Is ServiceNow stock still a buy in 2027?

The two ways to own this name — and what each one is actually betting on

Almost every argument about whether ServiceNow stock is still worth owning collapses into two distinct positions that get confused with each other constantly. Separating them is the first useful thing an operator or investor can do, because they have different entry rules, different failure modes, and different holding periods.

Option A is the compounder hold. You own ServiceNow because it is one of the stickiest enterprise platforms ever built, you expect subscription revenue to keep compounding in the high teens to low twenties, and you assume free cash flow margins in the low thirties persist. Under this thesis the multiple is a tax you pay for durability. You size it as a core position — call it 3–5% of a software sleeve — and you do not trade around earnings. The bet is that a decade of workflow entrenchment inside the Global 2000 means revenue keeps arriving whether or not any given quarter's guide lands cleanly. The failure mode for Option A is slow: net revenue retention drifts down a point or two a year, growth decelerates from ~20% toward the low teens, and the multiple compresses from 15x to 10x over three years while the business itself never actually breaks. You lose money on a company that is still fine.

Option B is the multiple-arbitrage trade. You are not primarily betting on the business — you are betting on the gap between where the forward-sales multiple sits today and where it sits historically. ServiceNow's five-year median forward-sales multiple has run closer to 13x than to 16x. If you buy at 12–13x and the market re-rates back toward 16–18x on an AI attach catalyst, you make 30–40% on multiple expansion alone before a single dollar of incremental revenue shows up. This is a fundamentally different trade: it has an entry price, an exit price, and a time limit. The failure mode is fast — the catalyst does not arrive, the multiple stays compressed for six quarters, and you are stuck holding a position you sized as a trade at a price you would not have paid for a hold.

The adjacent version of this question matters too, because most people asking whether ServiceNow stock is still a buy are also asking whether they should own the workflow-platform *category* at all. That's a third framing: platform exposure versus point-solution exposure. Owning ServiceNow is a concentrated bet that enterprises consolidate service operations onto one system of action rather than stitching together Atlassian for engineering, a separate HR case tool, a separate GRC vendor, and a separate ITOM observability stack. If consolidation is the trend, ServiceNow captures budget that used to be spread across four line items. If the trend reverses toward best-of-breed — as it periodically does when a platform vendor's pricing gets aggressive — the same entrenchment that looks like a moat starts looking like a target for procurement.

There is also a quieter fourth option that gets underweighted: owning the customer instead of the vendor. If you work in RevOps or IT operations, your exposure to ServiceNow's trajectory is already substantial through your own career and your employer's tooling spend. A RevOps leader whose entire operating stack runs on the Now Platform, whose team's skills are ServiceNow-specific, and whose bonus depends on service-delivery metrics measured inside that platform, is already long the name in ways a brokerage statement does not capture. Adding a large equity position on top concentrates a bet that was already concentrated. That's not an argument against owning it — it's an argument for sizing it against total exposure rather than against the rest of the portfolio.

How to decide between the compounder hold and the multiple trade

The decision is mechanical if you write down the inputs first. Three things determine which option fits: the current forward-sales multiple, the trajectory of committed remaining performance obligations (cRPO), and your holding horizon.

Start with cRPO growth in constant currency. This is the single most predictive leading indicator for a subscription business of this shape, because it captures contracted revenue that has not yet been recognized — it tells you what next year looks like before next year happens. Sustained cRPO growth around 20% year over year supports the compounder thesis outright. Growth in the 18–20% band is neutral: the business is fine, but deceleration is visible and the multiple will start pricing it. Anything printing below 18% for two consecutive quarters is the first genuine warning shot, and it converts the compounder hold into a much less attractive proposition regardless of how good the story sounds on the call.

Next, layer in the multiple. Below 13x forward sales with cRPO above 18%, both options are live and they point the same direction — buy. Between 14x and 18x with healthy cRPO, only the compounder hold makes sense; the arbitrage trade has no edge because the gap you would be arbitraging is already closed. Above 18x, neither option is attractive at new money: the compounder is paying too much for durability and the trade has negative expected value.

Third, be honest about horizon. The compounder hold needs three to five years to work, because the entire mechanism is revenue compounding faster than the multiple compresses. If you cannot hold through two bad quarters and a de-rating cycle in the broader software basket, you do not have a compounder hold — you have a trade you have mislabeled, and you will sell it at the exact moment the thesis requires you not to.

The diagram encodes one non-obvious rule worth stating plainly: retention direction overrides multiple. A cheap multiple on a business whose net revenue retention is sliding is not a bargain, it is a value trap with better marketing. Conversely, an expensive multiple on a business whose retention is inflecting upward has historically been forgiven quickly, because expansion revenue arrives at very high incremental margin and compounds into the out-years.

One more decision input that practitioners systematically underweight: what the vendor's own sales motion tells you. If your account executive is suddenly authorized to discount 25% to close a renewal that used to close at list, that is a demand signal reaching you months before it reaches a filing. Conversely, if pricing is firming and the vendor is bundling modules into higher tiers rather than discounting, that is expansion pricing power showing up in real time. Practitioners sitting across the table from enterprise software vendors have an information edge here that no screener replicates — and it applies just as much to adjacent names in the observability, CRM, and security-workflow categories as it does to this one.

The concrete numbers behind each path

Specificity is what separates a framework from a vibe, so here are the numbers that actually drive each scenario.

Revenue and cash generation. ServiceNow's subscription revenue guide has been in the $13.0–13.1B range, with a free-cash-flow margin profile around 32%. That arithmetic implies roughly $4.2B or more in annual free cash flow. Operating margin guidance has run near 30%. Those three figures — subscription revenue, FCF margin, operating margin — are the load-bearing inputs for every valuation scenario below, and they are also the ones most worth verifying against the latest filing before acting, because guidance moves.

The bear scenario, roughly 20% probability. The multiple compresses to 10–12x forward sales. The trigger is a cRPO miss combined with AI attach stalling below 20% of net-new annual contract value. Under that combination the market stops paying for optionality and prices the business as a mature mid-teens grower. Add a scenario where AI inference costs on the assistant products eat 200–400 basis points of subscription gross margin, cracking the 80%-plus gross margin narrative that underpins the whole premium, and the de-rating accelerates. Mid-market compression from Microsoft's bundled Power Platform and Copilot tooling is the mechanism most likely to produce that attach stall — the "we already have this in our E5 license" objection lands hardest in the segment where switching costs are lowest.

The base scenario, roughly 55% probability. The multiple holds 14–16x forward sales. Subscription growth guides in the high teens, cRPO stays around 20%, net revenue retention holds near current levels, and free cash flow compounds. Nothing breaks and nothing inflects. In this world your return approximates revenue growth minus a point or two of multiple drift, plus whatever buyback shrinks the share count. With $2–3B of annual repurchase capacity against the company's cash position and modest debt load, buybacks alone can retire on the order of 2–3% of shares annually — that is 2–3 points of earnings-per-share growth with zero operational improvement required.

The bull scenario, roughly 25% probability. The multiple expands to 18–20x. The mechanism is specific: AI assistant attach crosses 30% of net-new ACV, integrated risk management and CRM-adjacent workflows cross 10% of revenue proving the platform-extension thesis beyond core IT service and operations management, and public sector keeps compounding at 25%-plus on the back of FedRAMP-cleared deployments. Customers adopting the AI modules have been showing meaningfully higher average contract values over multi-year terms — driven both by per-user licensing on AI features and by faster time-to-value pulling forward expansion. If a dedicated premium AI tier lands and captures even 10–15% of the installed base, that is a material increment on top of the base subscription guide.

Switching-cost math, which underwrites all three. A large enterprise typically runs 8–12 distinct workflows on the platform after four to six years — IT service management, IT operations management, HR service delivery, customer service management, security operations, and custom applications built on the low-code app engine. Those custom apps are the sharpest tooth in the moat: they are not portable. A full rip-and-replace runs several times the annual subscription in internal labor and downtime, which is why gross churn in large accounts stays low single digits. That is the structural reason the base case is the fat part of the distribution.

Balance sheet quality. Cash and marketable securities in the $6–7B range against a modest convertible note, negative cash conversion cycle (collecting from customers before paying vendors), and days sales outstanding in the 65–75 day range. DSO drifting above 80 days would be a genuine yellow flag — it usually means either collection problems or quarter-end deal-pulling, and both precede guidance trouble.

Relative positioning. Microsoft trades at a materially lower forward-sales multiple with hyperscaler tailwinds and enormous diversification. Salesforce trades lower still, offering agentic-AI optionality at a value price with slower growth. ServiceNow sits between them on multiple and at the top of that group on growth durability. If you cannot decide, that comparison is itself the answer: owning a smaller ServiceNow position alongside one of the other two partially hedges the open question of who wins enterprise AI workflows, which is genuinely unresolved and will not be resolved by anyone's spreadsheet.

Implementation and sequencing — how to actually put this on

Frameworks fail at the execution layer, so here is the sequencing that turns the decision into positions.

Step one: establish your baseline before you buy anything. Pull the most recent quarterly filing and write down five numbers: subscription revenue guide, cRPO growth in constant currency, net revenue retention, operating margin guide, and free cash flow margin. Then compute the forward-sales multiple yourself rather than accepting a screener's number, because screeners disagree on whether "forward" means next fiscal year or next twelve months, and the difference is often a full turn of multiple. That single discrepancy explains most of the arguments people have about whether this stock is expensive.

Step two: size against total exposure, not just the portfolio. If you work in RevOps, IT operations, or any function whose stack and career trajectory are tied to this platform, count that as existing exposure. Someone whose employer runs its entire service-delivery motion on the Now Platform, and whose own skill set is platform-specific, should probably run a smaller equity position than someone with no professional connection to the category at all.

Step three: scale in on a rule, not on a feeling. A workable approach is thirds — one third of the intended position at the current price if the multiple is at or below your hold threshold, one third on a 15% drawdown, one third on a 25% drawdown or a confirming catalyst such as retention inflecting upward. This is deliberately boring. Its purpose is to remove the decision from the moment when the decision is hardest to make well.

Step four: define the exit before you need it. Write down, in advance, the two conditions that make you sell: cRPO below 18% for two consecutive quarters, or a gross-margin break below the 80% subscription threshold that management cannot explain as a one-time mix effect. Also write down the condition that makes you *add*: net revenue retention printing back above 100% alongside AI attach continuing to climb. Pre-committing to these removes the most common failure in a high-multiple name, which is retroactively rewriting the thesis to justify whatever you already own.

Step five: set a quarterly review cadence, and review the right things. The metrics worth tracking every quarter, in priority order: cRPO growth in constant currency, AI attach as a percentage of net-new ACV, net revenue retention trajectory, named large expansions in the prepared remarks, operating margin against the roughly 30% guide, public sector growth rate, and the gap between operating income and free cash flow — which tells you about working capital quality and stock-based compensation.

Step six: consider the hedge. If the unresolved question is who wins enterprise AI workflows, a paired position across two of the three major contenders converts a binary bet into a spread. You give up upside if your primary pick runs; you avoid being wrong-footed entirely if the category leadership shifts. For a position sized as a core hold rather than a speculation, that trade-off is usually worth making.

Adjacent angles that change the answer more than most people expect

Three neighboring dynamics deserve weight, because each one can move the thesis without ever appearing in a valuation model.

Consolidation versus best-of-breed, and where the budget actually sits. The platform thesis depends on IT budgets continuing to consolidate. What's underappreciated is that consolidation is not just an IT decision anymore — HR service delivery, procurement operations, legal operations, and increasingly customer-facing service workflows are all being pulled onto the same backbone. Every additional department that lands on the platform raises the switching cost superlinearly, because the integrations multiply rather than add. Watch for evidence that non-IT workflows are growing faster than core IT service management. That mix shift, more than any single quarter's revenue, is what determines whether the ten-year story is real.

The mid-market squeeze is a segment story, not a company story. Companies below roughly $500M in revenue have materially lower switching costs and far more sensitivity to bundled alternatives. A vendor whose enterprise business is untouchable can still see growth decelerate if the mid-market expansion motion stalls — and the deceleration shows up in aggregate numbers that get misread as enterprise weakness. When growth slows, ask *which segment* slowed before concluding the moat is cracking. The remedies are completely different: enterprise weakness is a thesis problem, mid-market weakness is a go-to-market problem that a pricing or packaging change can fix within a year.

AI commoditization risk is real but frequently overstated. The bear argument is that generic agentic AI from a hyperscaler or model provider undercuts the platform's AI premium. The counter-argument is specific and worth understanding: the value of workflow AI is not the model, it is the proprietary process context — the configuration, the historical ticket corpus, the approval chains, the CMDB relationships. A general model with no access to that context cannot resolve an incident correctly. That said, the premium is not infinite. If comparable agentic capability shows up at meaningfully lower per-user pricing, the sustainable premium compresses, and the AI-driven ACV uplift baked into the bull case shrinks with it. The honest position is that a moderate premium is defensible and a very large one is not.

There is also a practical read-through for anyone running RevOps rather than a portfolio. The same signals that inform the equity view inform vendor strategy: rising discount authority means better renewal leverage for you; aggressive bundling into higher tiers means budget your renewal for an increase; heavy AI-SKU push means the vendor needs attach numbers and there is room to negotiate a pilot at favorable terms. The investment question and the procurement question run on the same underlying data, and someone with a seat at the renewal table has the earlier read.

Related questions

What single metric matters most for this thesis?

Committed remaining performance obligations growth in constant currency. It captures contracted-but-unrecognized revenue, so it forecasts next year before next year arrives. Sustained ~20% supports the premium; two consecutive quarters below 18% is the first real warning that the growth-durability argument is weakening.

Is a lower multiple automatically a better entry?

No. A compressed multiple alongside falling net revenue retention is a value trap, not a discount. Retention direction overrides price. Only treat a low multiple as an opportunity when the leading indicators — cRPO growth and retention — are stable or improving at the same time.

How does this compare to owning a hyperscaler instead?

A hyperscaler gives diversification, infrastructure exposure, and a lower forward-sales multiple. ServiceNow gives a concentrated workflow-platform bet with higher sensitivity to AI attach outcomes. Owning both partially hedges the unresolved question of who ultimately wins enterprise AI workflows.

Does working in RevOps change the sizing decision?

Yes. If your employer's operating stack and your own skills are platform-specific, you already carry significant exposure through your career. Size the equity position against that total exposure rather than against the rest of the portfolio alone.

What would genuinely break the moat?

Sustained net revenue retention decline across large enterprise accounts — not mid-market softness, which is a go-to-market problem. Enterprise retention erosion means the multi-workflow entrenchment is failing, and that is the one failure mode no pricing change fixes.

FAQ

Is ServiceNow stock still a buy in 2027?

Conditionally. The workable framework is accumulate below roughly 14x forward sales, hold between 14x and 18x, and trim above 18x. The business quality is not the open question — growth durability and free-cash-flow generation are strong. The open question is whether the current multiple already prices several years of that durability, which it often does.

What separates the base case from the bull case?

Attach rate on the AI modules crossing roughly 30% of net-new annual contract value, plus adjacent workflow categories such as integrated risk management and CRM-adjacent products reaching double-digit revenue contribution. Those two together prove the platform-extension thesis and historically justify multiple expansion rather than just revenue growth.

What is the most likely way this thesis fails?

Slowly, not violently. Net revenue retention drifts down a point or two annually, growth decelerates from around 20% toward the low teens, and the multiple compresses over several years while the underlying business remains healthy. Investors lose money on a company that never actually breaks — which is why entry price matters more here than in a lower-multiple name.

How much does the buyback contribute to returns?

Materially but not dramatically. Repurchase capacity in the low single-digit billions annually can retire roughly 2–3% of shares per year, which translates into 2–3 points of earnings-per-share growth without any operational improvement. The risk is a pivot from disciplined buybacks toward large acquisitions, which would strain that arithmetic.

Does Microsoft's bundling actually threaten the business?

In the mid-market, meaningfully. Below roughly $500M in customer revenue, switching costs are low enough that a bundled alternative already included in an existing enterprise agreement is a real objection. In large enterprises running eight to twelve interconnected workflows plus custom low-code applications, the same objection rarely survives contact with the migration cost estimate.

What balance-sheet signal should trigger concern?

Days sales outstanding drifting above roughly 80 days. The company has historically run in the 65–75 day range with a negative cash conversion cycle. A sustained increase usually indicates either collection difficulty or quarter-end deal-pulling, and both tend to precede guidance problems by a quarter or two.

Sources

• https://investors.servicenow.com/ — ServiceNow investor relations: filings, earnings releases, and guidance.
• https://www.sec.gov/edgar/searchedgar/companysearch — SEC EDGAR full-text search for 10-K and 10-Q filings.
• https://www.gartner.com/en/information-technology — Gartner IT research, including service management market analysis.
• https://www.forrester.com/research/ — Forrester research and total economic impact studies on enterprise software.
• https://www.reuters.com/markets/ — Reuters markets coverage for earnings and analyst commentary.
• https://www.cnbc.com/technology/ — CNBC technology and enterprise software market coverage.
• https://www.wsj.com/finance/stocks — Wall Street Journal stock market and equities coverage.
• https://www.morningstar.com/stocks — Morningstar equity research, valuation, and fair value estimates.

Related on PULSE

• Is Salesforce stock still a buy in 2027?
• Is Snowflake stock still a buy in 2027?
• Is Datadog stock still a buy in 2027?
• Why did ServiceNow's stock drop after Now Assist launch?
• Why did Datadog stock drop after Bits AI launch?
• Is a ServiceNow AE role still good for my career in 2027?