ServiceNow is one of the largest enterprise software firms in the world. It automates workflows, tech stacks and projects to augment customer efficiency. For this reason, it calls itself the “leading digital workflow company.” Workflow automation buckets include: service management, operations, asset management, security, customer management, employee management and creator management. These are further grouped into workflow buckets like “customer & employee workflows,” “creative workflows” and “technology workflows.” Two products to know within the tech workflow category include Information Technology Operations Management (ITOM) and Information Technology Service Management (ITSM). The names of these products tell you exactly which types of workflows they’re meant to automate.

All products and services are neatly tied into its “Now Platform.” The firm describes this overarching ecosystem layer as a way to “optimize processes, connect silos and accelerate innovation on a single unifying platform.” That’s a fancy way of saying that it makes every piece of work more seamless and expedient.

To bolster automation capabilities, ServiceNow has been hard at work on GenAI innovation. Its Now Platform “Vancouver” release got the ball rolling by consolidating all GenAI model and app projects into an intuitive set of products. It recently built on that debut with a Now Platform “Washington D.C.” release. This is essentially a large batch of GenAI-inspired upgrades to the Now platform. It adds to the progress of the Vancouver release. Washington D.C. more seamlessly ties together NOW’s product categories to drive better interdepartmental work and communication. It makes using all of its tools and capabilities across teams more intuitive and obvious. 

ServiceNow also offers the “workflow studio” as a unified workspace to manage productivity across teams. It seamlessly refreshes databases without complex coding; it goes deeper in terms of intelligently automating customer service and order management workflows. This is a constant pursuit to make its clients happier and their lives easier.

These platforms form the foundation for its own GenAI apps. A key example of these apps is “Now Assist AI,” which is close to $100 million in average contract value (ACV) on its own already. This is ServiceNow’s GenAI assistant/companion being infused across most of its products. Several iterations of this product have been launched for various ServiceNow workflow categories. More GenAI product examples include:

“Plus SKUs” are how ServiceNow bundles all of its GenAI work into subscription packages. It up-charges clients for access to these SKUs, as its approach to GenAI monetization has been more aggressive than most. These Plus SKUs do things like automate customer service, expedite issue resolution, guide workflows and provide more conversational fetching/querying of a firm’s data.

Now trades for 61x forward EPS. EPS is expected to compound at a 25% clip over the next two years.

Every quarter, ServiceNow reminds us how obvious of a workflow platform play it is. As I’ve spoken about for years now, platforms mean vendor consolidation, broader interoperability, better outcomes and lower costs. They allow companies to do “more with less” and unlock durable demand amid the more challenging macro backdrops. This quarter, we simply got more data confirming this enterprise as a true software platform.

It signed its 2nd largest new logo ever, included ITSM and ITOM in 15 of its 20 largest deals and cross-sold its security and risk products in 14 of its 20 largest deals. Customer, employee and creator workflows were all in 13+ of its largest 20 deals (customer workflow volume rose 50% Y/Y). Finally, 18 of its top 20 deals included 7+ modules. It continues to grow its list of $1 million+ customers and even added 6 contracts worth $10 million+ during the quarter.

And when ServiceNow doesn’t have the module a customer wants? It partners openly with world-class firms in complementary areas of software to ensure customers have open access to whatever they need… with one interface and one user experience… all on ServiceNow.

The Now Assist AI product continues to be its “fastest growing product ever.” It added 44 customers with contracts worth more than $1 million, with 6 of those being brand new logos.

To build on this momentum, as well as Vancouver and Washington D.C. updates, ServiceNow released the Xanadu update. The release comes with 100s of new tools and a completion of the previously announced Microsoft Copilot integration to “meet employees wherever they work.” The update works to infuse GenAI even more deeply into NOW’s product suite in two new areas. First, this adds GenAI capabilities for security and threat management workflows. The models are trained on a client’s own data and enriched with NOW’s overarching ecosystem to “accelerate incident response and improve threat containment.” It also offers transferable, automated incident summaries to help with prioritizing threats.

Secondly, the release adds new GenAI capabilities within financial and supply chain workflows – starting in sourcing and procurement. Procurement, per Gartner, is among the earliest areas of concrete value creation and material enterprise demand; ServiceNow now has the tools to help. The firm sees “immense opportunity for intelligent automation of complicated, fragmented processes.” This will “handle procurement requests, improve compliance and drive tangible business value.” That’s absolutely vital if NOW will continue to monetize its GenAI work in a hesitant macro backdrop. It has monetized better than pretty much anyone else thus far because it has found ways to actually create return on investment for customers.

Aside from these two expansion markets, the Now Assist Skill Kit is a tool to help developers deploy new GenAI prompts and workflows. ServiceNow has templates for pretty much all common needs, but it cannot possibly build models for every niche workflow. That’s where this comes into play. More new tools include:

ServiceNow debuted the Workflow Data Fabric. This is a “new data layer to unify business and technology data across enterprises; it powers AI agents and workflows with real-time access to information and does things like cut transaction times by 53%. Simply put, it’s a “unified data management platform” to ensure all processes are guided by a firm’s own data and context. This works perfectly with the recently upgraded RaptorDB database used to bolster processing speeds. Per leadership, Raptor processes transactions 12x faster and data analytics 27x faster than legacy solutions.

Over the last few quarters, data has re-emerged as a vital ingredient for innovation. It never went away… focus just temporarily shifted. In a highly competitive, somewhat commoditized world of language model (LM) building, access to more relevant data is what differentiates one elite LM vs. another. I will keep saying it. Furthermore, “bringing data to the work,” as we hear from so many CEOs matters. Native integrations between databases and other software applications have a way of driving broader cohesion, more valuable querying, better decisions and lower costs. It’s conceptually similar to platform plays driving better inter-product communication.

GenAI apps and models rely on voracious data processing; when that data is closer to the final destination and more interoperable, transfer costs are lower. Zero copy partnerships with both Snowflake and Databricks for this product help a lot too. That’s why Salesforce worked so hard on their Data Cloud; it’s why Zscaler and CrowdStrike have focused so much on their own data storage scalability.

“So this is going to be an upsell opportunity for every existing installation across the world for ServiceNow.”

CEO Bill McDermott

Agentic AI allows companies to move from prompt-based chatbots to a “goal oriented” model with levels of autonomy needed to pursue that goal the best way possible. This allows the world to “evolve from the more familiar prompt-based activity to deeper contextual comprehension.” Agentic AI is designed to arrive at a concrete objective and is less commonly utilized for writing code or generating content. You can see how a model that can push towards a goal would be compelling for customers. All that client needs to say is “hey I’d like this inefficient workflow to be better” and Agentic AI can generate ways to bring that to life. NOW announced a new Agentic AI partnership with Nvidia to co-create AI Agents with Nvidia’s Inference Microservices (NIMs).

“Xanadu showcases our leadership in Agentic AI. Until now, GenAI required human prompts to initiate action. We’re deploying autonomous Agentic AI products that work with people, not just for them… It’s like hiring more people to support your workers by doing the jobs they’ve never wanted to do.”

CEO Bill McDermott

In a world desperate for evidence that GenAI is worth the investment, NOW continues to provide highly compelling case studies showing its tangible value. That’s why it’s having an easier time monetizing than basically every other software firm. For a “multinational conglomerate,” its GenAI products have cut its case deflection rates by 75% and for another large tech firm, it boosted employee satisfaction rates by 30% by automating tedious workflows.

“When public cloud solutions came online during the last secular shift, they created sprawling governance challenges. CIOs were asked to get things under control and ServiceNow played that role. Now the C suite is looking to ServiceNow to prevent another mess with AI. Leaders see the risk that every vendor's bots and agents will scatter like hornets fleeing the nest and they trust ServiceNow as the governance control tower, which is a privileged position for our platform.”

CEO Bill McDermott

On its Carahsoft Relationship Impacting Results as that Firm Deals with Public Sector Contract Price Fixing Allegations:

“We didn't see any impact in Q3, and we're currently not involved in any issues regarding Carahsoft with the U.S. federal business. So you should just take that one off of any concern list that you might have and as I look at the business in Fed, it could not possibly be stronger.”

Another rock solid quarter from this special software compounder. Aside from maybe Microsoft and Palantir, there is no software company realizing more success with monetizing GenAI work; there’s also no software company with more numerous incremental opportunities to keep doing just that. Continued 20%+ growth with operating leverage and a beautiful balance sheet… continued elite execution… more of the same. Enough said.

The stock could easily chop around or cool off at this multiple and after a fantastic run, but the company keeps chugging ahead.

Tesla needs no introduction.

Detailed coverage of the We Robot event can be found in section 4 of this article.

More deliveries, energy revenue, software revenue recognition and regulatory credits all helped revenue during the quarter. Lower average selling price (ASP) across its models due to mix, pricing and financing options more than offset those tailwinds to lead to the small miss.

While Tesla’s Auto GPM has fallen over the last couple years, it remains the only profitable, scaled EV program in the world. Furthermore, there have been bountiful news articles on price cuts, cheaper financing options and more incentives to spur demand. This has been the key source of negative sentiment. Beating margins by this wide of a gap despite all of that is a notable surprise.

Tesla continues to work hard on lowering cost per vehicle. That set a new low for the company this quarter at $35,100, as it flexes the advantages of its vertically integrated supply chain. This gives it a birds-eye view of what can be optimized and direct control over that optimization process too. The large margin lead Tesla enjoys over everyone else is a byproduct. At the same time, it’s not only using this strength to preserve margin. It will continue to balance pocketing part of this while also giving some of the savings to consumers while cost of capital remains very elevated.

“Note that we are focused on the cost per vehicle and there are numerous work streams within the company to squeeze our costs without compromising our customer experience.”

Vaibhav Taneja

Tesla reiterated all previously-issued guidance, including plans for affordable model production during the first half of 2025. It also raised its annual CapEx guide from $10 billion to $11 billion. It continues to invest in GPU capacity with needed caution and is basing new orders off of near-term needs.

On the call, Elon also told investors to expect 20%-30% auto volume growth for next year. This is well above the high-teens estimates I see from the Street.

As of right now, Tesla trades for 78x forward EPS, with -26% EPS growth expected this year and 38% EPS growth expected next year. These large profit beats should result in upward profit revisions during the coming days.

Considering the planned $25,000 model wasn’t mentioned at We Robot, inclusion of this in the report is encouraging and should have bulls relieved. This car is an important part of the 2025-2026 growth engine. It must carry part of the financial burden for the overall company until Tesla starts producing its next-generation models at scale (hopefully in 2026).

While 2026 will potentially bring brand new cars, the 2025 bridge model is an iteration of existing models. It borrows from existing manufacturing lines, with some new manufacturing techniques for the 2026 vehicles used. Relying on some of the 2026 methods will result in lower near-team cost per vehicle savings for Tesla than previously thought. At the same time, it will “enable them to prudently grow vehicle volumes in a more CapEx efficient manner.”

Per Musk, new manufacturing techniques for the 2026 models “aren’t in the same league” as other processes in terms of production capacity and efficiency. During Tesla’s We Robot event, CEO Elon Musk told us to expect Cybercab production to begin by 2027. Two weeks later, he told investors to expect volume production to begin in 2026. I’m not sure where the incremental optimism came from since the recent event, but this is certainly positive.

Elsewhere, the planned Tesla semi-truck is “on track to ramp production throughout 2026. The team is confident that this product “won’t be demand limited.” In early testing on pilot models with Pepsi, drivers are concluding that they “never want to go back.” Between using excess brake heat to recharge batteries, anti-jack knifing technology and more, the team sees this product as miles ahead of any other. This is a $220 billion industry in the USA alone.

The newest version of Tesla’s Full Self Driving (FSD) software (12.5 version 13) offers a 5x-6x improvement in miles per intervention and 5x more parameters than previously used models. Two quarters ago, Tesla really got into the weeds on the neural networks and highway/city software integrations that make this newer software better. That commentary can be found in section e of this article.

Tesla continues to expect FSD to offer safer rides than humans by Q2 2025 through continued tech improvements. It also expects to achieve full, unsupervised autonomy next year in California and Texas (maybe some other states too) with paid ride sharing rolled out in those states next year too. That’s sooner than expected. These rides will be offered by existing models (which are capable of unsupervised FSD) to start. Elon acknowledged that Tesla’s hardware 3 may never get to full autonomy like he thinks hardware 4 will. If that happens, hardware 3 customers will receive a free retrofit.

In other FSD news, Tesla debuted “Smart Summon” to call your car to pick you up from parking lots. The team also told us that “FSD take rate has improved substantially, especially after the October 10th event.”

Tesla’s 4680 cell battery is giving Elon confidence in this product being the most competitive in the USA on a kilowatt per hour basis. Cost, density and range characteristics are excellent vs. competition to allow for longer miles per charge. It’s used across Megapack (battery storage), Powerwall (rechargeable home battery) and its vehicles like Cybertruck. Tesla will continue to rely heavily on 3rd party cells to power its products too, but this will become an increasingly important source of its capacity needs.

We also got more updates on Megapack production at its Lathrop and Shanghai factories. Lathrop is now up to 40 gigawatt hours (gwhrs) per year in terms of energy storage capacity and Shanghai will ramp to 20 gwhrs per year throughout 2025. Tesla is well on its way to 100 gwhrs in Megapack storage capacity, which features lower energy conversion waste and higher reliability than alternatives.

The energy segment set a new record for gross margin with a 30.5% result vs. 24.5% last quarter. This was despite lower megapack shipment volumes (quarterly lumpy). Powerwall deployments set a new record and the Powerwall 3 ramp (newest version) continues to go very well.

Last quarter, market share declines in China and Europe were a key concern. Conversely, North American market share rose Q/Q. This quarter, North American share dynamics were stable, while China returned to growth and Europe flattened out. Stronger showing.

Macro headwinds have not gotten any better. Rate cuts should begin to help the sector, but those were very recent, and it takes time for accommodative policy to take effect in areas like this. At the same time, Tesla is using its strong margin profile and balance sheet to stay on the offensive with a goal to extend its competitive lead as times get easier once more.

Tesla added new Model 3 & Y trims during the quarter, as well as more paint options for its models. In the USA, the ramp up of the previous model 3 refresh is going well and Cybertruck production tripled Q/Q. It is on track to be profitable by year’s end. A clear example of the power of the Tesla ecosystem: This will reach profitability far more quickly and at smaller scale than others.

“We released FSD for Cybertruck and other features like actually smart summon like Pilar talked about in North America, which contributed $326,000,000 of revenues in the quarter.”

CFO Vaibhav Taneja

In China, cost of goods sold per car set new lows and leadership told us the market is outperforming both Europe and the USA. Impressive given the fierce competition there. In Europe, Model Y became the most sold car in several countries and was the top seller in all of Europe during September.

This was a strong showing for Tesla given all of the negative sentiment. Macro remains difficult for this sector and Tesla continues to outperform its peers by a country mile. Whether that makes this a good investment at its valuation and reliance on future products is up to you. It’s firmly in my “too hard” pile.

From a company execution standpoint for this specific quarter, this was great in my view. The small revenue misses are minor negatives compared to the large profit beat positives. The GPM beats were specifically notable to me, especially considering many optimistic sell-siders were warning of probable disappointment. Timelines for new launches (always uncertain with Elon) were also moved up to add to the positive vibes exiting this call. Congrats shareholders.