The average company’s SaaS portfolio did not grow last year. According to Zylo’s 2026 SaaS Management Index, the average organization runs 305 applications (median 240), and the total app count dipped 0.07%. Read that headline alone and SaaS sprawl looks solved. It isn’t. The same report shows average annual SaaS spend rising 8% to $55.7M and AI-native app spend more than doubling, up 108%. Sprawl did not shrink; it moved into places a simple app count no longer sees.
That shift changes how a finance or IT leader should define the problem, find it, and govern it. The rest of this guide covers all three.
What SaaS sprawl means now
SaaS sprawl is what happens when software adoption outpaces an organization’s ability to track, own, and price what it uses. The old definition focused on volume: too many apps, too many duplicates, too many forgotten subscriptions. Those still matter, but in 2026 sprawl shows up across five dimensions, and only the first is about counting.
- Redundant tools. Three project management platforms, two whiteboarding tools, a handful of file-sharing services, each adopted by a different team for a defensible reason at the time.
- Unused licenses. Zylo’s data puts average license utilization at 54%, up from 47% the year before (Zylo SaaS statistics). That is real progress, and it still means close to half of paid seats sit idle.
- Shadow purchasing. The expense channel is where the app count is still climbing. Zylo found the average number of expensed apps rose from 125 to 138, while non-expensed apps fell from 229 to 206 (Zylo SaaS statistics). Expensed tools are a small slice of spend (3.7%), but they are where the next contract you don’t know about starts.
- Shadow AI and embedded AI features. 77% of IT leaders discovered AI-powered features or apps operating without IT’s awareness, and 60% lack visibility into all the generative AI tools in use (Zylo SaaS statistics). This is the fastest-growing form of sprawl and the one least visible to traditional discovery. We covered why it belongs on the finance agenda in Finance Can’t See a Third of Your AI Spend.
- Metered sprawl. Hybrid seat-plus-usage pricing means a single, fully sanctioned app can sprawl on its own. 78% of IT leaders reported unexpected charges tied to AI features or consumption pricing, and 61% cut projects because of unplanned SaaS cost increases (Zylo 2026 Index).
The last two dimensions are the reason the bill rose while the app count stayed flat. An organization can consolidate from 300 apps to 250 and still lose control of spend, because one consumption meter on a tool everybody already uses can outgrow ten forgotten subscriptions. If that sounds abstract, Microsoft’s decision to switch on Copilot’s usage meter by default for new Business seats is a live example (our breakdown here).
Why sprawl persists even in disciplined organizations
Most finance and IT leaders know sprawl exists. It persists because the forces creating it are structural, and the controls built to stop it were designed for a different kind of software.
Purchasing authority is still decentralized
A $15 per user per month tool for a ten-person team costs $1,800 a year, which sits below most expense approval thresholds. Repeat that across 50 cost centers and you have $90,000 of spend that never touched a purchase order. Business units drive a large share of software decisions, and the Zylo statistics roundup cites Capgemini research finding that 48% of SaaS expenditure is driven by business units outside IT’s control.
Freemium converts quietly
Tools enter free, gain users, then convert to paid tiers on a stored card. By the time finance sees the charge, the tool is embedded in workflows and politically hard to remove. AI tools follow the same path, only faster, because the free tier is often good enough to become a daily habit within a week.
AI features arrive inside apps you already approved
This is the new root cause. A vendor ships an AI assistant into a platform you licensed two years ago, prices it as an add-on or a consumption meter, and enables it for some or all users. Nobody submitted a purchase request, so no intake process fired. The outcome-based pricing shift (HubSpot billing per AI resolution, for example) makes this even harder to forecast.
Renewals run on their own calendar
The average organization manages 211 SaaS renewals a year, and 79% of IT leaders saw price increases at renewal in the past 12 months (Zylo SaaS statistics). That is close to one renewal every business day. Without a central calendar, auto-renewal executes by default and locks in another term before anyone checks usage.
Discovery tools were built for installed software
Traditional IT asset management tracked installation footprints. SaaS leaves no such trace, and embedded AI features leave even less. Flexera’s 2026 State of ITAM research found that only 66% of organizations have full visibility into their SaaS environment (Flexera).
Running IT operations at a large telecom, the renewal problem was the one that cost us most, and it was rarely the big contracts. The enterprise agreements had owners, calendars, and a negotiation plan. The damage came from mid-sized tools with 60 or 90 day cancellation notice clauses that someone had signed on behalf of a team that had since reorganized. By the time anyone noticed, the notice window had closed and the decision had been made for us. That experience is why the governance section below treats renewal ownership as a named person, not a field in a spreadsheet.
A five-phase sprawl assessment
Before adding controls, build an honest baseline. This assessment runs about eight weeks for a mid-sized organization, and the phases overlap.
- Financial discovery (weeks 1 to 2). Pull 12 months of accounts payable data, corporate card transactions, and expense reports. Search for recurring charges from known SaaS and AI vendors, and flag any subscription-pattern charge to an unknown vendor. Given Zylo’s expense-channel data, give the card and expense feeds as much attention as AP.
- Technical discovery (weeks 2 to 3). Export SSO and identity provider authentication logs. Review OAuth grants in the Microsoft 365 or Google Workspace admin console, since that is where AI assistants and browser-based tools connect to corporate data. Add a browser extension or CASB if your privacy policy allows it. Apps with no SSO integration will only show up here or in the finance feeds.
- Usage and metering analysis (weeks 3 to 5). For each discovered app, pull login frequency from the IdP and admin dashboards. This is the step the old version of this assessment missed: for every app with a consumption or AI component, also pull the metered usage and the current billing rate. Seat utilization tells you about waste; metered usage tells you about exposure.
- Contract inventory (weeks 4 to 6). Locate every agreement in legal repositories, procurement systems, and email archives. Record term length, renewal date, cancellation notice period, price escalation clauses, and any usage-based or overage terms. Expect a meaningful share of subscriptions to have no locatable contract at all, running on click-through terms.
- Rationalization mapping (weeks 6 to 8). Group apps by function, identify overlap, and document integration dependencies that complicate consolidation. The output is a ranked list of consolidation candidates with estimated savings and migration effort. Our guide to SaaS rationalization covers how to sequence the cuts.
Treat the first pass as a baseline, not a cleanup. The number that matters most is not the app count you end with, but the share of spend that now has a named owner and a known renewal date.
SaaS management platforms: what changed in 2026
Specialized SaaS management platforms (SMPs) automate discovery, usage tracking, and renewal workflows. The 2024 view this article originally carried is out of date: the market shifted in 2026, through one shutdown, one acquisition, and a broader push into AI governance.
- Productiv shut down on August 6, 2026, with roughly four days between the announcement and the sunset, according to a customer migration guide published by competitor Zluri. Zluri’s guide notes that usage and engagement history was the one asset customers could not rebuild from other systems.
- CoreStack acquired BetterCloud on March 31, 2026, combining cloud governance with SaaS management in what the companies describe as a single governance layer across cloud, SaaS, and AI, serving more than 2,000 customers and partners (announcement via Newswire).
- The 2026 Gartner Magic Quadrant for SaaS Management Platforms, dated June 18, 2026, evaluated 16 vendors (Flexera). Flexera, Zylo, and Torii each announced Leader placements. The report itself is paywalled, so read vendor claims about it as vendor claims.
| Platform | Discovery approach | Best fit | Watch for |
|---|---|---|---|
| Zylo | Financial, SSO, and API integrations | Large enterprises with complex portfolios and heavy renewal volume | Implementation effort; plan for a multi-week rollout |
| Flexera | SaaS management inside a broader ITAM and FinOps suite | Organizations that want SaaS, on-premise licensing, and cloud in one place | Breadth can mean longer configuration time |
| Torii | Browser extension, SSO, and financial data | Mid-market teams wanting workflow automation | Extension rollout can meet employee privacy pushback |
| Zluri | SSO, API, and browser extension, now positioned around identity governance | Security-led buyers focused on access reviews | Spend features are secondary to the identity pitch |
| BetterCloud (CoreStack) | SaaS-to-SaaS workflow automation | IT operations teams prioritizing automation | Roadmap and packaging may change under new ownership |
| Vendr | Financial discovery plus managed buying | Organizations wanting negotiation support alongside data | More services-led than self-service |
Two lessons come out of the year’s churn. First, export your usage history on a schedule, regardless of vendor. Productiv customers learned that the irreplaceable data in an SMP is the usage trend, not the inventory. Second, ask every vendor how it discovers embedded AI features and consumption meters, because that is now where sprawl grows. A platform that counts apps well but cannot see a metered add-on inside an approved app will report a clean portfolio while the bill climbs.
No platform achieves complete discovery. Finance integrations catch paid tools but miss free tiers. SSO misses tools that use direct login. Browser extensions see the most but face deployment friction. Layer methods regardless of which platform you pick. For smaller portfolios, a disciplined manual process (expense reporting, IdP exports, quarterly owner surveys) can deliver most of the value of a dedicated platform at a fraction of the cost.
Governance that stops sprawl from returning
Discovery fixes the past. Governance prevents recurrence without blocking the speed that made SaaS attractive in the first place. A full template lives in our guide to building a SaaS management policy people actually follow; these are the controls that matter most.
Tiered approval by risk and spend
- Tier 1 (under $5K a year, no sensitive data): department head approval, registered in the central inventory within 30 days.
- Tier 2 ($5K to $50K a year, or touches employee or customer data): IT security review plus finance approval, standard contract terms required.
- Tier 3 (over $50K a year, regulated data, or any AI tool that processes company data): full procurement, legal review, and vendor security assessment.
The AI clause in Tier 3 is new. It reflects that a cheap AI tool can carry the data exposure of an expensive one.
A metered-spend rule
For any app with a consumption or AI component, require a named owner, a monthly cap or alert threshold, and a decision on whether new features are enabled by default. Treat the metered portion as its own budget line, separate from the seat subscription, so a usage spike is visible as a usage spike instead of being buried in “software.”
A central renewal calendar with named owners
Keep one source of truth for renewal dates with alerts at least 90 days out, earlier than the cancellation notice period on each contract. Assign a renewal owner who must validate usage before each renewal, and make auto-renewal require an affirmative decision. Bring usage data, not seat counts, to the negotiation; our guide on negotiating SaaS contracts at renewal covers the playbook.
Quarterly usage reviews
For Tier 2 and Tier 3 apps, review utilization every quarter. Licenses below 60% utilization for two consecutive quarters trigger a rightsizing review. This cadence catches waste before the next renewal locks it in.
An approved catalog
Publish a short list of pre-vetted tools per category, including approved AI assistants. When a team needs a capability, it picks from options already negotiated and secured. People route around governance when the sanctioned path is slower than a credit card, so the catalog only works if requests against it are approved in days, not weeks.
Offboarding that reclaims licenses
License reclamation belongs in the employee offboarding workflow, automated through the identity provider. Former employees who keep access to SaaS apps are both a security risk and a recurring cost.
The KPIs that show whether it’s working
App count was the headline metric for a decade. On its own it is now misleading, because it can fall while spend rises. Track these five together:
- SaaS spend growth versus headcount growth. If spend grows faster than headcount with a flat app count, the growth is coming from price increases or metering. Investigate which.
- License utilization. Share of paid seats active in the last 30 days, defined consistently. Zylo’s 54% average is a useful external comparison; a mature program should aim well above it.
- Owned spend share. Percentage of total SaaS spend with a named owner and a known renewal date. This is the single best indicator of control.
- Metered spend share and variance. Portion of SaaS spend that is consumption or AI based, and how far it ran from forecast each month.
- Renewal outcome. Savings or avoided increases at renewal versus the vendor’s opening offer. With 79% of IT leaders seeing increases at renewal, holding flat is itself a result worth reporting.
SaaS sprawl is a condition to manage, not a project to finish. The organizations that keep it contained build discovery, metering controls, and renewal ownership into their monthly operating rhythm, pairing SaaS spend management with disciplined software license management and a regular SaaS audit. For the broader picture of why the bill keeps rising when portfolios don’t, see Your SaaS Portfolio Stopped Growing. The Bill Climbed 8% Anyway, and for detection methods in depth, our guide to shadow IT discovery.
