[WDAY, CDAY, PAYC, ULTI, ORCL, SAP] Workday and Competitive Dynamics in HCM

Workday’s commanding position in enterprise HCM partly traces back to the design choices it made at the very start.  At the time, behind most enterprise software products were long bodies of code interacting with huge relational databases (columns and rows that store customer data) made up of thousands of tables.  The complexity of the application was tied to the complexity of the database, and modifications to the application architecture mandated expensive and time-consuming changes to thousands of underlying tables.  Rather than have a database structure determine the structure of its program, Workday’s architecture conformed to an Object-Oriented model¹, whereby applications were defined by “objects” – concepts with real world meaning such as employees, budgets, positions, benefits, etc. – and relationships between those objects.  With the application abstracted away from the database, a developer could change the former without restructuring the latter.    

The flexibility afforded by this modular configuration shortened development cycles and made it easier to alter programs and build new features.  Workday was introducing three new major updates every year at a time when competitors were launching ~one every two years.  Moreover, combining data and logic into objects and managing these objects “in-memory” (that is, closer to the CPU and main memory)² as Workday did, sped up data retrieval and reporting relative to traditional schemas that joined a bunch of tables related to a single object and drew data from disk storage.  Object orientation + in-memory were major advantages to the architectures that competitors were running at the time and laid the groundwork for not only deeper HCM functionality but also more ambitious forays into ERP and analytics/planning, with all that stuff unified and tied to a single transactional system rather than dependent on feeds, batched imports, and integrations between different platforms.

I wouldn’t say that Workday’s architecture, as a static concept, is itself a sustainable moat; but I do think early differentiators like this matter in a space where success begets success.  Rapid product innovation produces happy customers; happy customers create reference points that give rise to broad adoption; broad adoption sustains an ecosystem of partners and system integrators, who support even more adoption³.  Innovative cultures and ambitious product roadmaps attract the best engineers, who make compelling products for a receptive customer base, which attracts the best quota carrying sales people.  It didn’t hurt that in 2005, the established guard, sated on 90% licensing margins and recurring maintenance fees, weren’t taking multi-tenant cloud delivery all that seriously.  The most visible cloud vendor at the time, salesfore.com, generated just $160mn in subscription revenue compared to the $1.3bn of its on-premise rival, Siebel Systems.  Even so, salesforce had just delivered its second profitable year on the back of 80%+ revenue growth (Siebel’s revenue, meanwhile, was flat/declining) and was gaining ground in large enterprises. 

Cloud delivered software makes too much sense to ignore.  Investments in technology, infrastructure, and support, once carried separately by each enterprise in the on-premise regime, are rendered redundant by a single cloud provider like Workday, who can bear those costs on behalf of all its customers, yielding system wide scale economies⁴. The upfront sales and marketing investment that Workday lays out to acquire a customer are recouped by the end of year 1, after which it realizes subscription revenue at 75%-80% contribution margins over a 10-year average life⁵.  Upselling additional modules makes an already lucrative financial arrangement even more compelling. The core HCM product that Workday launched in November 2006 has been supplemented over the ensuring years with payroll processing, performance management (monitoring and measuring employee performance), succession planning (identifying high performers for leadership positions), time tracking (managing employee schedules and labor costs) and other modules. 

Prior to adopting Workday, a large enterprise might have a dozen siloed on-premise point solutions with different security models handling different HR and payroll functions in different parts of the world, presenting multiple network security vulnerability points that are laid further bare when emails – attached with spreadsheets containing sensitive data – act as the primary bridge between these disparate applications.  It’s cleaner and safer and more efficient to onboard employees; loop them into payroll; track their performance; compare performance across teams and team members; and store and analyze data in a single well-integrated system whose architecture flexes to meet future demands.  It’s easier for employees too.  Rather than manage 401(k) allocations through fidelity.com or scrawl out a time-off request – with pen, on paper – it’s easier to just handle these mundane tasks through a single Workday dashboard. 

There are several HCM vendors out there who cater to small/mid-sized businesses.  Some of them – BambooHR, Zoho, Trakstar, ClearCompany – were founded before or not long after Workday.  None of them offer the native payroll integration that seems pivotal to winning larger engagements and their core HCMs are not as comprehensive as Workday’s (Trakstar is geared towards performance reviews and goal tracking; Zoho is mostly about recruiting).  Intuit has a payroll solution, but its adoption is linked to that of QuickBooks accounting software and so will be relegated to customers that are down market to Workday.  Two notable venture capital backed darlings, Zenefits⁶ and Gusto (formerly ZenPayroll), founded in 2013 and 2011, respectively, do bundle payroll with HCM, but like the older HCM pureplays, they too serve SMBs with products that, though beautifully splashed with clean user interfaces that nail the “consumerization of enterprise software” thing, are not as feature dense as Workday’s.

It’s possible that whatever Zenefits and Gusto have got today is part of an ambitious product development arc that cascades to seven figure deals with large enterprises.  But keep in mind that Workday’s current success in mid/large enterprise – it claims half the Fortune 50 and nearly 200 of the Fortune 500 as customers – traces back to an earlier time, a simpler time.  Workday introduced a superior architecture and delivery model to a space primed for SaaS adoption (after salesforce.com had shown it to be viable in CRM for large enterprises) and was mostly contending with the listless efforts of entrenched on-premise incumbents.  The conditions are different today.  As far as I can tell, neither Gusto nor Zenefits, nor any of the other private HCM vendors I’ve mentioned, are attacking the space with a technology or business model or market approach that is differentiated enough to suggest a serious upmarket challenge to Workday, who is aggressive and on its game in a way that Workday’s 2005 incumbent challengers were not. 

Anyways, in speculating about G/Z’s chances against Workday, I’m sort of shadow boxing, fighting an opponent that doesn’t exist.  I bring all this up only because as a vendor matures and its product capabilities deepen, it often seems that the only way forward is up.  Workday started off selling to companies with 1,000-5,000 employees, always with the aim of migrating to large enterprises; Paycom has been pulled upmarket over the past few years, beyond its target market of companies with fewer than 2,000 employees; Ceridian, whose average customer has fewer than 1,000 employees, has recently begun targeting strategic accounts with workforces greater than 6,000. 

On the other hand, ADP, Paychex, Ultimate Software, and Intuit have been around for decades and have thrived without relentlessly marching upmarket.  Excluding one-person operations, there are ~6mn businesses with fewer than 500 employees in the US.  Zenefits has 7,000 customers (as of 2017) while Gusto has 60,000.  Persistent share donor ADP, by far the leader in this space, serves 570k small business HCM + Payroll clients.  If we accept that serving small businesses requires a different organizational setup and go-to-market approach than serving large enterprises – I can’t think of any instances where a system of record vendor is addressing both extremes of the market equally well (unless you count collaborative productivity apps that get organically adopted from the bottom -> up, but that’s not what this is) – then why reformat a sales and service motion to pursue capable incumbents at the high end when the low-end is so massive and underserved?

Of course, it works the other way too.  The US workforce is both heavily concentrated in a few large enterprises and diluted over many small ones: the 60mn employees working at the 10,000 largest firms in the US nearly rivals the 66mn staffed at the smallest 6mn.  The table below, from the 2016 US Census, breaks down the number of firms according to the size of their workforces.  You’ll notice is that while firms with greater than 5,000 workers (what would be considered “large enterprise”) make up just 0.04% of firms, they comprise 35% of the workforce.  On the other hand, firms with fewer than 20 employees (what would be considered “small businesses”) make up nearly 90% of US firms and account for 47% of the workforce.  So, over 80% of workers are employed at either really small or really large firms.  You can see where I’m going with this.  Compared to the deep labor pools found on the extreme ends, the middle is a desert.  It seems that a software vendor is fated to either concentrate on serving the low-end or migrate incessantly toward the high end. The organizational momentum of a thriving mid-market vendor will push it towards bigger, not smaller, deals. 

Workday, laden with a commission hungry sales apparatus, is not equipped to win $10-$15 monthly subscriptions from a 20-person firm, just as PeopleSoft, organizationally configured around license and maintenance revenue, could not enthusiastically embrace the lower margins, deferred cash payments, and in-house support costs demanded of cloud.  Workday may skirmish with Ultimate and, increasingly, Paycom for mid-sized enterprises (between 2,000 and 5,000 employees) – selling into medium companies is a different enough proposition from selling into large enterprise that in early 2017, Workday divided its salesforce in two, one for each group – but I think the opportunity set is wide enough at the tails to sustain several winners.