The State of SaaS: Build vs. Buy, October 2026

· HissingKitty ·

13 min read Original article ↗

Executive Summary

Everyone is arguing about whether AI makes building cheaper than buying. This report shows what teams actually pay for software, and what the few who built their own say it cost them. It covers 404 data points collected from October 17, 2026 to October 31, 2026.

For Finance Leaders

No verified price increases logged yet this month. Adobe leads the Software Hate Index at 21%. How to cut software costs · Run a Spend Audit.

Key Findings

  1. What teams pay. Buyers shared 176 yearly bills across 67 tools; ServiceTitan has the highest median, about $120,000 a year (n=3).
  2. "Build part of it" is the most common verdict. Our editors rate 62 categories: 15 "Build it", 35 "Build part of it", 12 "Keep paying". Most teams keep the core system and replace the costly slice of seats or features around it.
  3. Most builds state no saving. 23 of 26 documented builds state no saving at all. Treat the savings that are claimed as early and self-reported.
  4. Alternatives exist at scale. 268 AI-native alternatives and custom builders are listed against 205 incumbent tools.
  5. Building is a mainstream option. 32% of respondents at organizations using AI say they skipped buying at least one software product or feature to build it with AI coding tools (McKinsey Global Survey on the State of AI, 2026, n = 1,719).

A Note on This First Report

We launched HissingKitty in October 2026. We expected teams to replace bloated software slowly, over several years. Instead, within two weeks, buyers shared hundreds of bills, reviews and builds. We did not see that coming.

This is our first report. It was a lot of work, and it is a passion project. The people behind it spent 20 years building, selling and growing enterprise software companies. We are not hobbyists guessing at how software is bought; we have sat on the vendor side of these renewals.

We believe this is the biggest shift in business software we will see in our careers. That is our view, not a finding.

Why This Research Matters

For two decades, renewing software was the default. The cost of building software has fallen sharply since 2024, which changes the math at every renewal. Finance and operations leaders need independent evidence of what peers pay, where dissatisfaction concentrates, and when replacing a tool is justified.

Methodology at a Glance

Data Points
404: 40 reviews, 162 public data points, 176 anonymous yearly bills, 26 documented builds

Window
October 17, 2026 to October 31, 2026

Coverage
67 tools with shared bills; 268 alternatives listed (listings that did not meet our rules, such as hosting companies, were removed during review, while new listings added more tools covered)

Independence
Reviews moderated before publication; no vendor pays to appear or rank. Full methodology

01

Headline Number#

$6,836,836 in yearly bills shared across 67 tools. The sum of 176 anonymous yearly bills shared in October 2026. It is neither market spend nor verified savings.

The short version of October 2026: 268 new alternatives and builders were listed against 205 tools, led by HubSpot and Salesforce, tied with 17 each. The largest shared bill was NetSuite at about $2,000,000 a year (one bill). The grid below places each tool with enough bills on both measures at once.

The Replacement Grid, October 2026

HubSpot and Salesforce Drew the Most New Alternatives

  1. 1

    ServiceTitan logo

    ServiceTitan
  2. 2

    HubSpot logo

    HubSpot
  3. 3

    Salesforce logo

    Salesforce
  4. 4

    Adobe logo

    Adobe
  5. 5

    ZoomInfo logo

    ZoomInfo
  6. 6

    Microsoft Teams logo

    Microsoft Teams
  7. 7

    AWS logo

    AWS
  8. 8

    Klaviyo logo

    Klaviyo
  9. 9

    Webflow logo

    Webflow
  10. 10

    Intercom logo

    Intercom
Tools with at least 3 shared yearly bills and one new alternative in October 2026. Up: shared yearly bill (median where more than one). Right: new AI-native alternatives and builders listed. Window October 17, 2026 to October 31, 2026. Source: HissingKitty Research.

Top 10 Tools Under Pressure, October 2026

#ToolYearly billNew altsNew alternativesScore
1ServiceTitan$120,000 (n=3)668
2HubSpot$42,800 (n=7)1768
3Salesforce$40,000 (n=5)1767
4Adobe$840 (n=9)1339
5ZoomInfo$33,500 (n=6)632
6Microsoft Teams$5,300 (n=6)1032
7AWS$6,500 (n=7)826
8Klaviyo$36,000 (n=3)118
9Webflow$550 (n=8)618
10Intercom$10,248 (n=3)416

Only tools with at least 3 shared yearly bills are ranked, so one large bill can't lead. Score out of 100: half from the tool's shared yearly bill compared with the month's largest, half from its new alternatives compared with the month's most. A ranking of pressure, not of product quality.

02

Key Findings#

  1. Buyers shared 176 yearly bills across 67 tools in October 2026, totaling $6,836,836.
  2. NetSuite has the largest shared yearly bill this month, about $2,000,000.
  3. LinkedIn is next, at about $262,854 a year (median of 2).
  4. HubSpot and Salesforce tied for the most new alternatives this month, 17 new listings each, out of 268 listings in total.
  5. Figma and HubSpot tied as the most reviewed tools, with 11 reviews and data points each.
  6. 26 documented builds this month, reporting about $13,300 a month in combined savings.
  7. No verified price increases were logged October 17, 2026 to October 31, 2026.
  8. Outside source: 32% of respondents at organizations using AI said they decided against buying at least one software product or feature to build it with AI coding tools instead (McKinsey Global Survey on the State of AI, 2026, 1,719 respondents).

No verified replacement reports yet, so realized build payback cannot be measured this month.

03

What Changed#

This first monthly edition sets the baseline, so there is no earlier month to compare.

What This Means

Every number in this edition becomes the baseline for the next. From the next edition on, this section compares the same measures month over month, so readers can see whether alternatives, bills and price increases are rising or falling.

04

What Buyers Paid#

ServiceTitan Has the Highest Median Yearly Bill, $120,000

Only tools with at least 3 shared yearly bills; n on each bar. 26 tools of 67 qualify. Window October 17, 2026 to October 31, 2026. Bills shared by buyers; bars show the median. HissingKitty Research, October 2026

The single largest bill, about $2,000,000 for NetSuite (one bill), is 29% of the month's $6,836,836 total, so the total says more about who shared than about typical spend.

Show all 67 tools and bills
Shared Yearly Bills, October 2026
ToolBillsYearly Bill
NetSuite1About $2,000,000
LinkedIn2Median $262,854
ServiceTitan3Median $120,000
WorkOS1About $103,000
Trimble1About $60,000
Drata1About $47,600
GitHub1About $47,184
Jobber2Median $45,174
HubSpot7Median $42,800
Salesforce5Median $40,000
Autodesk1About $40,000
Dialpad1About $38,700
Klaviyo3Median $36,000
ZoomInfo6Median $33,500
Retool1About $25,000
Vanta2Median $18,442
Zendesk1About $13,416
Intercom3Median $10,248
ClickUp3Median $9,400
ScrapingBee1About $7,200
AWS7Median $6,500
Microsoft Teams6Median $5,300
Mailchimp4Median $4,800
Asana1About $4,800
Riverside1About $4,800
Shopify6Median $4,650
Chargify1About $4,400
Apollo2Median $3,060
Slack4Median $2,974
Hootsuite1About $2,388
Atlassian5Median $2,300
Zapier2Median $2,158
Kajabi1About $2,148
Beehiiv8Median $2,028
Ahrefs2Median $1,968
Loom3Median $1,800
QuickBooks7Median $1,680
Semrush1About $1,679
Typeform3Median $1,548
monday.com3Median $1,500
DocuSign3Median $1,440
Figma2Median $1,346
Jotform2Median $1,008
Adobe9Median $840
Bitrix241About $828
Linear1About $750
Clio1About $720
Webflow8Median $550
Supabase2Median $510
Squarespace3Median $480
Airtable1About $432
Wix1About $380
Descript1About $348
HeyGen2Median $348
Xero1About $324
Cursor2Median $300
ClickFunnels1About $300
Notion5Median $288
Calendly3Median $240
Vercel3Median $240
PandaDoc1About $223
Harvest1About $200
Canva2Median $192
OpusClip1About $180
Evernote3Median $170
Todoist1About $120
Macrofactor1About $100

What This Means

NetSuite sits at the top of the sample at about $2,000,000 a year, followed by LinkedIn at $262,854. 38 tools have more than one bill, so their figures are medians: LinkedIn (2), ServiceTitan (3), Jobber (2), HubSpot (7).

Editor's view: a large bill is where the question of building starts, not where it ends. The tools at the top of this list tend to be systems of record (CRM, data, compliance), where switching costs, integrations and audit needs matter as much as the license. Compare your own bill against these figures on each tool's page before deciding.

The Software Hate Index, End of October

The board as it stood on October 31. This snapshot no longer changes. This is the first edition, so it sets the baseline.

  1. 1

    Adobe logo

    Adobe21% of the hate
  2. 2

    Microsoft logo

    Microsoft13% of the hate
  3. 3

    Webflow logo

    Webflow11% of the hate
  4. 4

    Salesforce logo

    Salesforce10% of the hate
  5. 5

    QuickBooks logo

    QuickBooks8% of the hate
  6. 6

    Shopify logo

    Shopify8% of the hate
  7. 7

    HubSpot logo

    HubSpot8% of the hate
  8. 8

    Atlassian logo

    Atlassian7% of the hate
  9. 9

    Figma logo

    Figma7% of the hate
  10. 10

    AWS logo

    AWS7% of the hate
See the live rankings →

05

What's Replacing Them#

231 AI-native alternatives and 37 builders listed in October 2026, grouped by the tool they replace. Votes are from visitors, one per person an hour.

Show 4 more

Most Reviewed This Month

ToolReviews and postsAverage
Figma114.0
HubSpot113.0
Intercom93.0
Adobe91.6
Salesforce91.5
Show all 12

n = 40 approved reviews plus 162 approved public posts in October 2026. Star averages come from written reviews only; public posts link to their source. HissingKitty Research.

What This Means

HubSpot and Salesforce tied for the most new alternatives this month (17 each). More listings mean more builders see room to compete on price or fit; they do not yet show that buyers are switching.

Votes are the early buyer signal. A listing with net positive votes has been checked by visitors who pay for the tool it replaces. Ties and low vote counts are common this early, so treat the order within a tool as provisional.

Monthly Badges

The top-voted listing for each tool earns "Most Voted Alternative" for the month, once it has at least 3 net votes.

06

What Teams Built Instead#

public build posts
public build posts

tools replaced
tools replaced

stated savings, from 3 posts
stated savings, from 3 posts

median build time, from 3 posts
median build time, from 3 posts

HubSpot Is the Tool Teams Built Away From Most

Posts per tool, n = 26 posts. A post that names several tools counts for each. 16 more tools in the full list below. HissingKitty Research, October 2026

Biggest Stated Savings

  • AWS logo

    Simon Høiberg left AWS · Oct 27

  • HubSpot logo

    snowflakesociety left HubSpot · Oct 23

  • ScrapingBee logo

    Eugene Zolotarenko left ScrapingBee · Oct 27

As stated in each post. 23 of 26 posts did not state a saving.

See all 26 posts⌄

These are builders' own claims from public posts, not verified replacement reports. Staff time freed is not necessarily a cash reduction.

What This Means

26 teams documented building their own tool this month; 23 of them state no saving at all, and the rest claim about $13,300 a month combined. These are the builders' own numbers, often shared the week the tool went live, before upkeep costs show up.

Editor's view: most builds replace a workflow or a reporting layer, not the system of record underneath it. That is why "Build part of it" is the most common verdict on our category pages.

What a Claimed Saving Leaves Out

  • Support: who fixes it when it breaks, including nights and weekends.
  • Maintenance: updates, security patches and staff hours every month.
  • Compliance: audits, certifications and data handling rules.
  • Migration risk: data export, downtime and retraining staff.
  • Usage: many teams use a small share of a tool's features, so the fair comparison is the cost of that share, not the whole license.

07

The Wider Market#

32% of respondents at organizations using AI say they skipped buying at least one software product or feature to build it with AI coding tools instead. That figure comes from an outside survey of 1,719 people, fielded May 4 to June 8, 2026. It is not HissingKitty data and is not counted in any of our numbers.

About a Third of Organizations Chose to Build Instead of Buy

Share of respondents, by industry

  • Technology41%n=199
  • Healthcare payers and providers39%n=47
  • Professional services38%n=205
  • Energy and materials38%n=153
  • Financial institutions36%n=148
  • Media and telecom34%n=71
  • Pharma and medical products33%n=55
  • Travel and logistics31%n=56
  • Consumer goods and retail29%n=116
  • Engineering and construction28%n=72
  • Advanced manufacturing27%n=95
  • Insurance19%n=40
  • Public and social sector17%n=71
  • All industries32%
Asked only of respondents whose organizations regularly use AI in at least one business function. Fielded May 4 to June 8, 2026. Source: McKinsey Global Survey on the State of AI, 2026. Chart redrawn by HissingKitty Research.

What This Means

Technology, healthcare and professional services lead, at 38% to 41%. Insurance and the public sector trail, under 20%, where rules and audits make owning software harder. Even in the lowest group, about one in six respondents say they have made this choice.

Editor's view: the survey asks whether teams skipped a purchase, not whether they cancelled a contract. Most of these builds are likely extra features or small tools, which matches what we see in the builds logged above. Read it as the "Build part of it" path going mainstream.

08

Price Increases Logged#

No verified price increases were logged October 17, 2026 to October 31, 2026. A change counts only when a buyer or vendor states both the old and new price for the window.

09

Notable Stories#

10

Proof Points#

3 public build posts this month state a monthly saving. Our 176 shared bills establish what this sample paid, but do not establish that replacing those tools would save money.

Against building, the editor's view is to keep paying for identity and security compliance. A high shared bill alone does not settle reliability, audit requirements or liability.

Our editors rate 62 categories: 15 Build it, 35 Build part of it, 12 Keep paying. These are judgments, not buyer outcomes.

11

Predictions#

Our view. These are forecasts for the next monthly edition, not measured findings.

Likely

Partial builds will remain the largest editorial category group.

Wrong if: A different group leading the next edition's category count.

Possible

The next edition will include a verified replacement report.

Wrong if: No verified replacement reports at publication.

Long shot

Verified replacement evidence will show staff time savings exceeding license savings.

Wrong if: No verified report demonstrating that comparison in the next edition.

12

What to Do Next#

Editor's view.

Build it

Start with a bounded internal workflow on data you control. Budget for maintenance and test an export before switching.

Build part of it

Keep necessary infrastructure and replace a workflow around it. Compare license savings and staff time separately.

Keep paying

Keep systems where security, certification or regulated operations exceed your ability to maintain them.

Check Your Own Stack

  1. Run a spend audit: find seats and tools you no longer need.
  2. Open the Price Radar: see if your vendors changed prices.
  3. Write a Review: add your rating so next month counts you.

13

Methodology and Limits#

Launch edition: data starts October 17, 2026, so this edition covers about two weeks. Bills include prices teams stated in public posts. November is the first full month.

Window October 17, 2026 to October 31, 2026. The bill sample is 176 anonymous yearly bills shared in that window. Every tool is named from its first bill, with the number of bills beside it. The number of distinct companies is unknown.

Public data points: 188 data points logged this month, each linked to its source. Claims are self-reported. Verified replacement reports: 0.

Limits of This Sample

  • Self-selected: people who share bills and post reviews may be less happy with their vendors than the average buyer.
  • Short window: about two weeks of data, so most tools have only a few data points.
  • Self-reported: build times and savings are the builders' own claims, usually posted soon after launch.
  • Survivor bias: teams whose builds failed rarely post about it.
  • Not measured: market pricing, representative savings, build failure rates, ongoing support costs or realized cash savings.

We do not publish private seat counts, headcounts or per-person prices. Read our methodology.

Data is summarized on this page; raw data points are not published, to protect the people who shared them.

14

Sources#

  1. McKinsey Global Survey on the State of AI, 2026

26 documented builds come from public data points. Each is linked next to its number in What Teams Built Instead. Reviews and bills come from HissingKitty submissions.

15

Updates to This Report#

  • October 31, 2026: Final edition, data through October 31. The pressure ranking now needs at least 3 shared bills per tool, so a single large bill no longer leads it. Added key facts, sources and this log.

16

Cite This Report#

For Press and Analysts

Free to quote and republish with a link. Every number shows its sample size and source.

HissingKitty, The State of SaaS: Build vs. Buy, October 2026, published October 19, 2026, final October 31, 2026, https://hissingkitty.com/research/build-vs-buy-october-2026

Share this report:XLinkedIn

Figures follow our methodology. Estimates are labelled until backed by submitted savings reports.