.

For Microsoft, the pickup of PromoteIQ (formerly known as Spotfront) has overlap with its search advertising business. Product ads, often delivered alongside e-commerce search results, are essentially an evolution of paid search advertising. The target’s software provides retailers with workflows and controls to manage their sponsored product listings.

Amazon, more than any other company, has realized the potential for retailers to monetize their apps and websites through sponsored products. As we noted in a recent report, the online retail giant’s advertising business has tripled in less than two years to about $10bn in annual revenue. Despite that growth, the category is nascent enough that there’s not yet a widely accepted name for it. PromoteIQ refers to it as ‘vendor marketing,’ while Criteo and Triad, two of its larger rivals, call it ‘retail media.’

Given the early stages of the market, there are few sizeable players remaining for would-be buyers. Several midsized firms such as Adzerk, Crealytics, Koddi, Playwire and SYNQY offer retail media products. Another vendor, OwnerIQ, enables retailers to monetize purchase intent data gleaned from shoppers on their sites. As Amazon’s advertising revenue continues to balloon, targets in this space could get a look from acquirers in search (Google, Pinterest), e-commerce software (BigCommerce, Shopify) and retail-focused advertising (Quotient, Valassis).

No hard turns for application software market

by Scott Denne

Both strategic acquirers and sponsors have increased their purchases of application software vendors as the market for those targets accelerates beyond last year’s record. As those buyers stayed active, the largest deals grew larger and multiples continued to climb. Rising valuations for growth companies in the public markets – both new offeringsand already-public businesses – have pushed up pricing for software targets and could help propel deal sizes through the rest of the year.

According to 451 Research’s M&A KnowledgeBase, 632 application software providers have been acquired for a combined value of $59.2bn, on pace to top last year’s record haul in both value and volume. In our analysis of last year’s activity (published as part of 451 Research’s Tech M&A Outlook 2019), we noted that 2018 marked the first year that acquisitions of application software vendors crossed 1,100 as both strategic buyers and private equity firms accelerated their purchases.

The same trend has driven this year’s market as well. Both categories of acquirers are up from last year’s pace. Yet the number of big-ticket transactions is down a bit. While 29 application software providers sold for $1bn or more in 2018, only nine have done so this year. But those that have crossed the 10-figure mark have done so in a big way. Since the bursting of the dot-com bubble, only four such targets have sold for more than $10bn and two of them (Tableau Software and Ultimate Software) did so this year.

To be clear, the relative decline in the number of big-ticket acquisitions hasn’t pulled down the total deal value. If the current pace of deal value were to hold through the end of the year, it would finish more than one-quarter higher than last year, which smashed the previous record by nearly 50%. Put another way, at the midpoint of 2019, the total value of application software transactions is higher than all but two full years, our data shows.

Although fewer software vendors have sold for north of $1bn, those that have are fetching higher prices. According to the M&A KnowledgeBase, the median multiple for those deals this year stands at 8.1x trailing revenue, nearly a full turn above last year’s total. That rise comes amid a 22% year-to-date increase in the Nasdaq index and a welcoming environment for new tech listings, including software companies like Slack and Zoom Video, which commanded 62x and 41x multiples at the midpoint of the year, following their recent public market debuts. Such a compelling alternative exit could continue to boost acquisition prices.

Software buyers surge

by Scott Denne

In 2018, it seemed that anyone who could buy a software company did. The reemergence of software M&A’s regular strategic acquirers, along with a continuing surge in private equity (PE) deals and a jolt from young public software companies, pushed the market for application software acquisitions well past its 2016 record.

According to 451 Research’s M&A KnowledgeBase, acquirers picked up 1,191 application software companies worth a combined $90.5bn. This marks the first time we’ve recorded more than 1,100 software companies sold in a single year, and represents a 42% increase over the previous record (from 2016) for total value of acquisitions in the category.

While there were certainly outstanding deals – SAP’s $8bn purchase of Qualtrics being the largest – it was the volume of acquirers willing to make big purchases that drove the year’s total to its heights. Since the end of the dot-com bubble, there have only been two years that have had more than 10 application software deals to print at or above $1bn, and never more than 14. This year, buyers announced 28 of them.

SAP bookended the year with a pair of such transactions: the $2.4bn acquisition of Callidus Cloud and the $8bn purchase of Qualtrics in November. Prior to that, you have to go all the way back to SAP’s 2014 acquisition of Concur for its last 10-figure deal. It wasn’t the only one of its peers to return to market in a big way this year. Adobe printed two $1bn application software deals (Magento and Marketo) – its first since 2009. At the same time, Cloudera, Twilio and Workday all printed their first-ever 10-figure deals.

Vertical software targets – companies building software for specific markets, such as construction, pharmaceuticals, education and government – took an outsized share ($30.4bn) of the annual total. Those targets continued to be a private-equity favorite, while attracting interest from companies outside the software industry. Roche, the pharma giant, opened the year with a $1.9bn purchase of electronic health records company Flatiron Health, while industrial conglomerate 3M closed the year with a $1bn acquisition of clinical documentation software developer M*Modal.

Even amid the return of the old strategics and the emergence of new cohorts of software buyers, PE continued to influence the software M&A market. More than one in three (37%) application software acquisitions was made by either a PE firm or a PE portfolio company, up from 33% in 2017 – the highest level up to that point. Still, the reemergence of strategic acquirers helped push up prices on those deals. The median multiple paid by a PE firm for a software company last year came in at 4.1x trailing revenue, clocking in above 4x for the first time on record.

Across the software market, rising multiples spurred the growth in deal value. Buyers paid a median 7.6x trailing revenue across $1bn-plus deals in the category last year, compared with each of the four preceding years, where median multiples on such deals fluctuated between 3.9x and 4.8x.

A bullish bet in a bearish market

by Brenon Daly

The recent rout of technology stocks didn’t actually provide much of a discount in Big Blue’s big bet on Red Hat. In the largest-ever software acquisition, IBM is valuing the open source software provider at its highest price since its dot-com-era IPO. Essentially, Big Blue had to make up Red Hat’s recent stock decline with a very generous premium. (Subscribers to 451 Research can look for our full analysis of the transaction and its implications on our website later today.)

Ahead of the announcement, shares of Red Hat had shed about one-quarter of their value just since mid-September. The stock bottomed out last week at about $117, its lowest level in a year. Under terms, IBM is paying $190 for each Red Hat share, which works out to a premium of more than 60% from the prior close. That’s about twice as rich as the typical premium in a significant software acquisition.

However, looking at the terminal value of a company relative to the market value of a company doesn’t make too much sense unless we also factor in the state of the overall market. Stock prices change every day, particularly for high-beta stocks like Red Hat. It just so happens that in recent sessions on Wall Street, virtually all of the changes have been marked in red.

In the case of Red Hat, it was riding high last summer, with shares peaking at about $177. At that level, IBM is paying a scant 7% premium on Red Hat’s market value. (That fact hasn’t been lost on plaintiff lawyers, who have already revved up their strike-suit machine to target this deal.)

Rather than comparing how IBM is valuing the company to how public market investors value the company, it’s more useful to look at how IBM is valuing Red Hat’s actual business. And by that measure, this is a pricey pairing.

IBM, which trades at less than 2x trailing sales, is valuing Red Hat at more than 10x trailing sales. That’s substantially higher than the average multiple of 6.6x trailing sales for the 10 largest software acquisitions recorded in 451 Research’s M&A KnowledgeBase.

HR software in demand

by Scott Denne

A tight job market is opening exit opportunities for HR software companies. The record-low unemployment rate – it recently dipped below 4% for the first time in more than a decade – has increased interest in owning software developers that help businesses find, retain and train increasingly scarce employees, pushing such acquisitions to a remarkable level.

In the latest example, Cornerstone OnDemand has reached for Workpop, a provider of software for hiring hourly and seasonal workers, as part of the buyer’s revamp of its recruiting suite. As we noted at the time of that company’s last acquisition – back in 2014 – Cornerstone OnDemand hasn’t been much of a buyer, although in making a purchase now, it’s joining a parade of dealmakers scooping up HR software targets.

According to 451 Research’s M&A KnowledgeBase, acquirers have hooked 70 HR software targets, more than any other full year in this decade. Still, in terms of deal value, 2018 isn’t likely to be a record. The year’s total stands at $2.35bn, while two earlier years (2012 and 2014) saw more than $5bn in HR software transactions.

While 2012 and 2014 each had a pair of $1bn-plus deals by Oracle, IBM, SAP and Charthouse Capital, this year’s boom has benefited midmarket targets as we’ve recorded 13 acquisitions valued at $100-500m, two more than any other full year. Although the jobs picture helps juice this market, much of the increase comes through the same trends that bolster the overall software M&A market – increasing activity from private equity firms and a surge in strategic buyers. The former category has already purchased more HR software businesses this year than ever before, while corporate acquirers are heading toward record territory.

For more real-time information on tech M&A, follow us on Twitter @451TechMnA.