Upwork Automation Market: Platform & Application Comparison
July 2026 Analysis
Report Date: August 2026 Data Period: June 2025 - July 2026 (baseline anchored to June 2025) Total Jobs Analyzed: 2,549 (platform set, July 2026)
Executive Summary
July 2026 is the first fully-clean month on the new hybrid scrape basis. On 2026-06-24 a dedicated platform query (Zapier/n8n/Make terms) began running daily alongside the broad automation query, stepping platform-tagged capture from ~52 jobs/day to ~85 jobs/day. That means July's platform total of 2,549 is not comparable to June's 1,795 — the +42% headline is almost entirely a measurement change, not demand. On a per-day basis over comparable post-cutover windows, volume is flat-to-marginally-softer: June 24–30 ran 85.7 platform jobs/day, July ran 82.2/day. The honest read of July volume is flat.
The capture-independent signals are the ones that matter across the cutover, and they are calm. The median hourly rate held at $35 — where it has sat in 11 of the last 12 months. The mean drifted to $41.41/hr, inside the same narrow $38–43 band the series has occupied all year; this is not a shift. Platform share is stable: Zapier leads at 42.1% of platform mentions, n8n sits at 36.3%, Make.com at 21.7% — the same standing structure as every prior month.
The one genuinely new, multi-month signal is fixed-price adoption. Fixed-price work is now 38.6% of postings, the third straight month tracking near 38–43%. Combined with the flat $35 median, clients are increasingly scoping automation as defined deliverables rather than open-ended hourly engagements. AI mentions remain heavy across the platform set — Claude/Anthropic (621) and OpenAI/GPT (643) are the most-referenced categories — and the AI tier is broadening, not just deepening.
Key July Insight: Ignore the volume number this month. July is a coverage cutover, not a rebound — per-day capture is flat. What's real and durable: the median rate has not moved off $35, platform share is unchanged, and the fixed-price share keeps climbing. From August 2026 onward, month-over-month platform volume will be comparable again on this clean basis.
Platform Comparison: Twelve-Month Evolution
Major Automation Platforms
| Platform | Aug'25 | Sep | Oct | Nov | Dec | Jan | Feb | Mar | Apr | May'26 | Jun | Jul | Jun MoM | Total (vs Jun'25) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Zapier | 2,663 | 2,233 | 2,132 | 1,662 | 1,753 | 1,767 | 1,721 | 1,827 | 1,785 | 1,021 | 972 | 1,223 | 🚀 +25.8%* | 📉 declining |
| n8n | 2,155 | 2,099 | 2,024 | 1,729 | 1,663 | 1,535 | 1,403 | 1,465 | 1,252 | 805 | 815 | 1,055 | 🚀 +29.4%* | 📉 declining |
| Make.com | 1,174 | 967 | 1,027 | 811 | 751 | 725 | 743 | 806 | 781 | 638 | 593 | 631 | 🚀 +6.4%* | 📉 declining |
| Power Automate | 47 | 33 | 42 | 20 | 28 | 37 | 22 | 29 | 30 | 20 | 18 | 30 | 🚀 +66.7%* | 🟢 flat/niche |
Status Indicators:
- 🚀 Strong Growth (>10%)
- 🟢 Stable/Growth (0–10%)
- 🟡 Slight Decline (0–5%)
- 📉 Declining (>5%)
- 💥 Collapse (>20% since June 2025 baseline, sustained)
*The Jun→Jul MoM figures are not real growth. July is the first full month on the hybrid scrape basis; June was under-captured on the old basis. The raw MoM deltas above are measurement artifacts of the 2026-06-24 cutover, not demand changes. Do not read them as market moves.
Platform Analysis
Zapier (Volume Leader — Standing Structure)
- Current: 1,223 jobs (42.1% of platform mentions)
- Trend: Jun→Jul delta is a coverage artifact; share is the real signal, and it's stable in the low-40s
- Total: Down ~30% from peak, in line with the other two platforms
- Assessment: Zapier remains the volume leader — as it has in nearly every month of the series. Not this month's news; standing structure.
- Recommendation: Primary platform for breadth and client volume.
n8n (Technical-Depth Option)
- Current: 1,055 jobs (36.3% of platform mentions)
- Trend: Share holding in the mid-30s; the Jun→Jul jump is capture, not recovery
- Total: Down ~30% from peak — the same magnitude as Zapier and Make
- Assessment: n8n is the technical-depth option, not a uniquely resilient platform. The three moved together off their peaks.
- Recommendation: Premium/self-hosted work; pair with a CRM specialty to insulate from platform wobble.
Make.com (Weakest of the Three)
- Current: 631 jobs (21.7% of platform mentions)
- Trend: Smallest platform, deepest drawdown from peak
- Total: Down ~30% from peak, consistent with the others
- Assessment: Make remains the weakest of the three by volume and share — a standing condition, not a fresh decline.
- Recommendation: Secondary skill, not a primary specialization.
Power Automate (Niche)
- Current: 30 jobs
- Trend: Bounces in the teens-to-thirties; too small to read monthly
- Total: Effectively flat across the window
- Assessment: Niche enterprise tool, not a meaningful share of market.
- Recommendation: Microsoft enterprise ecosystem only.
Application Ecosystem: July 2026
Top 15 Applications by Job Volume
| Application | Aug'25 | Sep | Oct | Nov | Dec | Jan | Feb | Mar | Apr | May'26 | Jun | Jul | Jun MoM* | Status |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GoHighLevel | 647 | 662 | 659 | 579 | 562 | 645 | 585 | 693 | 687 | 476 | 483 | 585 | 🚀 +21.1% | Durable CRM leader |
| Excel | 626 | 549 | 568 | 403 | 382 | 316 | 280 | 251 | 235 | 108 | 171 | 353 | 🚀 +106.4% | Coverage-inflated |
| HubSpot | 490 | 450 | 447 | 346 | 333 | 350 | 349 | 375 | 378 | 205 | 243 | 322 | 🚀 +32.5% | Steady strength |
| Google Sheets | 815 | 719 | 726 | 597 | 552 | 537 | 495 | 526 | 532 | 313 | 261 | 268 | 🟢 +2.7% | Stable |
| Airtable | 731 | 604 | 611 | 505 | 492 | 467 | 424 | 435 | 447 | 308 | 266 | 255 | 🟡 -4.1% | Structural decline |
| Slack | 519 | 460 | 461 | 354 | 321 | 307 | 316 | 336 | 323 | 218 | 157 | 237 | 🚀 +51.0% | Coverage-inflated |
| Notion | 483 | 325 | 414 | 321 | 341 | 273 | 243 | 229 | 230 | 157 | 111 | 132 | 🚀 +18.9% | Long-run decline |
| Salesforce | 146 | 149 | 106 | 81 | 96 | 88 | 110 | 88 | 102 | 57 | 64 | 104 | 🚀 +62.5% | Small-number swing |
| ClickUp | 199 | 166 | 182 | 135 | 169 | 146 | 134 | 112 | 103 | 58 | 81 | 94 | 🚀 +16.0% | Structural decline |
| Zoho | 163 | 124 | 141 | 128 | 122 | 82 | 120 | 101 | 119 | 41 | 66 | 82 | 🚀 +24.2% | Volatile, small |
| QuickBooks | 123 | 93 | 95 | 79 | 64 | 77 | 73 | 94 | 95 | 47 | 70 | 70 | 🟢 0.0% | Flat |
| Monday.com | 125 | 98 | 97 | 101 | 90 | 102 | 101 | 104 | 111 | 60 | 49 | 61 | 🚀 +24.5% | Most stable alt |
| Asana | 100 | 93 | 97 | 67 | 71 | 64 | 65 | 49 | 72 | 27 | 24 | 49 | 🚀 +104.2% | Small-number swing |
| Trello | 78 | 75 | 55 | 48 | 47 | 26 | 24 | 31 | 41 | 11 | 17 | 26 | 🚀 +52.9% | Near-extinct |
| Xero | 58 | 43 | 51 | 40 | 34 | 30 | 37 | 43 | 42 | 17 | 20 | 20 | 🟢 0.0% | Flat, small |
*Every Jun→Jul MoM here is inflated by the same coverage cutover that inflated the platform counts. The large positive prints on Excel (+106%), Asana (+104%), Salesforce (+62%) and Slack (+51%) are dominated by the capture step-up, not by demand. Treat these as measurement, not market moves. Share and multi-month direction are the honest reads.
CRM Platform Deep Dive
Dedicated CRM Systems
| CRM | Jun'25 basis | May'26 | Jun | Jul | Standing Read |
|---|---|---|---|---|---|
| GoHighLevel | Durable leader | 476 | 483 | 585 | Strongest sustained CRM in the dataset — holds near its highs even in soft months |
| HubSpot | Steady | 205 | 243 | 322 | Rock-steady enterprise anchor; strongest CRM after GHL |
| Salesforce | Volatile, small | 57 | 64 | 104 | Small-number swings; don't chase the monthly print |
| Zoho | Volatile, small | 41 | 66 | 82 | Bounces around; ended the prior year down — noise, not a trend |
Key Insight: The July level-ups on all four CRMs are partly the coverage cutover. What survives the cutover is the standing structure: GoHighLevel remains the most durable specialization in this dataset, and HubSpot the steadiest second. The month-to-month CRM "rotation" — Salesforce up, then Zoho up — is small-number noise on counts in the double digits; do not pivot on it. Pick a CRM and go deep; GoHighLevel-for-agencies is the best-supported call in a year of data.
Generic "CRM" Mentions
| Month | Mentions | Note |
|---|---|---|
| May 2026 | 817 | New-basis level |
| June 2026 | 808 | Flat |
| July 2026 | 1,084 | +34.2% MoM — coverage-inflated |
Generic CRM mentions read 1,084 in July, up sharply from June's 808. The jump tracks the same capture step-up that lifted every count this month; it is not a demand surge. Converted to a share of postings — the capture-independent read — CRM has actually declined three months running: 47.2% of jobs in May, 45.0% in June, 42.5% in July. CRM remains the single largest sub-category in the data by a wide margin, but its direction is gently down, not up. The count says growth; the share says otherwise, and the share is the one to trust.
Alternative CRM Solutions
| Application | May'26 | Jun | Jul | Assessment |
|---|---|---|---|---|
| Airtable | 308 | 266 | 255 | Structural decline intact |
| Notion | 157 | 111 | 132 | Long-run erosion; July up on coverage |
| ClickUp | 58 | 81 | 94 | Structural decline; small counts |
| Monday.com | 60 | 49 | 61 | Most stable alternative |
Insight: Alternative CRM tools remain in structural decline over the full series — Airtable and Notion are both far below their 2025 levels regardless of the July capture bump. Monday.com stays the standout among alternatives for stability. Dedicated CRMs continue to outperform the alternative category; that gap is the durable story.
AI Tier
AI keywords are tracked across two populations, kept separate. The platform set (the 2,549 jobs naming Zapier/n8n/Make — the canonical, headline dataset) and the broader automation set (4,145 jobs, the wider net used only for the AI picture).
AI Mentions Within Platform-Automation Work
| AI Category | May'26 | Jun | Jul | Standing Read |
|---|---|---|---|---|
| OpenAI/GPT | 531 | 469 | 643 | Most-referenced AI category |
| Claude/Anthropic | 448 | 464 | 621 | Close second, rising steadily |
| AI Agents | 294 | 326 | 413 | Broadening agent demand |
| RAG | 213 | 266 | 375 | Retrieval work climbing |
| Voice AI | 157 | 192 | 246 | Vapi/Retell/ElevenLabs niche |
| MCP | 51 | 54 | 112 | Small but roughly doubled |
| LangChain | 56 | 54 | 108 | Framework demand present |
| Vector DBs | 38 | 59 | 92 | Infrastructure signal |
| AI Builders | 39 | 40 | 65 | Lindy/Relevance/Gumloop tier |
| Embeddings | 10 | 12 | 29 | Smallest category |
Two-population read: Within the platform-automation set, OpenAI/GPT (643) and Claude/Anthropic (621) are neck-and-neck as the most-referenced AI categories. Across the broader automation market (4,145 jobs), the same two lead — Claude/Anthropic at 783 and OpenAI/GPT at 738 — with AI Agents at 533 and RAG at 501. The July level-ups partly reflect the coverage cutover, so treat the counts as directional, not a clean MoM. The durable signal is composition: agent, RAG, and voice work are broadening beyond the two frontier-model names, and MCP/Vector DB/LangChain mentions — the plumbing of production AI — are now non-trivial. This is the clearest emerging signal in the data: AI is table stakes across platform work, and the premium is in building production systems, not mentioning a model.
July Market Dynamics
Volume — Flat on a Comparable Basis (Do Not Read the Cutover as Growth)
Volume Trends (platform set):
- May 2026: 1,730 jobs
- June 2026: 1,795 jobs (under-captured; cutover landed 2026-06-24)
- July 2026: 2,549 jobs (first full month on the hybrid basis)
Assessment: The raw June→July delta (+42%) is almost entirely a measurement change, not demand. The dedicated platform query began 2026-06-24, stepping platform capture from ~52 to ~85 jobs/day. The honest comparison is per-day over post-cutover windows: June 24–30 ran 85.7 platform jobs/day; July ran 82.2/day — flat, marginally softer. Absolute counts also aren't comparable back across the May-2026 measurement change to Aaron's earlier series. July 2026 is the new like-for-like anchor; from August 2026, month-over-month platform volume is comparable and will be narrated normally.
Rates — Median Flat at $35 (Lead With It)
| Month | Median Rate | Avg Rate | High-Paying Count ($75+) |
|---|---|---|---|
| Jan 2026 | $35 | $40.20 | 165 |
| Feb 2026 | $35 | $43.27 | 170 |
| Mar 2026 | $35 | $40.66 | 167 |
| Apr 2026 | $35 | $39.46 | 139 |
| May 2026 | $38 | $42.20 | 67 |
| Jun 2026 | $35 | $41.83 | 82 |
| Jul 2026 | $35 | $41.41 | 105 |
Analysis: The median hourly rate held at $35 — where it has sat in 11 of the last 12 months. That is the signal that survives every measurement change: it is capture-independent, and it has not moved. The mean at $41.41/hr sits inside the same $38–43 band the series has occupied all year; February's $43.27 was the lone outlier and reverted. Do not read the mean as a shift — the typical job pays $35/hr, and has all year. July's 25th percentile at $20 and 75th percentile at $50 confirm the low end persists; commodity work did not disappear. High-paying ($75+) counts read 105 in July, up on June's 82, but that count moves with capture — treat it as directional.
Contract Type — Fixed-Price Keeps Climbing (The Real Multi-Month Trend)
| Month | Fixed-Price Share | Hourly Share |
|---|---|---|
| May 2026 | 42.9% | 57.1% |
| June 2026 | 40.1% | 59.9% |
| July 2026 | 38.6% | 61.4% |
Assessment: Fixed-price work is 38.6% of July postings. Across the SOTU-era months it has run in the high-30s to low-40s — a real, multi-month lean toward defined-deliverable contracts, and the strongest concrete case for package pricing. Combined with the flat $35 median, the read is consistent: clients increasingly want a scoped outcome, not an open-ended hourly engagement. For consultants, that argues for outcome/package pricing far more concretely than any "raise your rate" exhortation.
Strategic Recommendations
Platform Strategy for Q3 2026
1. Zapier: Volume Leader
- 1,223 jobs (42.1% of platform mentions), stable share in the low-40s
- Down ~30% from peak, same as the other two
- Recommendation: Primary platform for breadth; the volume cushion is real.
2. n8n: Technical-Depth Option
- 1,055 jobs (36.3% share), share holding mid-30s
- Down ~30% from peak — no uniquely resilient platform here
- Recommendation: Premium/self-hosted work; pair with a CRM specialty.
3. Make.com: Weakest of the Three
- 631 jobs (21.7% share), smallest and deepest drawdown from peak
- Recommendation: Secondary skill only.
4. CRM: GoHighLevel and HubSpot Lead
- GoHighLevel (585) is the most durable specialization in the dataset
- HubSpot (322) the steadiest second
- Recommendation: Pick a CRM and go deep; GoHighLevel-for-agencies remains the best-supported call.
Application Strategy
High-Value Specializations:
-
CRM + Automation Specialist ($75–200/hr)
- GoHighLevel durable, HubSpot steady, generic CRM demand resilient
- The most reliable sub-segment across a full year
-
AI Production Engineering ($120–300/hr)
- July premium tier is dominated by AI work: $300/hr Claude automation architect, $200/hr AI outbound specialist, $150/hr AI ops architects and fractional CAIO roles
- The premium is in building production agents/RAG/voice systems, not mentioning a model
-
Accounting Automation ($60–120/hr)
- QuickBooks and Xero steady at small volumes; recurring finance-ops demand
- July's $150/hr finance-operations architect role is representative
-
Multi-System Integration ($80–150/hr)
- Smartsheet, custom CRM, and integration architecture postings recur at $120–125/hr
Avoid/Deprioritize:
- Excel-only, Notion-only, ClickUp-only positioning: all in structural multi-month decline regardless of July's capture bump
- Trello: near-extinct, single-to-double-digit counts, not material
- Generalist platform-only positioning: the market rewards CRM/AI specialization
Rate Strategy for Q3 2026
Current Market (platform set):
- Median: $35/hr (unchanged in 11 of 12 months)
- Average: $41.41/hr (inside the standing $38–43 band)
- 25th pct: $20/hr · 75th pct: $50/hr
- High-paying ($75+): 105 · Ultra-premium ($150+): 9 · Max: $999/hr
Positioning:
- Entry: $20–35/hr — the market's gravity; escape requires real specialization
- Experienced: $50–75/hr — requires a named specialty to compete
- Specialist: $75–130/hr — CRM + platform + AI or industry
- Premium: $130–300/hr — production AI (agents, RAG, voice) + proven builds
Key: The typical job pays $35/hr and has all year. The premium tier is real but scarce (9 ultra-premium postings). Win it with proof of production work, not positioning language — and quote packages, because fixed-price is where the market is drifting.
Q3 2026 Outlook
Expected August–October Trends
- Volume: August is the first month with a clean MoM comparison to July on the hybrid basis; expect it to read as the true near-term signal
- Rates: Median $35/hr likely holds; mean stays in the $38–43 band — treat any single-month mean move as noise until confirmed
- Fixed-price: The high-30s/low-40s fixed-price share is the trend to watch; if it holds, package pricing is the correct posture
- CRM: GoHighLevel and HubSpot remain the durable specializations; ignore the small-count CRM rotation
- AI: Composition broadening — agents, RAG, voice, and the MCP/Vector/LangChain plumbing — is the clearest emerging signal
- Platforms: Report share and shared trend, not a monthly leadership race; all three sit ~30% below peak
Market Structure Evolution
What's Standing (on capture-independent signals):
- Median rate flat at $35/hr — the through-line of the entire series
- Platform share stable: Zapier volume leader, n8n technical depth, Make weakest
- Fixed-price share drifting up toward ~40% — the real multi-month client-behavior shift
- AI is table stakes across platform work; premium is in building, not mentioning
- CRM (GoHighLevel, HubSpot) the most durable specialization
What's Next:
- August restores clean month-over-month volume comparison
- Package/outcome pricing over hourly hand-wringing
- Production AI + a CRM specialty as the highest-margin combination
Methodology
Data Source: Upwork job postings scraped via Apify (platform set: jobs mentioning Zapier / n8n / Make.com / Power Automate) Analysis Method: Keyword search in job titles and descriptions (substring / word-boundary matching) Baseline Anchor: June 2025 (first complete month; May 2025 is structurally invalid — missing its first five days) Rate Calculation: Median and average of posted hourly maximums where present (July sample: 948 hourly postings)
Coverage Notes (important):
- Two measurement changes affect this series. (1) Starting May 2026, the SOTU scrape captures a narrower slice than Aaron's earlier exports — absolute counts are not comparable across that boundary. (2) On 2026-06-24, a dedicated platform query began, stepping platform capture from ~52 to ~85 jobs/day. July 2026 is the first full month on the complete hybrid basis. The Jun→Jul volume increase is a coverage artifact, not demand.
- Across both boundaries, the capture-independent signals are median rate and platform share % — lead with those.
Limitations:
- Jobs may mention multiple platforms (overlap exists); counts are keyword mentions and overcount
- Rates are posted maximums, not contracted rates
- Substring matching may catch incidental mentions (e.g., "excel" in "excellent")
- Sample represents the Upwork marketplace only
Report Generated: August 2026 Next Update: August 2026 data (first clean MoM comparison on the hybrid basis) Questions/Feedback: Via GitHub Issues
This analysis is part of the ongoing Upwork Automation Market Analysis project tracking the evolution of the automation consulting marketplace.