From Madhur Jain | Product & Market Analysis

SaaS Spend Management Data: AI-Native Apps Explain 11% of Enterprise Growth

On this page

AI-native apps account for about 11% of the growth in large-enterprise SaaS budgets, on a calculation from Zylo's own 2026 figures. The other 89% arrives through less visible lines. SaaS spend management data from Zylo and Vertice points to AI being added to budgets, with no measured evidence of it being funded by cancelled apps. Productiv, the third index in this comparison, no longer operates.

Key takeaways

  • AI-native apps are a small part of the visible increase. At companies above 10,000 employees, Zylo's figures imply AI-native apps added about $3.75M of roughly $33.8M in new annual SaaS spend, or about 11%.
  • App counts do not agree on direction. Zylo reports 305 apps per company, down 0.07%, while Vertice reports 144, up from 131 in a year. The two vendors serve different customers and count differently.
  • No public index page reviewed here publishes an average renewal uplift. Zylo's blog states that 79% of IT leaders met a renewal increase, which is a frequency. Vertice publishes an inflation rate without stating its method on the page.
  • The displacement question is unmeasured. None of the indexes publishes apps cancelled to pay for AI. Cledara and YipitData describe AI as mostly additive, and the main counter-figure, 45% from Redpoint, reaches us only second-hand.
11%AI-native share of new large-enterprise SaaS spend. Source: Zan Digital calculation from Zylo's 2026 index.
14.3%SaaS inflation, September 2026, early data. June set a 16.4% record. Source: Vertice, October 2026.
305 vs 144Average SaaS apps per company at Zylo and at Vertice. Source: Zylo and Vertice, 2026.

Short answer

AI tools are mostly being added to software budgets rather than paid for by cancelling other apps, as far as public data can show. At large enterprises, AI-native apps explain about 11% of spend growth. The larger AI cost probably sits inside existing vendor contracts, and no index isolates it.

SaaS management index data, side by side

Three vendors built their reputations on contract and expense data, and each one reports a different thing under a similar label. The table below sets each headline next to the method behind it, because a figure without its sample cannot be compared with anything.

Every row is a vendor describing its own customers. That is the single most important fact about this table, and it is why I would not quote any cell without its source name attached. BetterCloud appears as a stand-in because the Productiv row is now a historical record.

SaaS spend indexes compared: stated method and headline figures, as read November 2026
IndexMethod and sample as statedAI-native spendApps per companyRenewal price
Zylo 2026 SaaS Management Index, 29 January 2026, with an AI update on 23 September 2026Platform data. 40M+ licences and $75B under management in the January release; the report page now says $100B+ discovered. Number of companies not stated.Up 108% overall and 393% above 10,000 employees (January). Up 334% and 507% (September).305 average, 240 median, down 0.07%79% of IT leaders met an increase at renewal, per Zylo's blog. Sample not stated. No average uplift.
Vertice inflation, AI and benchmark pages, October 2026Contract data. Over $75bn processed, 250,000+ contracts, 32,000+ vendors, 32 markets. AI index window December 2024 to June 2026. Number of companies not stated.Average AI-native spend per company up 41% in Q3 2026, tripled in 12 months.144 in Q2 2026, up from 131 in June 2025. 146 is a modelled figure for September.SaaS inflation 14.3% in September (early), 16.4% June record. Method not stated on the page.
Productiv State of SaaS, latest edition traced is 2024. Company in assignment for creditors since 25 June 2026Usage telemetry. Nearly 100M licences and 100B+ data points (June 2023 release). Number of customers not stated.No AI-native spend figure found. ChatGPT ranked 11th among shadow apps in 2023.371 in the 2023 release; 342 in the 2024 analysis, as cited by Torii from CIO Dive.No figure found.
BetterCloud State of SaaS 2026 (stand-in)Survey. 525 respondents. Sourcing not stated on the page.27 AI-powered apps on average; 22% of tech stacks AI-powered.118 projected for 2026, up 11% on the chart.No figure on the page.

Read from public pages on 8 November 2026. Zylo's full PDF and Vertice's inflation report are gated, and they were not read. A blank here means the figure is absent from the public page, not that it does not exist anywhere.

Productiv's series ended with the company

Productiv made a general assignment for the benefit of creditors effective 25 June 2026, according to the notice held by its claims agent. 1Password reported that customers were told on 2 August that the platform would close on 6 August. Torii and Josys, both competitors, report that access ended on 5 August and quote Productiv as saying all systems and backups were destroyed.

The productiv.com address now redirects to a Dallas logistics company. Any buyer still citing a Productiv benchmark is citing a dataset that cannot be rechecked or extended. I found no 2025 or 2026 edition, and the numbers in the third row should be treated as a closed series.

AI app spend growth: why 108%, 334% and 3x are different numbers

AI app spend growth is reported at wildly different rates because the vendors measure different things across different windows. Zylo's January release gave 108% overall. Its September release gave 334% overall and 507% above 10,000 employees. Neither release explains how the first number became the second.

Vertice reports a third shape. Average AI-native spend per company rose 41% in Q3 2026 and has tripled in 12 months, with 55% of companies paying for at least one AI-native tool. Tripled is close to Zylo's September rate only by coincidence of arithmetic, since the two populations are not the same.

A further reason for caution is that these rates start from a very small base. Growth of 393% on a line under $1M is a large percentage and a modest cheque. Percentages travel well in board decks, and I would ask for the dollar figure every time one appears.

AI-native spend and Applied AI spend are separate lines

Zylo defines AI-native as apps where AI is core to the product. Vertice uses the same idea and separates it from Applied AI, which it describes as paid-for AI capabilities added to an established SaaS tool. That split decides almost everything in this debate.

Vertice's widely quoted claim that nearly 42% of SaaS spend is directed towards AI technology sounds like a budget takeover. The page also says Applied AI contracts include the traditional SaaS suite, so a whole contract may be counted because one module carries AI. I read the 42% as the share of spend exposed to AI-selling vendors, and I would not read it as spend on AI.

The purchasing route has moved too

Zylo's January release says expense-based SaaS spend rose 267%, with ChatGPT now the most expensed app. AI tools bought on a corporate card miss the procurement step where a cancellation would normally be negotiated. That matters to the funding question, because a tool that nobody approved cannot have been funded by a tool that somebody cancelled. The site's analysis of how SaaS sprawl gets rationalised covers the audit side of that problem.

AI-native apps explain about 11% of enterprise SaaS spend growth

AI-native apps account for about 11% of new spend at the largest companies. This figure is a calculation of ours rather than a number Zylo publishes. Zylo gives enough inputs to derive it. Spending at the largest companies grew by about $33.8M, and AI-native apps supplied about $3.75M of that increase.

Working: AI-native share of new large-enterprise SaaS spend
InputAI-native appsTotal SaaS
Current average annual spend$4.7M (10,001+ employees)$245M (large enterprises)
Year over year growth, as stated393%About 16%
Implied prior year (current divided by 1 plus growth)$0.95M$211.2M
Dollars added in the year$3.75M$33.8M
AI-native share of dollars added$3.75M divided by $33.8M = 11.1%

Inputs: Zylo's report page chart text ($4.7M), January release (393%), report page FAQ (about 16%), and Zylo's February 2026 blog ($245M). Zylo does not define "large enterprise", so pairing $245M with the 10,001+ band is an assumption. Moving the 16% to 12% or 20% puts the answer at 14% or 9%.

Where the new large-enterprise SaaS dollars went Average added annual spend per company, about $33.8M, derived from Zylo 2026 index figures All other SaaS: $30.05M (89%) AI-native: $3.75M (11%) The blue block includes AI features sold inside existing suites, usage charges, price rises and volume. Source: Zylo 2026 SaaS Management Index; share computed by Zan Digital.
Illustrative of the arithmetic, not a Zylo chart. The red block is the spending people argue about. The blue block is the one that lands on a renewal.

What sits in the other 89%

The remaining 89% is not AI-free. Zylo attributes the roughly 16% rise to "the growing impact of AI features and consumption-based pricing", which places AI inside the other block as well. An add-on sold into a suite you already own shows up as a larger bill from an existing vendor, and it never appears as a new application.

That is my main conclusion from this data. The AI line finance can see is small, and the AI cost that does the budget damage is likely to be the one disguised as a renewal. Treat 11% as a floor for AI's contribution to growth, since Applied AI, credits and bundled uplift all sit outside it.

Apps per company: flat at Zylo, rising at Vertice

Apps per company point in opposite directions depending on the index, so a flat line at one vendor cannot be read as market consolidation. Zylo reports portfolios holding at 305 apps, with a change of 0.07%. Vertice reports 144 in Q2 2026, up from 131 in June 2025, a rise of 9.9% by my arithmetic. BetterCloud's survey chart shows 11% annual growth to a projected 118.

One question, six answers: SaaS apps per company Each figure describes that vendor's own customers or respondents Productiv 2023 release371 Productiv 2024 analysis342 Zylo average, 2026305 Zylo median, 2026240 Vertice, Q2 2026144 BetterCloud, projected 2026118
Grey bars belong to a company that has since closed. The spread is a customer-mix effect, so benchmark against the vendor whose customers look like you.

Why a flat count can hide a lot of churn

Zylo's release also says large enterprises add an average of 21 applications per month, which is 252 a year. If totals are flat and that holds for the same companies, roughly as many apps leave as arrive. Zylo calls this "ongoing tool turnover as teams make room for AI apps", and I would treat that sentence as interpretation. The release publishes additions only, so removals are inferred.

An app count also says nothing about seats or dollars. A company can keep all 305 apps, downgrade 40 of them, and report a flat portfolio. That is why the next section, on price against spend, carries more weight than this one.

SaaS renewal price increases: frequency is published, size is not

SaaS renewal price increase data is thinner than the volume of commentary suggests. Zylo's statistics blog says 79% of IT leaders encountered price increases at SaaS renewal in the past 12 months. That tells you how often a rise occurred. It does not tell you how large it was.

Other Zylo survey findings are close but distinct. In a survey of 218 IT leaders, 78% reported unexpected charges tied to consumption or AI pricing, and 61% cut projects because of unplanned SaaS costs. Those are useful warning signs, though neither is a price uplift. For what happens when a legacy rate expires, the piece on grandfathered pricing at renewal goes further, and the tally of AI price increases at seven vendors gives named cases.

Vertice's inflation series is the nearest thing to a size measure. It reached 16.4% in June 2026, fell to 12.9% in July and 12.7% in August, then rose to 14.3% on early September data. Vertice notes that a renewal quoted in June faced a market rate nearly 4 points higher than one quoted in August, so timing alone can move a renewal by real money.

Prices rose far faster than spend per employee Vertice SaaS inflation by month, 2026, against spend per employee growth, annualised from Q1 to Q3 16.4% 12.9% 12.7% 14.3% Spend per employee, about 5.5% annualised JuneJulyAugustSeptember (early) Source: Vertice, October 2026. Annualised rate computed by Zan Digital.
Both series come from one vendor, which is what makes the gap worth noticing. The annualised dashed line is our arithmetic, and Q3 data is preliminary.

The gap between price and spend inside Vertice's own data

Vertice reports SaaS spend per employee of $9,450 in Q3 2026, up 2.7% from Q1. Annualised, that is about 5.5%. Inflation over the same months ran between 12.7% and 16.4%. Taken together, those figures imply buyers received about 6% to 9% less software per employee than a year of flat volume would have delivered.

This is the most interesting figure in the dataset, and it is an inference of mine, not a Vertice finding. It suggests buyers are absorbing price rises by trimming seats, tiers and scope inside contracts they keep. If AI is being funded by cuts anywhere, this is where I would look first, because it is the only series pair that shows budget pressure in the data.

Software budget consolidation: who says AI is replacing tools

Software budget consolidation is widely claimed and thinly measured, and the indexes themselves do not settle it. Neither Zylo nor Vertice publishes apps cancelled to fund AI. Vertice's AI index identifies no category losing spend, while Zylo's February blog answers the question directly with "AI is expanding SaaS, not replacing it". Zylo sells software to manage SaaS, so that answer comes from an interested party.

Is AI spend additive or funded by cuts? Evidence read November 2026
SourceDataFindingMain weakness
YipitData, May 2026, updated August100-company cohort from a 1,300+ panel; 5 interviews, midsize firmsAI growth "generally isn't being funded by cutting software"AI spend means OpenAI, Anthropic and Cursor only; the qualitative layer is 5 people
Cledara, March 2026Anonymised platform transactions, sample not statedAI is "mostly additive"; 26.4% of purchases by countShare is by transaction count, and the page also cites 12.5% of transactions and 5.6% of spend
Redpoint, March 2026141 CIOs, reported by secondary coverage45% tied AI budgets to existing software lines; 54% consolidating vendorsOriginal was not readable; coverage disagrees on whether 45% is of CIOs or of budgets

The honest summary is that the transaction data leans additive and the one survey leans toward reallocation, and those are not contradictory. A CIO can tell a researcher that AI money comes from software lines while the apps stay on the books at reduced scope. Vertice's price and spend gap would be consistent with exactly that.

For measurement tooling on the AI side, the review of FinOps gaps in AI spend tracking is the place to continue. For an independent spend series that is not one of these vendors, see the analysis of the Ramp AI Index.

Where this analysis is weakest

This analysis rests on vendor data, partial public pages and a calculation of my own, and each of those can be wrong. The 11% figure pairs numbers that Zylo never set side by side. If the $245M average covers a different group from the 10,001+ band, the result moves, and Zylo does not define the term.

The sellers of the cure wrote the diagnosis

Zylo and Vertice both earn money when buyers fear waste and inflation. Their samples are their customers, and those customers joined because they already had a spend problem. Neither vendor states how many companies sit behind its averages. Under this site's sourcing rules, the figures here are reported as what each vendor says, and they should be treated as directional.

The gated reports may close these gaps

Zylo's full PDF and Vertice's full inflation report sit behind forms and were not read for this post. Either may contain a renewal uplift, a methodology appendix or a company count. Stating that no public page reviewed publishes an average uplift is accurate. Stating that no such number exists would not be, and the two claims should not be confused.

Frequently asked questions

What is the average SaaS spend per company in 2026?

Zylo's 2026 SaaS Management Index puts average annual SaaS spend at $55.7M, up 8% year over year, with a median of $20.6M. Vertice reports a different unit: $9,450 per employee in Q3 2026, on preliminary data. The averages are skewed by very large customers, so the median or the per-employee figure is the better benchmark for most companies.

How fast is AI-native app spend growing?

Zylo reported growth of 108% overall and 393% at companies above 10,000 employees in January 2026, then 334% and 507% in a September 2026 release. Vertice reports average AI-native spend per company rising 41% in Q3 2026 and tripling in 12 months. The figures differ because the samples and periods differ, and neither vendor reconciles them.

How much do SaaS prices increase at renewal?

No public index page reviewed here publishes an average renewal uplift. Zylo's blog says 79% of IT leaders encountered price increases at renewal in the past 12 months, which measures how often, not how much. Vertice publishes a SaaS inflation rate of 14.3% for September 2026, preliminary, without stating on that page how the rate is calculated.

Is AI replacing SaaS tools or adding to software budgets?

Most of the transaction evidence points to addition. YipitData says AI growth generally is not being funded by cutting software, and Cledara calls AI tools mostly additive. Zylo's app count was flat while AI-native spend doubled. The counter-evidence is a Redpoint survey of 141 CIOs, reported second-hand, in which 45% linked AI budgets to existing software lines.

What happened to Productiv's State of SaaS report?

Productiv made a general assignment for the benefit of creditors effective 25 June 2026, according to the claims agent Stretto. Customers were told on 2 August 2026 that the platform was closing. No 2025 or 2026 edition of its State of SaaS series could be found, and competitors report that its underlying data has been destroyed.

How many SaaS apps does the average company use?

It depends on whose customers are counted. Zylo reports 305 on average and 240 at the median. Vertice reports 144 in Q2 2026. BetterCloud's 2026 survey charts a projected 118. Productiv's last figures were 371 in a June 2023 release and 342 in its 2024 analysis. Compare yourself to the index whose customer base resembles your company, not to the largest number.

Where to start on your next renewal

Take your 10 largest renewals in the next two quarters and split each year-on-year change into 3 columns: price, volume and AI features. Most finance teams can do this from the last two order forms. If your vendor bundles AI so that the third column cannot be separated, ask for it in writing before you sign.

Then calculate your own version of the 11%. Divide the dollars you added to AI-native apps last year by the dollars you added to software overall. If the answer is small and your budget still jumped, the pressure is coming from existing contracts, and that is where your negotiating hours belong. The comparison of AI bundling and metering sets out which contract structure to push for.

References

  1. Zylo, The 2026 SaaS Management Index, public report page, read 8 November 2026. Used for spend, app counts, the AI-native definition, the 16% figure and the $4.7M chart text.
  2. Zylo, 2026 index press release, 29 January 2026, and AI-native spend release, 23 September 2026. Used for growth rates, licences, the 218-person survey and app additions.
  3. Zylo, SaaS statistics, updated 8 February 2026, and 2026 predictions, 4 February 2026. Used for the 79% renewal figure and the $245M average.
  4. Vertice, SaaS inflation rate, spend per employee and apps per business, all last updated October 2026. Q3 figures are preliminary or modelled.
  5. Vertice, AI Market Intelligence Index, 12 October 2026. Used for AI-native spend, the 42% share and the contract sample.
  6. Stretto, Productiv claims notice; 1Password, shutdown post, 5 August 2026; Torii, timeline and statistics; SalesTechStar, Productiv 2023 release, 21 June 2023.
  7. BetterCloud, State of SaaS 2026, read 8 November 2026. Used for the 525 respondents and the app figures, including chart text.
  8. YipitData, AI early adopter series, 13 May 2026; Cledara, AI eating SaaS budget, 25 March 2026; Let's Data Science, coverage of the Redpoint update, 29 March 2026, secondary only.

Weakest thing about this source base: every index here is vendor research on the vendor's own customers, none states a company count, and the 45% Redpoint figure could not be traced to its original. All figures last verified 8 November 2026. Vertice's series update monthly, so recheck them before quoting.

AV
Madhur Jain
Founding Member, Zan Digital. Writes about AI product economics, B2B software markets and what the numbers behind vendor claims actually say.

Related reading