Appen Limited Overview
Appen is a 30-year-old Australian-listed provider of human-generated and human-validated data for the artificial-intelligence lifecycle, spanning data sourcing, data annotation and model evaluation across text, speech, image, video and multimodal content. It operates through two reporting segments: Appen Global, which serves predominantly US-based technology companies and frontier AI laboratories, and Appen China, which serves Chinese large-language-model builders and has become the group's growth engine (revenue up 75% in FY25 and 80% in H1 FY26). The company sits between the AI model builders and a self-reported crowd of over one million contractors in more than 200 countries and territories speaking over 500 languages, orchestrated through its proprietary Mercury, CrowdGen and ADAP platforms. After a collapse from US$447m revenue in FY21 to US$231m in FY25 — driven by digital-advertising weakness and Google's 2024 contract termination — Appen is attempting a profitable-growth turnaround anchored on generative-AI work in both the Western and Chinese AI ecosystems.
Company's own description
In its 2025 Annual Report and ASX boilerplate, Appen describes itself as a global market leader in data for the AI lifecycle, drawing on three decades of experience in data sourcing, data annotation and human model evaluation to enable organisations to launch innovative AI systems. It emphasises its crowd of more than one million skilled contractors (self-reported), coverage of over 500 languages and over 200 countries and territories, and its AI data platform, and lists technology, automotive, financial services, retail, healthcare and government as its customer verticals (Source: FY25 Annual Report; H1 FY26 Release boilerplate). In 2026 communications management added a distinctive framing: Appen claims to be one of the few AI data partners working across both "frontiers" of AI — Western and Chinese model builders (Source: Q2 FY26 4C; H1 FY26 Release).
Independent characterisation
| Revenue model element | Description | Disclosed share of revenue |
|---|---|---|
| Project-based data services (collection, annotation, evaluation) | Bespoke datasets and human judgments priced per task, per hour, or per project | Not publicly disclosed; implied as the large majority |
| Ongoing/"run-rate" programs | Multi-year relevance or evaluation programs (historically Google, now other large clients) | — |
| Platform/software (ADAP) | Self-serve or managed use of the annotation platform | — |
| Off-the-shelf datasets | Licensed pre-built speech/language/image corpora | — |
Appen is, in economic substance, a tech-enabled managed-services business. It wins projects from AI developers (historically search-relevance, ads-relevance and speech projects for the US "Big Tech" platforms; now increasingly LLM supervised-fine-tuning, RLHF, evaluation, red-teaming and multimodal data collection), designs task workflows, recruits and qualifies crowd workers, runs the work through its platforms, applies quality control, and delivers datasets or judgments. The dominant cost line is crowd payments, which makes gross margin structurally modest (Appen-defined gross margin of 40.3% in FY25 and 36.7% in H1 FY26, per FY25 and H1 FY26 Releases) and operating leverage dependent on scale and automation.
The revenue model is overwhelmingly project-based services. Management notes that Appen Global revenue is mostly derived from project-based work with seasonality skewing revenue toward the second half (Source: H1 FY26 Release). Software subscription (ADAP platform licences, historically the Figure Eight business) and off-the-shelf dataset licensing are minor contributors; mix percentages are not publicly disclosed. In Q2 FY26 the company disclosed investment in off-the-shelf datasets to be sold, signalling an intent to grow a licensing line (Source: Q2 FY26 4C).
Value-chain position
Appen sits in the "data layer" of the AI stack: upstream of model training and evaluation, downstream of labour supply (crowd marketplaces) and compute. It does not build foundation models and does not own proprietary data at scale (aside from its off-the-shelf catalogue). Its bargaining position is therefore weak relative to hyperscaler customers, and it competes against venture-funded specialists (Scale AI, Surge AI, Mercor, Turing), BPO/digital-services firms (TELUS Digital, TaskUs), and in-house annotation teams.
Customer types and end-markets
| Customer type | Examples / description | Segment |
|---|---|---|
| US frontier AI labs and Big Tech | Model builders requiring LLM training/evaluation data; Appen does not name current customers, but disclosed that Q2 FY26 included expanding projects with leading AI labs in the USA | Appen Global |
| Chinese LLM builders | Domestic Chinese AI model developers; Appen does not name them | Appen China |
| Enterprise | Corporates building or fine-tuning AI applications (e.g., disclosed case study partner ReflexAI) | Appen Global |
| Government | Historically US/other government programs; management cited policy uncertainty reducing government revenue in 2025 | Appen Global |
| Automotive | Data for ADAS/autonomy perception and in-car voice | Both |
Customer concentration is high. Google alone represented US$82.8m of FY23 revenue before terminating its contract effective March 2024 (Source: Appen ASX release, 22 January 2024, as reported by Business News Australia). In Q2 FY26 management disclosed that Appen Global revenue excluding its largest client grew 65.2% quarter-on-quarter versus 20% including it — implying the largest client's spend declined (Source: H1 FY26 Interim Report / StockWireX summary).
Strategy
Stated strategy themes
Across the FY25 Annual Report, 2026 AGM materials and the H1 FY26 Release, management's strategy can be summarised in the following themes (paraphrased):
-
Revenue growth from global AI demand — capture generative-AI data spending from foundation-model builders in both the US and China.
-
Ongoing underlying EBITDA profitability — maintain positive underlying EBITDA while growing.
-
Data quality as differentiator — position on quality of bespoke datasets and expert crowd specificity (crowd requirements becoming more specialised).
-
Technology-led efficiency — AI-enabled operations, automation and the Mercury/CrowdGen/ADAP platform stack to lower delivery cost.
-
"Both frontiers" positioning — being one of few vendors serving both Western and Chinese AI labs.
-
Organic focus — H1 FY26 call emphasised organic growth and technology-led efficiencies rather than M&A.
The 2023 strategy refresh (under Ahmad) set four pillars that remain relevant: customer diversification, margin improvement through technology, revenue visibility through more recurring products, and strong ROI on GenAI investments.
Initiative register (last 24 months)
| Initiative | Timing | Description / target | Status |
|---|---|---|---|
| Mercury and CrowdGen platform rollout | 2024–2025 | Crowd experience, automation, operational efficiency | Launched 2024 |
| A$65m growth capital raise | Oct–Nov 2024 | Fund working capital and GenAI opportunities | Completed |
| Cost-efficiency program | Identified H1 FY25 | ~US$10m annualised cost-out via technology and automation | ~US$7m complete by Q3 FY25; program implemented in FY25 per FY25 Annual Report summary |
| Pilot-to-production conversion | Q3 FY25 | US$10m+ annual-revenue project moved from pilot to full production | Achieved |
| Agentic AI data and evaluation offering | Nov 2025 | Custom data and end-to-end evaluation for AI agents | Launched |
| Safety research (refusal strategies) | Dec 2025 | Thought leadership on LLM safety evaluation | Published |
| Board renewal | Sep 2025–Jan 2026 | New Chair | Completed |
| Off-the-shelf dataset investment | Q2 FY26 | Build reusable licensable datasets | In progress |
| US frontier-lab expansion | 2,026 | Expand projects with leading US AI labs | In progress (Q2 FY26 commentary) |
| HQ consolidation | Aug 2026 | Move to Sydney CBD | Completed |
| Net Zero Roadmap | Ongoing | Net zero emissions by 2030 | In progress |
Guidance and medium-term targets
| Item | FY26 guidance | Status |
|---|---|---|
| Group revenue | US$270–300m | Set 25 Feb 2026; reaffirmed 30 Apr, 22 May, 29 Jul and 27 Aug 2026 |
| Underlying EBITDA (before FX) margin | ~5–10% | Same |
| Implied H2 FY26 revenue | US$150–180m (vs US$128.7m H2 FY25) | Own calculation |
| Implied FY26 underlying EBITDA | ~US$13.5–30m | Own calculation |
No formal medium-term (multi-year) financial targets are disclosed in reviewed sources. Third-party analyst projections cite revenue of ~US$362m by 2029 (Yahoo Finance-sourced summary) — this is not company guidance.
Products & Services
Appen does not publish price lists; pricing is project-specific. Launch years are given where disclosed.
Platforms (shared across segments)
| Product | Description & capabilities | Target user | Launch / latest version | Pricing model |
|---|---|---|---|---|
| Mercury | Back-end workforce and project-management system used to design projects, manage crowd workers and automate operational workflows; integrated with ADAP and CrowdGen | Internal delivery teams | Launched 2024 (FY25 AGM CEO address) | Internal (embedded in service pricing) |
| CrowdGen | Crowd-facing interface for recruiting, onboarding, qualifying and paying contributors; rebranded from Appen Connect and enhanced in 2024 | Crowd contractors | Enhanced/relaunched 2024 | Internal |
| AI Data Platform (ADAP) | Highly configurable annotation interface enabling internal teams and external customers to design, manage and execute human-in-the-loop data workflows; descended from the Figure Eight platform acquired 2019 | Internal teams; enterprise and AI-lab customers (managed or self-serve) | Figure Eight acquired 2019; ADAP branding current | Platform licence and/or bundled into managed projects — not publicly disclosed |
| AI-assisted annotation tooling | Model-assisted pre-labelling, quality scoring and automation within ADAP/Mercury; central to the stated cost-efficiency program | Internal | Ongoing; emphasised in FY25–H1 FY26 | Internal |
Generative-AI and LLM data services (Appen Global and Appen China)
| Offering | Description | Target customer | Year | Notes |
|---|---|---|---|---|
| Supervised fine-tuning (SFT) data | Expert-written prompts and responses across domains and languages | Foundation-model builders | 2023 onward | Core growth driver in both segments |
| RLHF / preference data | Human ranking and comparison of model outputs | Foundation-model builders | 2023 onward | — |
| AI Chat Feedback | Product for evaluating and improving LLM conversational responses | LLM developers and enterprises | Launched 2023 (Business Wire) | Positioned around helpful, harmless, honest outputs |
| Benchmarking | LLM benchmarking/evaluation product | LLM developers and enterprises | Launched 2023 (Business Wire) | — |
| Model evaluation and red-teaming | Human evaluation of safety, accuracy and bias; adversarial testing | AI labs, enterprises | Ongoing | Appen published research on refusal as a safety strategy (Dec 2025) |
| Agentic AI data and end-to-end evaluation | Custom data and evaluation for AI agents | AI labs, enterprises | Announced November 2025 (appen.com) | — |
| Multimodal data (image/video/audio + text) | Collection and annotation for multimodal models | AI labs | Ongoing | — |
Traditional and legacy services
| Offering | Description | Target customer | Status |
|---|---|---|---|
| Search and ads relevance evaluation | Human raters judging search/ads/feed relevance | Large US platforms | Long-standing strength cited in H1 FY26 Release; materially reduced after Google termination (2024) |
| Speech and language data | Speech collection, transcription, pronunciation lexicons in 500+ languages (self-reported) | Voice assistants, ASR developers, automotive | Long-standing strength cited in H1 FY26 Release |
| Computer vision annotation | Bounding boxes, segmentation, point clouds for ADAS/robotics | Automotive, robotics, retail | Ongoing |
| Location / POI data (ex-Quadrant) | Mobile location and point-of-interest data | Analytics, advertising | Acquired 2021; included in 2022–23 impairments; current status not publicly detailed |
| Synthetic + real data (Mindtech partnership) | Combined synthetic and real image data for computer vision | CV developers | Partnership since 2022; current activity not publicly detailed |
Off-the-shelf (OTS) datasets
| Offering | Description | Status |
|---|---|---|
| OTS speech, language, image datasets | Pre-collected licensable corpora (catalogue historically marketed via datasets.appen.com) | Q2 FY26 4C discloses fresh investment in OTS datasets "to be sold" — a strategic step toward reusable data IP |
Crowd as a product attribute
| Attribute | Figure | Source |
|---|---|---|
| Crowd size | >1 million contractors (self-reported) | FY25 Annual Report |
| Languages | >500 (self-reported) | FY25 Annual Report / H1 FY26 Release |
| Countries | >200 including territories (self-reported) | FY25 Annual Report |
| Domain specialisations | >100 | FY24 Annual Report |
Financial Narrative
Income statement — Appen-reported measures
Sources and caveats:
-
FY21 revenue: US$447.3m (Reuters, Oct 2022). FY22: US$388.5m (FY22 results call); some materials cite US$388.8m.
-
FY21 underlying EBITDA before FX: implied as US$79.1m from FY22's reported 82.8% decline to US$13.6m; Reuters cited US$78.9m — discrepancy of US$0.2m noted.
-
FY23 underlying EBITDA before FX (US$20.4m loss) from FY24 Results Release comparative table.
-
FY23 underlying NPAT (US$52.8m loss) derived from FY24's reported US$42.3m improvement to US$(10.5)m.
-
FY23 statutory NPAT derived from FY24's reported US$98.1m improvement to US$(20.0)m (S&P shows -118.08).
-
FY25 statutory loss US$21.818m (Appendix 4E, 25 Feb 2026).
Income statement — S&P standardised: notes and sources
Source: S&P via StockAnalysis (pre-tax income computed from reported pre-tax margins × revenue). FY21 diluted EPS shown equal to basic as diluted figure was not separately verified.
Dividends
Source: S&P via StockAnalysis (FY21, USD-equivalent of A$-declared dividends). The interim FY22 dividend was scrapped (August 2022); no dividends have been declared since, and no final dividend was declared for FY25 (Appendix 4E).
Balance sheet
Source: S&P via StockAnalysis. Appen has no bank borrowings; "total debt" consists of lease liabilities. Net tangible assets per share fell from 32.27 US cents (FY24) to 29.67 US cents (FY25) (Appendix 4E).
30 June 2026 balance sheet (H1 FY26): cash US$44.71m; total debt (leases) US$7.43m; net cash US$37.28m; receivables US$72.08m; accounts payable US$35.67m (up sharply from US$12.93m); equity US$93.01m; shares 268.52m (S&P via StockAnalysis; H1 FY26 Release confirms cash US$44.7m and no debt).
Cash flow
Source: S&P via StockAnalysis. Caveat: Appen historically capitalised significant product-development costs (presented within investing activities as intangibles); S&P's capex line may not capture these, so true FCF in FY21–FY23 was likely lower than shown.
Equity issued (not shown in chartable form): FY23 ~A$90m gross (A$60m in May at A$1.85; A$30m in Nov/Dec at A$0.55); FY24 ~A$65m gross (A$50m placement and A$15m SPP at A$1.92); FY25 nil (Sources: ASX equity raising announcements, 17 May 2023, 9 June 2023, Nov 2023, 14 Oct 2024, 7 Nov 2024). Shares on issue more than doubled from 123m (FY22) to 269m (H1 FY26).
H1 FY26 cash flow: operating cash flow was negative in H1 FY26 (investing.com/de summary); cash fell US$14.3m in Q2 FY26 alone, from US$59.0m to US$44.7m, which management attributed partly to working capital in Appen China and investment in off-the-shelf datasets (Q2 FY26 4C). TTM operating cash flow to June 2026: US$5.67m; TTM FCF US$0.43m (S&P).
Ratios
Sources: S&P via StockAnalysis for ROE, ROA, ROIC, current ratio, D/E and asset turnover; DSO = receivables ÷ S&P revenue × 365 (own calculation).
Ratios not meaningful or not disclosed: Net debt/EBITDA — not meaningful (net cash every year). Interest coverage — not meaningful (no borrowings; only lease interest). Cash conversion cycle — not computable in the conventional sense (negligible inventory; crowd payables not separately disclosed); DSO used as the working-capital proxy.
Commentary on trends, inflections and drivers
FY21→FY23: the collapse. Revenue fell 39% over two years and underlying EBITDA swung from ~US$79m to a US$20m loss. Drivers per company disclosures and contemporaneous reporting: weaker digital advertising demand among Big Tech customers (ads-relevance work), Apple's privacy changes reducing targeting-related projects, customer in-sourcing and pricing pressure, and a cost base swollen by investment in China, Quadrant and product transformation. Two impairments followed — US$204.3m in FY22 (New Markets ex-China CGU) and further write-downs in FY23 which eliminated remaining goodwill (S&P balance sheet shows goodwill falling from US$53.1m to nil). Equity fell from US$392m to US$93m.
FY24: survival and reset. Google's termination removed roughly US$83m of annual revenue, yet underlying EBITDA turned positive (US$3.5m) because operating expenses were cut ~26% and new generative-AI work grew; GenAI projects were 28% of H2 FY24 revenue versus 6% in H2 FY23 (FY24 results call). The A$65m 2024 capital raise rebuilt cash to US$54.8m.
FY25: mix shift to China. Group revenue was flat but composition changed dramatically: China +75% to US$102.9m with EBITDA up 640%; Global -21% ex-Google. Gross margin improved 100bps to 40.3% on GenAI mix. Operating cash flow recovered to US$23.0m — management cited roughly 100% conversion of EBITDA to operating cash flow after adjustments (FY25 call). The statutory loss widened slightly to US$21.8m because of D&A, share-based payments and restructuring items outside "underlying".
H1 FY26: growth with cash strain. Revenue +17.5% and underlying EBITDA +US$7.5m to US$5.3m, but operating cash flow turned negative and cash fell to US$44.7m as China's rapid growth absorbed working capital (receivables +US$15m since December) and OTS dataset investment. Payables also rose sharply (US$35.7m vs US$12.9m), which partly funded growth — a reversal would pressure cash further.
Key inflection to monitor: Appen China's margin structure is inherently lower than Appen Global's peak margins (management states China margins are traditionally lower), so a China-heavy mix caps group margin unless Global volumes recover.
Unlock Company Financials
Unlock this company profile to view its financial data.
View Subscription PlansFinancial Detail
Segment Overview
| Segment | Contents | Products/services | Customers |
|---|---|---|---|
Appen Global | All operations outside the China business, including US Big Tech and frontier-lab programs, enterprise, government, global product (ADAP) | LLM data (SFT, RLHF, evaluation, red-teaming), search/ads relevance, speech and language data, multimodal collection, ADAP platform, off-the-shelf datasets | US and international technology companies, AI labs, enterprises, governments |
Appen China | Appen's China-based operations serving domestic customers | LLM training and evaluation data, speech, multimodal data collection and annotation | Chinese AI model builders and technology companies |
Corporate / unallocated | Group costs and minor other income | — | — |
Revenue by Segment
| Segment | FY2024 | FY2025 |
|---|---|---|
Appen China revenue (USD Mn) | 58.9 | 102.9 |
Appen Global revenue incl Google (USD Mn) | 175.4 | 127.9 |
Appen Global revenue ex Google (USD Mn) | 160.2 | 127.9 |
Group operating revenue (USD Mn) | 234.3 | 230.8 |
Unlock Company Financials
Unlock this company profile to view its financial data.
View Subscription PlansResults by Segment
| Segment | FY2024 | FY2025 |
|---|---|---|
Appen China underlying EBITDA (USD Mn) | 1.4 | 10.6 |
Appen Global underlying EBITDA (USD Mn) | 9.2 | 5.8 |
Corporate and other (derived, USD Mn) | -7.1 | -4.2 |
Group underlying EBITDA before FX (USD Mn) | 3.5 | 12.2 |
Segment Margins & Growth (%) and Revenue by Segment (%)
| Metric | FY2024 | FY2025 |
|---|---|---|
Appen China EBITDA margin (%) | 2.4 | 10.3 |
Appen Global EBITDA margin (%) | 5.2 | 4.5 |
Appen China share of group revenue (%) | 25.1 | 44.6 |
Appen Global share of group revenue (%) | 74.9 | 55.4 |
Revenue by Segment (Half-Yearly)
| Segment metric | H1 FY2025 | H1 FY2026 |
|---|---|---|
Appen China revenue (USD Mn) | 42.2 | 76.2 |
Appen Global revenue (USD Mn) | 59.8 | 43.7 |
Group revenue (USD Mn) | 102.1 | 119.9 |
Appen China underlying EBITDA (USD Mn) | 2.9 | 12.1 |
Group underlying EBITDA before FX (USD Mn) | -2.2 | 5.3 |
Revenue by Segment (FY2022–FY2024)
| Legacy segment | FY2022 | FY2024 |
|---|---|---|
Global Services revenue (USD Mn) | 299.8 | 118.1 |
New Markets revenue (USD Mn) | 88.4 | 116.2 |
Revenue by Segment (Quarterly)
| Quarter metric | Q1FY25 | Q2FY25 | Q3FY25 | Q4FY25 | Q1FY26 | Q2FY26 |
|---|---|---|---|---|---|---|
Group revenue (USD Mn) | 50.3 | 51.8 | 55.3 | 73.4 | 54.8 | 65.1 |
Appen China revenue (USD Mn) | 18.6 | 23.6 | 28.7 | 32.0 | 34.9 | 41.3 |
Appen Global revenue (USD Mn) | 31.6 | 28.2 | 26.6 | 41.4 | 19.9 | 23.8 |
Group underlying EBITDA before FX (USD Mn) | -1.5 | -0.7 | 1.0 | 13.4 | 1.0 | 4.3 |
Income Statement and Margin Analysis (%)
| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|---|
Operating revenue (USD Mn) | 447.3 | 388.5 | 273.0 | 234.3 | 230.8 |
Underlying EBITDA before FX (USD Mn) | 79.1 | 13.6 | -20.4 | 3.5 | 12.2 |
Underlying NPAT (USD Mn) | 40.6 | -22.8 | -52.8 | -10.5 | -21.8 |
Statutory NPAT (USD Mn) | 28.5 | -239.1 | -118.1 | -20.0 | -21.8 |
Underlying EBITDA margin before FX (%) | 17.7 | 3.5 | -7.5 | 1.5 | 5.3 |
Statutory net margin (%) | 6.4 | -61.5 | -43.3 | -8.5 | -9.5 |
Unlock Company Financials
Unlock this company profile to view its financial data.
View Subscription PlansIncome Statement and Per-Share Data
| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|---|
Revenue incl other income (USD Mn) | 447.26 | 388.31 | 273.79 | 235.22 | 232.67 |
Gross profit S&P basis (USD Mn) | 107.87 | 56.38 | 22.17 | 37.86 | 44.69 |
Operating income (USD Mn) | 43.80 | -28.31 | -55.56 | -19.85 | -19.15 |
Pre-tax income (USD Mn) | 35.87 | -239.08 | -124.96 | -19.99 | -21.78 |
Net income (USD Mn) | 28.52 | -239.07 | -118.08 | -20.01 | -21.82 |
EPS basic (USD) | 0.23 | -1.94 | -0.83 | -0.09 | -0.08 |
EPS diluted (USD) | 0.23 | -1.94 | -0.83 | -0.09 | -0.08 |
Gross margin S&P basis (%) | 24.12 | 14.52 | 8.10 | 16.10 | 19.21 |
Operating margin (%) | 9.79 | -7.29 | -20.29 | -8.44 | -8.23 |
Financial Analysis
| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|---|
Dividend per share USD equivalent | 0.073 | 0 | 0 | 0 | 0 |
CAGR by Period
| Period | CAGR (%) | Basis |
|---|---|---|
FY21–FY25 (4 years) | -15.2 | Operating revenue US$447.3m → US$230.8m (own calculation) |
FY23–FY25 (2 years) | -8.1 | US$273.0m → US$230.8m |
FY24–FY25 | -1.5 | Reported; +4.5% ex-Google |
FY25–FY26 guidance midpoint | 23.5 | US$230.8m → US$285m midpoint of US$270–300m |
Balance Sheet and Share Count (M)
| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|---|
Total assets (USD Mn) | 502.16 | 254.94 | 155.16 | 170.59 | 152.47 |
Cash and equivalents (USD Mn) | 47.88 | 23.43 | 32.15 | 54.81 | 59.82 |
Total debt incl leases (USD Mn) | 15.06 | 10.18 | 12.43 | 11.04 | 9.40 |
Short-term lease liabilities (USD Mn) | 5.00 | 3.15 | 3.13 | 3.58 | 4.15 |
Long-term lease liabilities (USD Mn) | 10.06 | 7.03 | 9.31 | 7.46 | 5.25 |
Net cash (USD Mn) | 32.82 | 13.25 | 19.72 | 43.77 | 50.42 |
Shareholders equity (USD Mn) | 391.87 | 147.99 | 92.80 | 114.32 | 94.60 |
Goodwill (USD Mn) | 241.82 | 53.11 | 0 | 0 | 0 |
Other intangibles (USD Mn) | 77.01 | 56.45 | 39.87 | 30.18 | 15.81 |
Working capital (USD Mn) | 97.37 | 59.66 | 59.10 | 85.31 | 75.31 |
Accounts receivable (USD Mn) | 97.78 | 91.57 | 63.25 | 65.08 | 52.70 |
Shares outstanding (M) | 123.07 | 123.45 | 211.47 | 260.74 | 265.61 |
Cash Flow
| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|---|
Operating cash flow (USD Mn) | 53.92 | 18.23 | -22.94 | -0.66 | 22.97 |
Capex PP&E per S&P (USD Mn) | -1.30 | -3.04 | -1.81 | -1.79 | -4.03 |
Free cash flow per S&P (USD Mn) | 52.62 | 15.19 | -24.75 | -2.45 | 18.94 |
Share buybacks (USD Mn) | 0 | 0 | 0 | 0 | 0 |
Returns & Leverage Ratios (%) and Financial Ratios (x)
| Ratio | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|---|
ROE (%) | 7.45 | -88.57 | -98.08 | -19.32 | -20.89 |
ROA (%) | 5.68 | -4.67 | -16.93 | -7.61 | -7.41 |
ROIC (%) | 10.07 | -11.47 | -53.47 | -27.64 | -33.38 |
Current ratio (x) | 2.51 | 1.92 | 2.24 | 2.95 | 2.59 |
Debt to equity (x) | 0.04 | 0.07 | 0.13 | 0.10 | 0.10 |
Asset turnover (x) | 0.93 | 1.03 | 1.34 | 1.44 | 1.44 |
Days sales outstanding own calc (days) | 79.8 | 86.1 | 84.3 | 101.0 | 82.7 |
Revenue by Region
| Region proxy | FY2024 | FY2025 |
|---|---|---|
China domestic via Appen China (USD Mn) | 58.9 | 102.9 |
Rest of world via Appen Global (USD Mn) | 175.4 | 127.9 |
Geographic Revenue
| Region proxy | FY2025 growth (%) | H1 FY2026 growth (%) |
|---|---|---|
China | 74.8 | 80.4 |
Rest of world (incl Google in base) | -27.1 | -26.9 |
Rest of world (ex Google base) | -21.1 | not reported separately |
Revenue by Region (%)
| Metric | FY2024 | FY2025 | TTM to June 2026 |
|---|---|---|---|
China share of group revenue (%) | 25.1 | 44.6 | 55.6 |
Share Price and Margin Analysis (%)
| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|---|
Year-end share price (AUD) | 9.79 | 2.20 | 0.63 | 2.64 | 0.80 |
Calendar-year return (%) | -54.5 | -77.5 | -71.4 | 319.0 | -69.7 |
Year-end market cap (AUD Mn) | 1,374 | 307 | 133 | 688 | 212 |
Share Price & Returns
| Horizon | Return | Basis |
|---|---|---|
YTD 2026 | +52.5% | From A$0.80 (31 Dec 2025) |
1-year | +23.4% | To 4 Sep 2026 (StockAnalysis) |
Since 31 Dec 2023 (~2.75 years) | +93.7% | From A$0.63 |
Since 31 Dec 2021 (~4.75 years) | -87.5% | From A$9.79 |
From Aug 2020 peak (>A$40) | ~-97% | Approximate |
52-week range | A$0.65 – A$1.975 | StockAnalysis (Sep 2026) |
Beta | 1.78 | StockAnalysis |
Valuation Snapshot
| Multiple | Value | Basis |
|---|---|---|
Market cap | A$327.6m / US$225.8m | 268.52m shares |
Enterprise value | ~US$188.5m | Market cap less US$37.3m net cash (June 2026, incl. leases) |
EV / TTM revenue | ~0.76x | TTM operating revenue US$248.6m (H2 FY25 128.7 + H1 FY26 119.9) |
EV / TTM underlying EBITDA | ~9.6x | TTM US$19.7m (H2 FY25 14.4 + H1 FY26 5.3) |
EV / FY26E underlying EBITDA (guidance midpoint) | ~8.8x | US$285m × 7.5% = US$21.4m |
EV / FY26E EBITDA range | ~6.3x – 14.0x | High end US$30.0m; low end US$13.5m |
Forward P/E | ~27–29x | StockAnalysis (Sep 2026), consensus-based |
P/B | ~2.4x | Equity US$93.0m (June 2026) |
P/TBV | ~3.0x | Tangible book US$75.4m |
Source Details
| Source | Coverage | Rating | Target |
|---|---|---|---|
StockAnalysis (Sep 2026) | 3 analysts | Buy | A$1.90 average (~+56% vs A$1.22) |
Motley Fool (Jul 2026) | Not stated | Mostly bullish | Lowest target A$1.96 |
Simply Wall St | 9 analysts cover; 4 contribute estimates | — | — |
Parqet | 1 Buy, 0 Hold, 0 Sell | Buy | — |
Capital Markets
| Item | Detail |
|---|---|
Dividend policy | No dividends since FY21; Board prioritises capital for growth (Appendix 4E FY25) |
Last dividend | FY21 (interim FY22 dividend scrapped Aug 2022) |
Credit ratings (Moody's / S&P / Fitch) | Not rated |
Debt maturity profile | No borrowings; lease liabilities of US$7.4m (June 2026), ~US$3.8m current |
Analyst Conclusions
Management guidance
FY26 revenue of US$270–300m with an underlying EBITDA (before FX) margin of ~5–10%, reaffirmed on 27 August 2026. Achieving even the low end requires H2 FY26 revenue of ~US$150m (+17% vs H2 FY25), heavily dependent on Appen Global's seasonal second-half projects; the China run-rate (>US$175m annualised at June 2026) alone implies ~US$88m of H2 China revenue if sustained.
Consensus expectations
Forward P/E of ~27–29x (StockAnalysis) implies consensus expects a return to modest statutory profitability in FY27. Third-party long-range projections cite ~US$362m revenue by 2029.
Bull case
-
China compounding at high margins. China revenue grew 80% in H1 FY26 with EBITDA margin rising to ~16%; if the >US$175m run-rate holds, China alone delivers ~US$25–30m annual EBITDA, more than the group's entire FY25 underlying EBITDA.
-
Global operating leverage on volume recovery. Q4 FY25 showed Appen Global generating US$10.2m EBITDA on US$41.4m revenue in one quarter; a repeat in Q4 FY26 plus incremental US frontier-lab work would push the group toward the top of guidance.
-
Valuation asymmetry. At ~0.76x EV/sales and ~9x forward EBITDA (guidance midpoint) with net cash, Appen trades at a fraction of private peers' multiples; prior approaches (TELUS 2022, Innodata 2023) demonstrate strategic interest.
Bear case
-
Appen Global may be in structural decline. Revenue fell 21% ex-Google in FY25 and 27% in H1 FY26, and the largest remaining Global client shrank in Q2 FY26, while expert-data specialists with billion-dollar revenues capture frontier-lab budgets.
-
China concentration is a geopolitical single point of failure. With China >50% of TTM revenue, escalation of US–China AI controls or Chinese data-security enforcement could impair the growth engine or its perceived value to Western customers.
-
Cash and dilution. Cash fell from US$59.8m to US$44.7m in six months during a growth phase; a guidance miss could force another equity raise, extending a history of dilution (shares up ~118% since FY22) and repeated guidance shortfalls (FY25 landed below its range).
Catalysts and monitorables (next 12 months)
| Timing | Catalyst / monitorable |
|---|---|
| Late Oct 2026 | Q3 FY26 4C: China run-rate (watch >US$175m), Global QoQ trend, cash balance |
| Q4 2026 | Evidence of Global seasonal surge (Q4 FY25 comparison: US$41.4m) |
| Late Jan 2027 | Q4 FY26 update — guidance delivery |
| 24 Feb 2027 (expected) | FY26 results; FY27 guidance; statutory profitability trajectory |
| Ongoing | US–China AI policy actions |
| Ongoing | Named customer wins or large-lab disclosures; OTS dataset revenue |
| May 2027 | AGM — remuneration structure after two strikes |
| Ongoing | Working-capital normalisation in China (receivables vs payables) |
Concluding analyst verdict
Appen is no longer the company that peaked above A$40 in 2020; it is a smaller, two-speed business whose value now rests disproportionately on a Chinese growth engine. The evidence of turnaround is real but uneven. On the positive side, Appen China has delivered six-plus quarters of profitable high growth, reaching an annualised run-rate above US$175m and a mid-teens EBITDA margin; group underlying EBITDA has moved from a US$20.4m loss in FY23 to US$12.2m in FY25 and US$5.3m in a seasonally weak H1 FY26; and the balance sheet carries net cash with no borrowings. On the negative side, Appen Global — once the entire franchise — continues to shrink against far larger, better-capitalised US expert-data competitors, statutory losses persist, cash is being absorbed by growth, and governance confidence is strained, as shown by consecutive remuneration strikes.
At roughly 0.76x EV/sales and ~9x guidance-midpoint EBITDA, the market is pricing a substantial probability that FY26 guidance is missed or that the China business is ultimately discounted for geopolitical risk. That pricing is not unreasonable given Appen's record of falling short of guidance. However, the asymmetry has improved: if H2 FY26 resembles H2 FY25's seasonal strength and China holds its run-rate, Appen should land within its US$270–300m guidance and approach statutory breakeven, which would likely re-rate the shares toward analyst targets near A$1.90.
Our synthesis: a high-risk, catalyst-driven special situation rather than a quality compounder. The investment case hinges on three observable variables — Appen Global's Q4 seasonal volume, China's run-rate and working-capital discipline, and cash preservation without new equity. Investors should treat the October 2026 and January 2027 quarterly updates as decisive checkpoints, size positions for binary outcomes, and assign an explicit geopolitical discount to the China contribution.
Unlock Company Financials
Unlock this company profile to view its financial data.
View Subscription PlansExecutive Leadership
Executive Leadership Team
| Name | Title | Tenure | Background | Education |
|---|---|---|---|---|
Ryan Kolln | CEO & Managing Director | CEO since 5 Feb 2024; joined Appen 2018 | SVP Strategy & Innovation (2023), then Chief Operating Officer (2023–Feb 2024); North America-based | MBA; B.Eng (Electrical) (Simply Wall St profile) |
Justin Miles | Chief Financial Officer | — | — | Not publicly disclosed in reviewed sources |
Board of Directors
| Name | Role | Board tenure | Background highlights | Shareholding |
|---|---|---|---|---|
Vanessa Liu | Independent Non-Executive Chair | Director since 2020; Chair from 1 Jan 2026 | 25+ years in global technology; former VP of SAP.io (SAP's early-stage venture arm); senior roles at Trigger Media Group; former McKinsey associate partner; co-founder and CEO of Sugarwork | 25,200 shares (30 Dec 2025) |
Ryan Kolln | CEO & Managing Director | Since Feb 2024 | See above | 498,782 shares (5 Jun 2026) plus performance rights |
Stephen Hasker | Independent NED | — | Senior global media/data executive | 58,333 shares (30 Dec 2025) |
Lynn Mickleburgh | Independent NED | — | — | 300,000 shares (30 Dec 2025) |
Robin Low | Independent NED | — | — | 272,676 (indirect, 30 Dec 2025) |
Stuart Davis | Independent NED | — | — | 88,455 (indirect, 30 Dec 2025) |
Directors' Remuneration
| Component | FY2025 (USD) | Source |
|---|---|---|
Total compensation | ~968,130 | Simply Wall St, based on FY25 remuneration report |
Mix | ~62% salary, ~38% bonus/equity | Simply Wall St |
Directors' Remuneration by AGM
| AGM | Remuneration report vote against (%) | Outcome |
|---|---|---|
16 May 2025 | ~27.7 (9.26m against vs 24.13m for, own calculation from ASX results) | First strike (>25%) |
22 May 2026 | 42.19 | Second strike; spill resolution put and defeated with 92.53% against |
Board & Management Changes
| Date | Change | Context |
|---|---|---|
Dec 2022 | Mark Brayan resigns as CEO; Armughan Ahmad appointed | Following failed TELUS approach and missed targets |
5 Feb 2024 | Ahmad departs; Ryan Kolln appointed CEO | Two weeks after Google termination |
24 Sep 2025 | Chair succession announced | Freudenstein reducing board load (Coles, REA, Cochlear) |
1 Jan 2026 | Vanessa Liu becomes Chair | US-based technology executive leads board |
Competitive Landscape
Key Competitors
| Competitor | Segment overlap | Type | Scale indicator | Positioning vs Appen |
|---|---|---|---|---|
Scale AI (US, private; Meta invested ~US$14bn in 2025) | Global | VC-backed data/infra platform | ~US$850m ARR (2024, media) to ~US$1.5bn ARR end-2024 (Sacra) — sources conflict | Historically market leader; some labs reportedly cut ties after Meta deal, opening share for rivals |
Surge AI (US, private) | Global | Premium expert RLHF data |
| Quality leader serving ~12 frontier labs; far larger than Appen Global |
Mercor (US, private) | Global | Expert marketplace for AI training | ~US$2bn annualised gross revenue by mid-2026 (CEO/Sacra) | Fastest-growing; gross revenue includes contractor pass-through |
Turing (US, private) | Global | Expert coding/RL data | — | Competes in coding/agentic data |
micro1 (US, private) | Global | AI-recruited expert data | ~US$500m gross run-rate (reported) | Emerging challenger |
TELUS Digital (Canada; AI Data Solutions) | Global | Digital services/BPO with AI data unit | — | Former bidder for Appen; deep crowd |
TaskUs (US) | Global | BPO with AI services | — | Trust & safety and AI ops |
Innodata (US, NASDAQ: INOD) | Global | Listed AI data engineering | — | Closest listed pure-play comparable; made 2023 approach to Appen |
Sama, iMerit, CloudFactory, Labelbox, Toloka, Defined.ai | Global | Annotation services/tools | — | Niche specialists |
Hyperscaler in-house tools (AWS SageMaker Ground Truth, Google) | Global | Self-serve labelling | Not applicable | Substitution risk |
Haitian Ruisheng / Speechocean (SSE STAR Market) | China | Listed speech/language data | — | Direct speech data competitor in China |
DataTang, Testin, Magic Data, Baidu Smart Cloud data services | China | Domestic data vendors | — | Domestic competition; policy-favoured locals |
Peer Comparison FY2025
| Company | Revenue | Growth | EBITDA margin | Basis |
|---|---|---|---|---|
Appen | 230.8 | +4.5% ex-Google (FY25); +17.5% (H1 FY26) | 5.3% underlying (FY25) | FY25 audited / H1 FY26 |
Surge AI |
| — | — | 2024, media/Sacra estimates |
Scale AI | ~850–1,500 ARR | — | — | 2024, conflicting estimates |
Mercor | ~2,000 annualised gross | ~2.6x vs end-2025 run-rate | — | Mid-2026, CEO/Sacra |
“—” = not publicly disclosed.
Recent Developments
--
