New Relic's long term thinking vs Wall Street's short term nerves despite strong Q1

New Relic turned in better-than-expected Q1 numbers yesterday, comfortably exceeding analyst expectations, but a jittery Wall Street took fright at some cautious forward near term guidance. 

The firm reported revenues of $162.59 million for the quarter ended June 2020, compared to $141.01 million for the same period a year ago, with earnings per share of $0.15.

The numbers came less than a week after New Relic repackaged its product line into three categories and overhauled its pricing model. CEO Lew Cirne took time on the post-results earnings call to talk analysts through what he pitched as necessary changes: 

As applications have become more complex, the market has evolved into a variety of monitoring tool to address discrete problems, creating product categories around log management, infrastructure monitoring, APM, and other adjacent categories. Accordingly, vendors in these categories today overwhelm IT and development teams with an array of adjacent, overlapping and disconnected tools.

Each of these tools employs a different user interface and fundamental architecture and each promises visibility, proactive detection, and a mythical single pane of glass through which teams can allegedly see the entirety of their application infrastructure footprint. Meanwhile, the heretofore standard pricing system for APM and infrastructure based on hosts has become increasingly costly and difficult to predict in an age of distributed architectures. It also correlates very poorly to customer value.

He added: 

Current monitoring pricing models create a financial disincentive for teams to instrument all of their applications and infrastructure. Fundamentally undermining the holistic visibility, these teams strive to consider when these solutions were in the first place. Furthermore, customers who adopt these tools encounter surprise overage charges that may cause them de-instrument software or infrastructure to scale back costs further restricting the visibility provided by today's monitoring offerings.

To cope, customers install even more tools, open-source tools, alongside commercial tools, sometimes augmenting and sometimes overlapping functionality. Fragments of critical operational data gets trapped in siloed databases, creating acute bind spots that cripple the team's ability to deliver and operate software effectively. The average organization can end up with dozens of tools to monitor different parts of their stack, forcing engineers, to scramble and switch between these disconnected and often expensive tools to investigate issues while precious time passes in which end users are stuck and disgruntled.

This leads to one conclusion, he said: 

We believe this model no longer works, engineers deserve better. End users deserve better. Customers deserve better. Traditional monitoring techniques are no longer sufficient to provide visibility into modern systems. Complex punitive pricing from monitoring vendors flies in the face of the mission to deliver critical operational visibility. It's time for a new approach. One predicated on a connected, real view of the entirety of an organization's operational data in one central store.

Observing

What’s needed is a move from monitoring to observability, he argued:

Observability describes how well your teams can understand the behavior of a complex digital system. If monitoring tells you when something is wrong, observability lets you ask why. Observability is predicated on upon your software and infrastructure being able to deliver the data, all of the data to provide an answer, supporting questions that lead to other questions that ultimately lead to the ultimate answer.

For a shift from traditional monitoring to observability to happen, our experience shows that customers require the following four things. One, a single source of truth for all your operational data, metrics, events, logs, and traces, not only coming from a vendor's agents but from any source of data; two, comprehensive visibility across the software stack in one unified user experience; three, AI and machine learning driven capabilities to detect anomalies and extract insights in real-time and at petabyte scale; and four, programmability so that customer and third-party developers can build applications and connectors to add functionality, creating an ecosystem of developers and partners. Furthermore, observability requires pricing and technical enablement such that teams can readily instrument all of their applications and infrastructure, then collect that data. 

New Relic’s solution here is New Relic One:

New Relic One is how we deliver observability to the market in a way that is unique from any other commercial vendor or open-source stack. New Relic One now includes three and only three products. The first product is telemetry data platform. Telemetry data platform ingests, visualizes, alerts on all your metrics, events, logs and traces from any source that could include our agents or to conclude third-party sources of instrumentation or telemetry data. It does this at petabyte scale and pennies per gigabyte.

Our second product full stack observability correlates that data from the APM, logs, infrastructure monitoring, client side monitoring and distributed tracing all the way into a single integrated user interface and product, that is easy to analyze and troubleshoot your entire stack in one connected curated experience. Unlike anybody else in our space, we price this product per seat.

Our third product, applied intelligence resolves incidents faster than with machine learning models trained using the entirety of a customer's operational data in the telemetry data platform. As is the case with other AI platforms, the more underlying data there is to work, the better the AI.

Short term vs longer term

The new model has been piloted and is proving appealing to customers, said Cirne: 

Overwhelmingly, customers demonstrated a strong preference for our new pricing model. And this resulted in a higher level of willingness to spend and broader platform adoption with us than what would have been the case, where they to have stuck with our previous model. In one or two cases, customers who had been planning to leave New Relic for a competitive solution decided to do an about face and standardize on the New Relic platform, specifically because of this new model.

But here’s the but that contributed to short term nerves on Wall Street being rattled: 

While we are very excited about what this will do for our customers and our business, note that it will take time for the benefits of this new model to show up in our financial results.

This message was added to by CFO Mark Sachleben, who said the new offering had taken up a lot of time in what is already the firm’s weakest quarter, one that this time around also had the added complication of COVID-19, which itself contributed to “$5 million to $6 million of downgrades”: 

As you can imagine, given the magnitude of our announcement last week, over the past four months, there's been a broad effort throughout the company working toward this launch. We included a lot of tests in discussions with customers and prospects and while this helped define our direction and provided more competence in the long-term success of that strategy, it did take away some of the focus of the team on Q1 results. 

That said, renewal rates across the quarter were up year-on-year and there’s a new head of US sales now in place, he added: 

We want people to adopt [New Relic One], so they expand on it. But that’s, we think, going to cause near term headwind to our results, but set ourselves up for success in the second half and accelerated growth as we get into next year.

Or as Cirne put it:

We are taking a temporary disruption in the short-term to set ourselves up for accelerating growth in the back half and then particularly as we head into fiscal 2022 and 2023 on our path towards getting back towards market rate growth rates. 

But he added: 

There's no reason to expect that all of our customers will suddenly reduce their spend in the short term, some might, but many will increase their spend, but certainly, we expect that this model will incent our customers to standardize on one place as a single source of truth, all their telemetry data that'll drive more users, which will also create a flywheel effect.

Longer term, the product and pricing changes will be a boon, he concluded: 

With New Relic One we have now simplified what has become too complex. We believe our strategy is one that our competitors can neither copy nor beat.

My take

As Phil Wainewright commented last week:

This seems like a bold move by New Relic, effectively telling its customers that the APM product they grew up with and for which the company is best known is no longer adequate for their needs…The calculation New Relic has taken is that it's better to disrupt its own product line and market, than let someone else come in and do it.

That’s a message to customers that Cirne took a lot of time out to explain - or try to explain - to Wall Street’s finest yesterday. The jittery response and the pressure on the share price afterwards suggests that it’s going to take time to get across to many, particularly those whose already instinctive short termism has been accentuated by the stresses of wider COVID-19 uncertainties. 

Validation, as ever, will come in the form of customer response to the new model. That’s something that will play out over the coming quarters. But returning to Phil’s analysis last week: 

Observability is no longer solely the preserve of bleeding edge technologists. It's emerging as something that's regarded as an essential part of the digital engineering toolbox, which suggests the time is right to package it up as a platform for broader adoption. New Relic has spotted the opportunity to establish itself in that role. A bold move indeed, but one that could prove very astute.