Recently, Kerrisdale Capital Management published an analysis of ServiceNow. The opening sentence made the tone of the report quite clear, “We believe that ServiceNow, Inc. (“NOW” or the “Company”) is highly overvalued. We are short the stock.”
Nevertheless, the analysis did end with recognition of ServiceNow’s impressive growth by congratulating the company for creating a niche IT Service Management product. Interesting though, is that the congratulation included a comment that the product was created with only $40 million of R&D.
The ServiceNow success is very impressive. They saw an underlying weakness among legacy vendors, in a space that was crying out for something new to free them from the expensive and highly unproductive cycle of increasingly difficult upgrades to decades old on-premise products. The largest customers were hurting the most and ServiceNow set their sights on these giant prizes (As per the ServiceNow S-1, “We target our sales efforts at large enterprises.“ and “A majority of our revenues come from large, global enterprise customers.”). Possibly through concern that the large enterprises would not easily embrace a multi-tenant environment (I can’t think of any other plausible reason), ServiceNow elected to build a “multi-instance” architecture (more commonly known as “single-tenant.”).
With this approach, ServiceNow was successful in replacing the “Big Four” legacy vendors in a signifcant number of large enterprises. But, the “multi-instance” model and the fact that ServiceNow has some on-premise customers (the architecture allowed them to do this), has lead to ServiceNow having all of the challenges that the replaced vendors have struggled with. The basic fact that ServiceNow has hundreds of instances makes them look and act more like an older legacy vendor than a modern SaaS vendor. Each new customer brings new expense. In addition, this complex management task means that ServiceNow has also been forced to only provide a “new release” every six months — not quite matching the rapid innovation cycles of multi-tenant providers where multiple releases every month is common.
Add the acknowledged complexity of implementation and we can start seeing the growth problems that ServiceNow is faced with. As per the S-1 and the Kerrisdale report, “Should a customer require customized functionality, as many do, it can take 6-8 months and tens of thousands of dollars in implementation costs to get NOW’s system up and running.” These implementation barriers put pressure on ServiceNow margins as can be seen in the Professional Services loss of $4.2M posted over the past twelve months.
What then can we take from this hard-hitting report? Firstly, let’s not dump completely on ServiceNow. Most of us in the SaaS world are envious of their growth and success. But, it came for all the wrong reasons when you look at it through the eyes of true SaaS customer promise. ServiceNow saw an opportunity — the Big Four had ignored the complaints from their largest customers for too long and there is no question that they were damaged. I submit that ServiceNow will continue to hurt them, even if it becomes increasingly difficult to do so.
What we can also conclude is that ServiceNow showed that an offering superior to the decades-old on-premise products will get the attention of large enterprise customers. Not surprisingly, today it is also evident that this is true of the mid-market and even the SMB market. Actually, the pain felt by these customers may be more severe than that of the much larger organizations. The resulting emergence of a small number of new multi-tenant SaaS vendors has created a substantial opportunity for companies in these segments to free themselves from the never-ending upgrade cycle while moving to solutions that are so much easier to implement and use — all at a total cost of ownership that in many cases is 50%-70% lower.
Some of the numbers in the report also surprised me. Kerrisdale reported that BMC had grown its RemedyForce customer base by 160 in six months, to a total of 475 SaaS customers. This from an industry giant and with the help of the sales team at SalesForce.
One final, important point — the Kerrisdale analyst gets it when it comes to the fact that multi-tenant is important to customers. Here is what he said:
“The differences between single and multi-tenant SaaS are subtle but matter when forecasting the future direction of the industry. Single-tenant SaaS generally refers to fully-customizable software that can vary by customer (including a separate database instance for each application); ServiceNow’s customizable code firmly places them in this category. Some of the independent experts that we’ve spoken with believe that a multi-tenant architecture, or one in which the underlying software is identical across customers, is a better way forward for the SaaS industry. Without variation in the underlying code, multi-tenant SaaS makes regular software updates much simpler for all end users. This issue is significant since many of NOW’s customer wins have come from clients who have been frustrated with the incumbents’ cumbersome upgrades, and are looking for predictable, straightforward software updates. Yet a single-tenant system would typically involve complex upgrades given that different clients have customized their NOW platforms with functionality specific to their operations. Additionally, single tenancy usually involves substantial upfront coding costs and longer implementation times. For these reasons, NOW may lose certain customers over time to SaaS providers who offer more attractive multi-tenant solutions.”