Summary

  • SailPoint reported more than US$70 million of AI-driven ARR in fiscal Q2 2027. AI-driven solutions supplied more than 30% of quarterly net new ARR, while SaaS supplied 97%.
  • The company’s ARR definitions allow a contract to remain counted after expiry while SailPoint is actively negotiating a renewal or new agreement, until the customer says it will not renew.
  • Contracts in that state were less than 1% of total ARR and less than 1% of SaaS ARR. That bounds the company-wide tail below US$12.307 million, but SailPoint does not disclose its amount inside AI-driven ARR.
  • No evidence says that a particular AI contract is expired, impaired or likely to churn. The issue is measurement: live service, contractual expiry, negotiation and eventual renewal are different states.
  • A renewal-status column would connect AI adoption to active entitlements, signed duration, expansion, recognised revenue and cash without exposing customer identities.

A new metric arrives with an old timing question

SailPoint’s fiscal second-quarter results give investors a more useful AI signal than a product roll-call. AI-driven annual recurring revenue exceeded US$70 million. AI-driven solutions represented more than 30% of net new ARR in the quarter. Existing customers that adopted one of those solutions increased annual spend by more than 60%, and more than two-thirds of completed migrations included an AI-driven product.

These are four different observations. The first is a stock of annualised contract value at a measurement date. The second is a share of the quarterly change in a larger operating metric. The third describes spend by a selected group of adopters. The fourth describes the product content of migrations that finished during the quarter. None is a recognised-revenue line, a cash receipt or a company-wide conversion rate.

That distinction matters because the company also provides an unusually explicit definition. AI-driven ARR covers Agentic Suites, SailPoint Agentic Fabric and agentic add-on modules. Contract value is divided by contract days and multiplied by 365. If a contract has expired but SailPoint is actively negotiating a renewal or new agreement, its annualised value can remain in the metric until the customer says it will not renew.

The rule does not make the metric false. Enterprise renewals do not always sign on the exact day a prior document expires. Procurement, security review, budget approval and legal redlines can leave a commercially continuing relationship in an administrative gap. Removing every contract at midnight on the expiry date could create volatility that says more about paperwork than demand. Keeping it indefinitely would create the opposite distortion. SailPoint’s definition chooses a stopping event: an explicit decision not to renew.

The investable question is what happens between those two moments.

The company-wide ceiling is reassuring but incomplete

The Form 10-Q says contracts kept in total ARR during active renewal or replacement negotiations were less than 1% at the dates shown. Total ARR was US$1.2307 billion at 31 July 2026. One percent of that balance is US$12.307 million, so the disclosed wording places the tail below that amount. It is an outer boundary, not a reported balance.

SaaS ARR was US$846.9 million, and SailPoint separately says the same renewal-negotiation population was less than 1% of that metric. Again, the filing does not publish an exact amount. The disclosure is nevertheless meaningful: it prevents a reasonable reader from treating the whole 25% rise in total ARR, or the whole 36% rise in SaaS ARR, as a product of expired contracts waiting for paperwork.

The narrower AI metric has a different denominator. At the disclosed floor, US$70 million equals about 5.69% of total ARR. Because SailPoint says “exceeded” and the inputs are rounded, 5.69% is only a minimum share. The company does not say what portion, if any, of the less-than-1% company-wide renewal tail belongs to AI-driven solutions.

It would be wrong to assume the tail is distributed proportionally. It would be equally wrong to assign the entire ceiling to AI. The total-ARR statement limits the maximum company-wide issue; it does not locate it. A small absolute tail could be immaterial to total ARR while still being more noticeable inside a young product cohort. Only a segmented status disclosure can resolve that question.

“Expired” is not the same as “gone”

Contract expiry is a legal date. Service state is an operational fact. Renewal intent is a commercial judgement. Revenue recognition and cash collection each follow their own rules. Those clocks can line up, but a mature measurement system should not presume that they do.

Consider four possible states. A customer may have signed before expiry and be fully active under a new term. It may continue receiving service while final signatures are pending. It may be negotiating a materially different product bundle. Or it may be disengaging without yet having delivered the formal notice that removes it from SailPoint’s metric. The public definition can encompass at least the middle states, but it does not publish their ageing or outcome.

This is not an allegation that SailPoint is extending access without authority or recognising revenue without a contract. ARR is an operating metric, not a GAAP account. The filing explicitly says it is not a forecast of future revenue and should not be combined with or substituted for revenue. The measurement question is narrower: how much annualised value has a current signed term, how much is inside a short administrative bridge, and how much remains exposed to an unresolved commercial decision?

An ageing band would help. Zero to 30 days after expiry can reflect ordinary procurement friction. A longer interval may still be benign for a complex enterprise migration, but it deserves a separate explanation. Outcome data would be better still: value renewed unchanged, expanded, contracted, replaced by a different agreement, or lost. No customer identity is required.

Net new ARR supplies momentum, not the missing denominator

SailPoint says AI-driven solutions accounted for more than 30% of quarterly net new ARR. It also says net new SaaS ARR grew 34% year on year and supplied 97% of net new ARR. The figures show where the incremental contract engine is pointing: overwhelmingly toward SaaS, with AI-driven products already contributing a significant share of the movement.

They do not disclose the absolute quarterly net-new-ARR denominator in the release. “More than 30%” therefore cannot be converted into a defensible dollar addition from the published sentence alone. Nor can it tell readers how much came from a new logo, an existing-customer expansion, a migration, a renewal uplift or currency.

The definition of net new ARR is also a change in a stock. It can be influenced by additions, expansions, contractions and removals. The final number is not a bookings ledger. A product can contribute positively even if a different cohort shrinks; a migration can raise ARR because added functionality carries a higher price even when no new customer has arrived.

This is why the renewal-status question is not answered by a strong net-new share. Flow and status are different axes. The flow shows how the total changed. Status shows how much of the ending balance is supported by a live signed term, a negotiated bridge or an unresolved decision.

Selected adoption cohorts need their own base

The more-than-60% annual-spend increase among existing customers adopting an AI-driven solution is commercially encouraging. It suggests that the product can enlarge an established relationship rather than depend solely on new-logo acquisition. But the statement does not disclose the number of customers in the cohort, their starting spend, the distribution of increases, product mix, contract duration or whether migration itself changed the price.

It is not a 60% expansion rate for all SailPoint customers. It is not a price increase. It is not proof that the AI component alone produced the full increase. An existing customer could add users, move from a customer-hosted product to SaaS, upgrade a suite and purchase an agentic module in the same agreement. The commercial result may be excellent while attribution remains mixed.

The migration statement carries a similar boundary. More than two-thirds of migrations completed in the quarter included an AI-driven solution. The denominator is completed migrations, not the installed base, pipeline or all renewals. It does not reveal how many projects finished, how long they took, which customers deferred, or whether the AI component was an add-on, a suite feature or the reason for moving.

These qualifications do not erase adoption. They describe the receipt needed to understand it: cohort size, starting ARR, product components, signed term, live entitlement, deployment state and renewal outcome.

SaaS changes the revenue clock as well as the product

SailPoint’s transition toward SaaS is material. SaaS ARR represented 69% of total ARR at July’s end, up from 63% a year earlier. Subscription revenue reached US$295.205 million in the quarter, up 19%, while total revenue was US$308.813 million, up 17%.

The 10-Q explains why ARR can run ahead of revenue during this transition. SaaS revenue is recognised ratably. A portion of term-subscription licence value can be recognised upfront when control passes. Moving a customer from a hosted licence to SaaS may increase annualised contract value because the cloud suite includes more functionality, while shifting the accounting timetable toward recognition over the service period.

SaaS also changes the cost boundary. SailPoint carries the hosting obligation. The company says a higher SaaS share can have a near-term negative effect on revenue growth and gross margin because of revenue-recognition differences and hosting costs. That means an AI-driven SaaS expansion has at least three relevant clocks: the annualised contract measure, recognised subscription revenue and the cost of delivering the service.

Gross margin alone cannot isolate the new products. Quarterly subscription cost of revenue was US$84.103 million and subscription revenue US$295.205 million. Acquired-technology amortisation was a large part of subscription cost, and hosting, software, staff, royalties and capitalised-development amortisation also sit in the line. SailPoint publishes no AI-only revenue or gross margin.

The GAAP operating loss of US$58.958 million and adjusted operating income of US$63 million likewise should not be assigned to AI. The reconciliation removes equity compensation, acquired-intangible amortisation and other specified items. These company-wide measures provide capacity and discipline context; they are not a product contribution statement.

Retention is broader and slower than a renewal-status receipt

Dollar-based net retention was 113%, down from 114%. SailPoint calculates it from the same subscription-customer cohort and reports a weighted average of the four most recent quarterly observations. The measure captures expansion, contraction and churn across a year-old base.

That makes it valuable and deliberately smooth. It also makes it unable to answer the immediate question about AI contracts at one quarter-end. The cohort includes products beyond AI-driven solutions. A trailing average can absorb a recent shift. Expansion elsewhere can offset a lost contract. Retention shows the health of a broad installed base, not the legal and operational status of each dollar in a new product metric.

The two measures should work together. Retention can test whether the company sustains expansion over time. A renewal-status column can show whether the ending AI balance is composed of signed current terms or temporary negotiation bridges. One provides outcome trend; the other provides measurement-state integrity.

The small receipt that would make the metric durable

SailPoint does not need to publish customer names or contracts. A compact quarterly table would be enough. For AI-driven ARR, it could show value under current signed terms; value beyond expiry during active negotiation; ageing bands for that bridge; value renewed, expanded, contracted or lost during the quarter; and the share already live in production rather than purchased but not deployed.

The table should also keep product motion separate: new-logo ARR, cross-sell to existing customers, migration uplift and renewal uplift. A bridge to recognised subscription revenue would not need to promise dollar-for-dollar conversion. It would only explain timing, term and major ASC 606 differences. An accompanying delivery measure—active agent identities, governed connections or another stable entitlement unit—could show whether contract value has become operating use.

That evidence would allow several outcomes to be distinguished. A small, quickly resolved renewal tail would support the view that the existing definition merely smooths administrative timing. A growing or ageing tail would make net new ARR less informative until outcomes arrive. Strong signed expansion with rising live use would validate the AI signal. Signed value without deployment might still be strategically important, but it would carry a different implementation and renewal risk.

SailPoint’s disclosure already contains the essential honesty: the rule exists, the company-wide tail is below 1%, and ARR is not revenue. The missing step is allocation. AI-driven ARR has crossed the threshold where its composition matters. A renewal-status column would turn a promising indicator into a better decision interface.

Sources