21 days ago
News

Copy Geek raises $2M pre-seed, no sales team disclosed

## Copy Geek raises $2M pre-seed, no sales team disclosed Sydney crypto trading software startup Copy Geek raised A$2 million in pre-seed funding from private investors led by Dr David Issa and Thomas Boga. One funding database lists the round as US$1.4 million, likely a currency conversion rather than separate financing. Founded in 2026 by Chris Ji, Copy Geek builds non-custodial copy-trading infrastructure for cryptocurrency exchanges. The platform is live on Bitfinex, WhiteBIT, Bybit and Hyperliquid. Users mirror verified traders through their own exchange accounts. The company charges no trading fee, taking an approved performance fee instead. ## The sales angle Copy Geek's go-to-market is B2B2C: win exchange integrations, not end users directly. The pitch to exchanges: attract funded, active traders without bearing full acquisition cost. Capital goes to product development, engineering hires and deeper exchange partnerships. What is missing: employee count, sales team size, named CRO or VP Sales, customer numbers, revenue, or broader ANZ headcount. For a pre-seed SaaS play, that tracks. Worth noting if you are evaluating sales roles in crypto infrastructure: this is early, product-led, partnership-focused. ## Market context Copy Geek sits in the crypto copy-trading and exchange infrastructure market alongside eToro, Bitget and Binance Copy Trading. The funded accounts and prop trading model is gaining traction in crypto, similar to forex prop firms that fund traders in exchange for performance splits. Ji's background is derivatives and product development. Tech adviser Adrian Tadros has two decades across cloud providers and financial institutions. The consortium backing them is Sydney-based private capital, not institutional venture. ## What it means If you are watching the crypto sales market: this is infrastructure, not consumer. Sales motion will be partnership-heavy, likely enterprise AE roles targeting exchanges, not SDR-led volume. Comp structure unclear. Early-stage risk, technical product, niche market. Pre-seed means no established sales org yet. If they scale, expect partnership and enterprise AE hiring in 12 to 18 months. Until then, this is a product and engineering story.

21 days ago
News

Four ANZ startups raise $31m: no sales team data disclosed

Four Australian startups raised A$31.9 million this week across energy storage, legal tech, preventative health and construction software. The rounds: MGA Thermal ($17m), Mary Technology ($7m), TMRW ($7m seed) and Scopey Onsite ($851k pre-seed). **MGA Thermal** is building thermal-energy storage for heavy industry. The $17 million supports commercial scale-up. Sales opportunity sits with industrial energy users and utilities, not traditional SaaS buyers. No revenue, headcount or sales team structure disclosed. Leadership: Mark Croudace and Shane Meaney. **Mary Technology**, a Sydney legal-tech company, closed $7 million to fund US expansion. The litigation-data platform targets law firms, making international go-to-market execution critical. Founded by Rowan McNamee, Luke Abagi, Harry Raworth and Daniel Lord-Doyle. No CRO, VP Sales, revenue or sales headcount data available. Competitive set includes legal research, litigation analytics and e-discovery platforms. **TMRW** raised $7 million seed for its preventative-health platform. CEO Mark Britt is expanding clinical footprint. Market position: longevity services meets digital health. No customer count, revenue or sales hiring disclosed. Competes with clinical longevity providers and health-screening platforms. **Scopey Onsite** closed €523k (A$851k) pre-seed for AI-powered construction data capture. Co-founder Gillian Laging (COO) is accelerating product development. Target market: ANZ construction contractors. Early commercial stage; no public revenue or sales team data. Competitive set: construction field-management and site-reporting software. **Worth noting:** The combined $31.9 million included zero public disclosure of sales team size, quota structures, CRO appointments or go-to-market strategy. For early-stage startups, that tracks. For a company announcing US expansion (Mary Technology), silence on VP Sales or revenue metrics raises questions about execution readiness. Mary Technology's legal-tech play is the most relevant to traditional B2B sales motion: targeting law firms requires enterprise sales capability. Whether they are building that team or relying on founder-led sales remains unclear. MGA Thermal and TMRW face complex, consultative sales cycles (industrial procurement and clinical partnerships respectively). Scopey Onsite is pre-revenue. None disclosed hiring plans for AEs, SDRs or sales leadership.

21 days ago
News

Guidewire pays 21x ARR for AI startup, double its own 10x multiple

## The Numbers Guidewire reported $1.24 billion ARR for fiscal 2026, up 19% constant currency. Total revenue hit $1.48 billion, up 23%. Subscription revenue grew 37% to $916 million while license revenue fell 7%. The company ended the year with 105 customers over $5 million in fully ramped ARR, up from 86 a year earlier. ARR gross attrition hit its lowest rate since Guidewire started measuring the metric. That matters when your average deal is a multi-year core system replacement, not a transactional SaaS contract. Operating cash flow margin was 26%. ## The Acquisition Math On 27 October, Guidewire agreed to pay $1.7 billion upfront for hyperexponential, an AI-powered underwriting and pricing platform, plus up to $400 million in contingent consideration tied to revenue targets. Guidewire did not disclose hyperexponential's ARR. Management said the deal should add at least 650 basis points to fiscal 2027 ARR growth. Applied to Guidewire's $1.24 billion base, that implies roughly $80 million of acquired ARR. On that figure, the acquisition multiple is about 21x ARR on the base price, potentially 26x if the full earnout pays. Guidewire itself trades at roughly 10.8x current ARR. A company the market prices as legacy just paid double that multiple for AI-native growth in its own vertical. ## What Changed Earlier this year, CEO Mike Rosenbaum said Guidewire would own core systems and partner for AI. Hyperexponential was one of those partners. Guidewire also already sells its own pricing product, PricingCenter, which closed 12 deals in fiscal 2026. Management decided buying the leader was faster than building for complex commercial risk. The balance sheet shifts. Guidewire plans to fund the deal with $1 billion in new term loans plus cash. After the acquisition, debt rises from roughly $704 million to about $1.7 billion. Over 15 months, Guidewire committed roughly $2.3 billion to buybacks and this acquisition. Fiscal 2026 operating cash flow was $390 million. ## The Sales Angle Guidewire reported $336 million of future ARR already signed but not yet recognised, a 27% cushion on current ARR. That gap grew from $255 million a year earlier. More than half of net new ARR in the fiscal 2027 guide is already contracted. The company sells multi-year transformation projects to some of the slowest-moving buyers in enterprise software. Low churn and a growing base of customers over $5 million ARR support the system-of-record positioning. The acquisition accelerates growth but also resets the comp story from organic execution to inorganic ARR and integration risk. Management updates fiscal 2027 guidance in December. That update will separate organic from acquired growth, and the market will price the organic number.

22 days ago
News

SaaStr cut sales team to 1.5 people, runs GTM with 20 AI agents

# SaaStr Cut Sales Team to 1.5 People, Runs GTM With 20 AI Agents Jason Lemkin's SaaStr now operates with three humans and 20 AI agents in production. The human sales team is down to 1.5 people. Revenue is higher than it was with the bigger team. The data: SaaStr's AI agents generated $4.8 million in additional pipeline, with approximately half closed. Deal volume and win rates doubled after deploying agents across GTM. The agents send 3,200 emails per month, against the 75 to 285 a human SDR sent at the same company size. "Agent time for each of us went from about 30 minutes a day a year ago to roughly 8 hours a day," Lemkin said. The company uses Salesforce Agentforce, Artisan, Qualified, Clay, Momentum, Gamma and Zapier, with Salesforce as the central CRM and agent hub. SaaStr's Chief AI Officer Amelia LeRutte moved from managing a 10-person team to overseeing 20 AI agents. The company's AI VP of Marketing now also runs quote-to-cash and commissions. When a contract is signed, the agent flips the deal to Closed Won, appends contacts, creates and sends invoices, and runs collections reminders. ## What Changed Across GTM Vercel took a 10-person SDR function to 1.25 people for $5,000 per year in infrastructure costs. The company moved the 10 people into higher-value roles instead of cutting them. Vercel's support agent now handles 93% of case load on deep infrastructure tickets, not password resets. Kyle Norton, CRO at restaurant-management software company Owner, deployed 20 agents and eliminated SDR roles. Anthropic reported 54% of new enterprise logos closed self-serve after rebuilding its sales organisation. In Emergence's survey of 560-plus companies, 36% had cut SDRs. Only 14% had cut sales engineers, and 28% had grown AE headcount. ## What the CRO Still Owns The CRO role survived because someone has to decide which motions go to agents, which leads go to which human, and when an agent's output is good enough to trust. At SaaStr, those decisions are Lemkin's and LeRutte's. The reps themselves are doing the same job about as well as they did two years ago. The gains came from decisions made above them: which accounts get nothing, where to concentrate the best opportunities, and how to measure agent performance against human benchmarks. Nobody has built a good AI account executive. Agents run the top and bottom of the funnel. The middle, and the person who owns it, is still human. Marketing ops, RevOps and finance ops are collapsing into a few agents, but the revenue number still needs one owner. For ANZ sales professionals, the implication is direct: SDR, support, CS and ops roles are moving to software. The jobs that survived are the ones that own revenue decisions, not execution. If your role is defined by volume, start looking at what owns the architecture above it.

22 days ago
News

$300 billion PE software debt hits 2028: what vendors your customers use

## The Numbers About $300 billion of PE-backed corporate debt matures from 2028 onward. Software accounts for roughly one-third of the riskiest loans due in 2028–29. That is $100 billion against B2B companies bought at peak multiples, now growing slower than their debt assumed. The math is straightforward: a company bought for $800 million on $320 million of debt could be worth $240–360 million if its revenue multiple falls to 2–3x. Lenders get paid first. Equity holders get what is left. Software loans increased from $8 billion in 2015 to more than $500 billion by end-2025, representing 19% of direct lending. About 21% of outstanding software loans mature in 2028, compared with 14% across all sectors. ## Why This Matters for Sales Teams Your customers rely on vendors that may not exist in 18 months. If a customer success platform, sales engagement tool, or revenue operations system goes into workout, support degrades before the default becomes public. The risk concentrates in mid-market PE-owned vendors: high leverage, slowing growth, limited refinancing options. Large public SaaS companies have recurring revenue and cash reserves to reduce spending before missing payments. Smaller PE-backed tools do not. Valuation distress is not the same as credit distress. Software stocks lost $1.2 trillion in market value between December 2025 and June 2026. Roughly $25 billion of software leveraged loans trade below 80 cents on the dollar. But public one-year default probability sits at 12 basis points, despite doubling from December. ## What Changes in 2027 Refinancing decisions happen in 2027 for 2028 maturities. Lenders will evaluate growth rates, customer retention, and whether AI-native competitors are taking share. Vendors that cannot refinance enter workout or sale processes. For sales teams: track which tools your customers use, who owns them, and when debt matures. If a core system goes into distress, migration risk becomes a sales conversation. Stability becomes a competitive advantage. The defaults have not started yet. Refinancing season determines who survives.

22 days ago
News

Databricks CRO: One SDR Books 400 Meetings Monthly, Reply Rates Hit 10%

## Databricks kept SDR headcount flat. Reply rates went from 2% to 10%. Ron Gabrisko joined Databricks as SVP of Worldwide Sales in 2016, later becoming CRO. He watched the company scale from under $1M ARR to a $190B valuation and $1.5B+ revenue run rate. The company still runs a large SDR team, but headcount stayed flat while AE and solutions architect teams kept growing. Ron credits AI-powered outbound: personalized messaging at individual level, automated signals on prospects, and campaign orchestration. Old model was 5,000 contacts into "the automated spammer" with 1-2% response. New model: 6-10% reply rates. At Monaco, co-founder Sam Blond (former Brex CRO) runs one SDR who books 400 outbound meetings per month. Account research, buyer identification, signal tracking, and message customization are automated. Cold calling still needs a human, but the system tells the rep who to call, when, and with what script. ## Every AE candidate builds a demo on the spot Databricks tests for AI-first selling in the interview loop. Candidates pick an account, deliver a CEO-level AI pitch for that company and industry, then build a demo in Genie (Databricks' AI product) and walk the panel through outcomes. After hiring, every rep gets certified on Genie and the CEO-level pitch throughout the year. Ron keeps the SDR team because it is where his AEs come from. Many Databricks SDRs get promoted to AE. Without the team, every AE hire comes from outside and no one mentors the next generation. Some of his best sellers today came up through SDR ranks and build their own programs using Genie and LLMs. ## 60% of rep time in front of customers Ron's rule: "Sell from your feet, not your seat." Target is 60% of rep time in front of customers. Monday and Friday for prep and admin. Tuesday through Thursday on the road. Databricks built agents for account research, presentation prep, exec briefing docs, and proposals. That time goes back into customer meetings. Demos are built on the customer's industry data. AEs create custom demos for specific accounts like Pepsi and Rivian. The company runs Databricks on Databricks, and reps use Genie on their phones. ## ANZ context Databricks appointed Adam Beavis as VP and country manager for Australia and New Zealand in 2023. The company reported more than 150 employees across ANZ at that time, over double its regional headcount from two years prior. Regional customers span financial services, retail, government, mining, utilities, and digital-native sectors. The company's consumption-based pricing model means sellers land with limited access or free seats, then expand usage. That aligns with the broader shift in enterprise software: initial workloads lead into data, analytics, and AI use cases. Principal competition comes from Snowflake and the major cloud providers bundling data and AI into enterprise relationships. ## Worth noting Ron expects the SDR team may shrink over time, but wants an academy for younger sellers with a path into AE or management roles. Sam thinks a bot cold calling on your behalf would be too dystopian today. He expects that could look different in two years.

23 days ago
News

Salesforce hits $800m Agentforce ARR, AI agents now qualifying leads for sales

Salesforce reported $800 million in Agentforce annual recurring revenue for FY2026, with 29,000 closed deals and 169% year-on-year ARR growth. The numbers matter because they show agentic AI moving from experiment to production revenue, and two Salesforce customers are running it directly against SDR and BDR workflows. ## TaskUs: 100% lead follow-up with an AI SDR TaskUs deployed an AI SDR called Liz to cover a sales gap. Sales teams were skeptical until it worked, then asked to expand it everywhere. Result: 100% follow-up rate on every lead, something no human SDR team could sustain at that scale, while human BDRs shifted to qualified, closer-to-purchase opportunities. The operational gain was real. TaskUs compressed newsletter production from 16-20 days down to 10-12 days, a 40% improvement, by re-platforming onto Marketing Cloud with Agentforce. ## Cloud Protection by WithSecure: identifying 98.5% of visitors before they fill a form Cloud Protection by WithSecure built a visitor-matching pipeline that identifies 98.5% of visiting organizations by company name in real time, before anyone fills out a lead form. Two full-time employees shipped it in six months, running website logs through Data 360 and ZoomInfo. They rebuilt lead scoring on 300+ signals: page views, email opens, high-intent actions that recalculate continuously and decay with inactivity. Score reflects right now, not last quarter. Sales stops chasing ghosts. They also automated two-way agentic email. An AI agent reads every reply, qualifies intent, routes sales-ready accounts to the right rep with campaign context, reply thread, and enrichment data attached. No lost threads, no "who owns this lead" scramble. The team running agentic workflows shipped features 3X faster per sprint: 5 before agents, 18 with an agentic team. ## What this means for ANZ sales teams Salesforce has dedicated ANZ leadership for Agentforce. Kevin Doyle runs Agentforce and Data Cloud for the region. Frank Fillmann, Salesforce's EVP and GM for ANZ, has positioned AI agents as a response to tight labour markets and stagnant productivity. The question for sales leaders: if an AI agent can handle 100% lead follow-up and qualify inbound faster than your SDR team, what does your SDR headcount look like in 12 months? TaskUs and Cloud Protection by WithSecure are not running experiments. They are running production workflows that replace manual SDR tasks at scale. Salesforce posted $41.5 billion revenue in FY2026, up 10%. Agentforce ARR approaching $1.4 billion combined with Data 360. The company acquired Qualified, an agentic AI marketing provider built to generate pipeline autonomously, moving directly into the AI SDR and revenue orchestration market. Comp implications: if AI agents handle outbound sequences, lead qualification, and reply routing, the market for $80k base SDR roles contracts. The market for AEs who can work qualified pipeline expands. Territory planning changes when one AI agent covers what used to take three SDRs.

23 days ago
News

Adelaide AI robotics startup raises $15m pre-seed, no sales team yet

**Adelaide AI robotics startup Metacognition raised $15 million pre-seed:** $10 million from Main Sequence in September, topped up with $5 million from AirTree in October. The company is building foundational AI for robotic navigation, including its open-source NavGPT-3 vision-language model and an operating system called Zeos. **Founded by Paul Dalby, Anton van den Hengel, and Stephen Gould,** the company is focused on making AI agents and robots reliable enough for real-world industrial tasks: managing open-cut mines, electricity grids, or warehouse operations using verbal instructions. NavGPT-3 reportedly outperformed models from Alibaba and France's Mistral in simulated navigation tests. **No sales team identified.** Available coverage shows no CRO, VP Sales, or enterprise sales organisation. This is a research and product-development stage company. Founders are academics and researchers; the focus is on building the technology, not yet scaling enterprise GTM. **The raise is large for pre-seed.** $15 million positions Metacognition among Australia's better-capitalised early-stage AI ventures. For context, **Alloy Robotics** raised $11.5 million seed after a $4.5 million pre-seed, focused on reducing robot-test analysis time. AirTree itself announced $250 million seed and $400 million growth funds in 2025, giving Metacognition access to a substantial ANZ investor network. **No public revenue, valuation, or customer numbers.** The company initially targeted industrial systems and verbal-command control for heavy machinery. AirTree co-founder Craig Blair framed the investment as strategic: "If all we do is keep the lights on for models trained elsewhere, the value walks out the door with the tenants." **For sales professionals:** This is not a hiring story. Metacognition is building foundational AI, not scaling a sales org. If you are tracking ANZ AI startup hiring, this company is in research mode. Worth watching if they pivot to enterprise GTM in 12 to 18 months, but no sales roles on the horizon yet.

24 days ago
News

Anthropic IPO targets $1.5T valuation, $80B H1 loss on $16B revenue

Anthropic filed its US IPO prospectus, targeting a Wall Street debut next month at a valuation around $1.5 trillion, down from $2 trillion a week ago. The company disclosed $16.3 billion in revenue for the first half of 2026, up from roughly $1.1 billion in the same period a year earlier, and an $80.7 billion loss driven largely by fair-value adjustments on convertible bonds. The numbers show exceptional growth but also the capital intensity of frontier AI. Anthropic has committed to more than $600 billion in future computing infrastructure and is raising public capital to fund it. Third-quarter revenue hit $16.8 billion, putting the annualised run rate at about $65 billion by July 2026. For sales professionals, the IPO sets a benchmark for how public markets value AI model companies. OpenAI, Google DeepMind, and Meta are Anthropic's direct competitors. Goldman Sachs and Morgan Stanley are leading the deal. What matters for ANZ: Anthropic's go-to-market relies heavily on cloud partners, primarily AWS and Google Cloud, to reach enterprise customers. Public disclosure on headcount, sales team size, or ANZ employee base is limited. The company is building a startup-sales function focused on convincing founders to build on Claude, creating downstream enterprise opportunities, but no major ANZ CRO appointment or local acquisition was identified. The IPO comes as market volatility has delayed other offerings. Anthropic raised a $65 billion Series H in May 2026 at a $965 billion valuation, following a $30 billion Series G in February at $380 billion. Total funding sits around $125 billion, with backers including Amazon, Google, Salesforce, Sequoia, Altimeter, Dragoneer, and Greenoaks. Worth noting: the loss figure is accounting-driven, not purely operational. The IPO prospectus shows a company scaling fast, spending heavily, and betting that public capital markets will fund the next phase. Whether that translates to ANZ sales hiring remains to be seen. For now, the path to enterprise customers here runs through cloud partners, not a disclosed local sales team.

24 days ago
News

US seed rounds hit record 8,625 in 2026, but small VCs are getting squeezed out

Through Q3 2026, an estimated 8,625 US startups closed their first venture round, already beating 2021's full-year record of 8,507. The dollars are up too: $34.3 billion through Q3, pacing ahead of 2021's $40.7 billion. Here is the problem for smaller VCs: they cannot get in the door. Median pre-seed valuations tripled from $4 million in 2020 to $12 million today. Median seed pre-money sits at $18.3 million, up from $16 million last year. Meanwhile, median seed check sizes fell from $3.4 million to $3.0 million, the only stage where deal size dropped in 2026. That $3 million now buys about 14% of a company versus 17.5% last year. Smaller checks, higher prices, less ownership, at the exact moment there are more companies to choose from than ever. The capital is consolidating. Funds of $500 million or more took 78.1% of capital while making up 6% of funds closed. Andreessen Horowitz alone closed $23.8 billion, more than every non-megafund in America combined. Established firms took a record 88.2% of capital. First-time funds raised just $4.9 billion across 81 funds, on pace for the fewest in a decade. Emerging managers closing a fund: 927 in 2022, 440 in 2025, 211 through Q3 2026. The median Series A now costs $17.4 million. A small fund that leads a $3 million seed cannot defend its position at that price. In Q3, deals of $100 million or more were 5% of deal count and 75% of deal value. For the year, 82.7% of US venture dollars went to AI. ANZ has local equivalents. Antler operates across Sydney, Melbourne and Brisbane, while Startmate runs a 14-week accelerator offering A$120,000 in initial funding. Active early-stage investors include Blackbird, Icehouse Ventures, GD1, Wavemaker Partners, OIF Ventures, Investible, Archangel Ventures and Eastend Ventures. Recent deals show the range: QuantX Labs raised about US$5 million at seed after bootstrapping and already counts SpaceX among customers. DataMasque raised US$4 million, with ARR up sixfold since its 2023 seed. Auckland's Partly raised US$50 million from DST Global and Blackbird. For sales teams at early-stage startups, this matters. Fewer small funds means longer fundraising cycles. Tighter capital means slower headcount growth. If your Series A costs $17.4 million instead of $12 million, that is months of runway difference and fewer SDR roles approved. The data says seed is getting cheaper per round while everything after it gets more expensive. Plan accordingly.

24 days ago
News

Anthropic S-1: $6.6M revenue per employee, 900% NRR, 250 accounts above $10M

## The Numbers Anthropic filed its S-1 this morning. First-half 2026 revenue: $16.3 billion, roughly 14× the same period in 2025. Full-year 2025 revenue was $4.6 billion, up from $386 million in 2024. Net revenue retention: approximately 900% as of September 30, 2026. That measures current-quarter revenue from existing customers against the same cohort a year earlier, excluding fixed-price Claude Pro and Max subscriptions. The usage business grew 9× inside existing accounts in twelve months. Customer count: more than 11,900 accounts generating $100k+ in trailing-twelve-month revenue, including 1,900 above $1M and roughly 250 above $10M. The top five customers represent about 25% of revenue for the nine months ended September 2026, though no single customer exceeds 10%. ## What Changed in Sales and Marketing Sales, marketing, and partnerships expense went from $120 million in first-half 2025 to $3.2 billion in first-half 2026. That is 26× growth and now sits at roughly 19% of revenue, up from 10%. The line is larger than total personnel cost ($3.15 billion). The spend covers forward-deployed engineers building custom solutions on customer infrastructure, partnerships with Accenture, Deloitte, KPMG, PwC, plus platform integrations with Salesforce and Snowflake. Anthropic also launched Ode with Anthropic, an AI services company backed by private equity firms to target mid-market accounts. Revenue per employee hit $6.6 million trailing twelve months as of September 30, 2026, up from $1.4 million a year earlier. Personnel cost was $3.15 billion in first-half 2026, with $2.2 billion of that in stock-based compensation. Roughly 70% of the personnel line is equity. ## The ANZ Angle Anthropic opened a Sydney office as its fourth Asia-Pacific location and appointed Theo Hourmouzis as ANZ general manager. Local customers include Commonwealth Bank, Quantium, Canva, and Xero. The company lists partnerships with ANU, Curtin University, and other research institutions. A posted ANZ partnerships role signals the region is being developed through both direct sales and channel leverage. ## What It Means Anthropic priced on usage from day one. That decision now has 250 customers paying $10M+ annually. Few software line items at any company reach that size. The money is coming from labor budgets and new project budgets, not traditional software spend. Compute and infrastructure cost fell from 201% of revenue in first-half 2025 to 55% in first-half 2026. Revenue grew 14× while compute expense grew roughly 4×. Customer cost per task dropped 40% with Claude Opus 5.5 versus the prior version. For sales teams watching the AI software space: Anthropic spent $3.2 billion in six months on go-to-market despite organic developer adoption. Product-led growth at enterprise scale still requires serious field investment, forward-deployed engineering, and SI partnerships. The 900% NRR shows what happens when pricing aligns with customer value expansion. Usage-based models remove artificial caps on account growth. When customers can send more workload without renegotiating contracts, they do.

24 days ago
News

Telstra building AI agents for sales teams, 8,000 staff already using tools

## Telstra pushing AI agents into sales workflows Telstra is building AI agents for sales teams as part of a broader push across its A$23.1 billion operation. The telco says 18,000 employees now have Microsoft Copilot licences, with over 80% using the tools weekly. More than 20,000 staff have completed Data & AI Academy training. The company is running AI initiatives across customer service, internal support, sales productivity and network operations. Its AskTelstra assistant is used by more than 8,000 frontline staff, cutting average customer call times by over a minute. A customer-facing assistant has nearly tripled the number of queries resolved without escalation. For sales specifically, Telstra is applying AI to proposal generation and customer engagement, according to its FY2025 annual report. The company's Connected Future 30 strategy includes partnerships with Microsoft, Accenture, Quantium, Ericsson, Red Hat, Dell and Cisco. ## Accenture partnership compressing timeline Telstra's 2025 partnership with Accenture gives it access to a stated US$3 billion AI investment pool. The goal: compress a five-year data-and-AI roadmap into roughly two years. Worth noting: these are company-reported figures and timelines. The Real Estate Institute of NSW (REINSW) is also expanding its AI approach. The industry body launched an AI board agent called Alice Ing in 2024 and now plans to roll out similar capabilities to its member base. The use case: listings, buyer and seller enquiries, compliance, scheduling and administrative workflows. ## What this means for sales teams Telstra operates as both a large B2B buyer and an AI-services competitor. Its enterprise sales opportunity spans connectivity, cloud, cybersecurity, managed services and AI transformation. Competitors include Optus, TPG Telecom, NBN Co in fixed connectivity, plus Microsoft, AWS, Google Cloud, Accenture and Salesforce in adjacent technology and AI services. The timeline matters. If Telstra can compress five years into two, that sets a benchmark for how fast enterprise buyers expect AI rollouts. For sales teams selling into enterprise, that means shorter evaluation cycles and higher expectations around proof-of-value. The REINSW use case shows how vertical-specific AI agents are moving beyond generic productivity tools. Real estate agencies need workflows that handle compliance, scheduling and enquiry management, not just email summaries. If you are selling AI tools to industries with specific regulatory or workflow requirements, generic demos will not cut it.

24 days ago
News

MP Space raises $23m, no CRO: ex-Tesla duo selling spacecraft batteries

## Ex-Tesla engineers land $23m, now they need to sell spacecraft batteries Sydney startup MP Space raised $23 million in seed funding to manufacture standardised satellite power systems in Australia. The round was led by Blackbird, with participation from Seraphim Space (its first Australian investment), Systemiq Capital, Speedinvest and MFV Partners. Founders Jay Manley (CEO) and Sam Paterson (chief commercial officer) are both ex-Tesla battery engineers. Manley spent eight years on Tesla's Powerwall products; Paterson worked on Model S, X and 3 battery packs before stints at Apple, Zoox and Sydney biotech Vow. They started MP Space in January 2024. The pitch: standardised spacecraft batteries that ship faster and cost less than bespoke systems. The company says its hardware has been designed into missions representing 500+ spacecraft, with the first units scheduled to fly later this year. Customer names, revenue and contract values have not been disclosed. ## The sales structure question Here is the interesting part: there is no CRO or VP Sales. Paterson, the CCO, is handling the commercial function. The publicly listed team sits at roughly seven people, mostly engineers and manufacturing leads. That is typical for a hardware startup at this stage, but selling into aerospace and constellation operators is a long-cycle enterprise motion. Worth watching how they scale that function as they move from early design-wins to production contracts. The capital is earmarked for Australian manufacturing expansion, additional power-system platforms and engineering hiring. The broader satellite power market includes established suppliers like Saft, EaglePicker and GS Yuasa. MP Space's differentiation hinges on speed, standardisation and production scale rather than pure battery chemistry. ## What this means for ANZ tech sales Space hardware is not a typical SaaS play, but the go-to-market dynamics are worth tracking. Long sales cycles, technical validation requirements, mission-critical reliability standards. If MP Space hits the 500-spacecraft pipeline they claim, they will need a structured sales organisation soon. That likely means enterprise AE hires with aerospace or deep-tech experience, probably based in Sydney with travel to the US and Europe where most constellation operators sit. Comp structure for those roles would be interesting: lower volume, higher ACV, longer close cycles. Expect base-heavy splits and quota cycles measured in quarters, not months.

25 days ago
News

Venture capital now only backs billion-dollar shots, leaving $20M ARR SaaS stranded

## The New Math A $20M ARR SaaS company, profitable and growing 6-7x year-on-year, is struggling to secure meetings with tier-one venture capital firms. The reason is not performance. It is fund arithmetic. A $5 billion venture fund raising every 18 months needs billion-dollar gains on individual positions, not billion-dollar exits. In their investment committee memos, a $1 billion gain is the base case. The aspirational case is $25 billion or more. If your realistic outcome does not deliver a position of that scale, you are pitching the wrong fund, regardless of your metrics. This has direct implications for sales professionals evaluating startup offers. Equity packages that looked attractive two years ago now carry significantly different risk profiles. A company unable to raise from top-tier funds may still succeed, but the exit timeline extends and the buyer pool narrows. ## ANZ Context: Anthropic and Modal Anthropic, which has raised more than $18 billion including $4 billion from Amazon, is opening a Sydney office and hiring across sales, research, engineering and leadership. The company grew from fewer than 1,000 Claude users two years ago to approximately 300,000 enterprise customers. That is the scale venture capital now requires. Modal raised a $355 million Series C at a $4.65 billion valuation, then a $750 million Series D five months later at a reported $15.75 billion post-money valuation. Annualised revenue moved from below $100 million to more than $1 billion. Developer-led infrastructure growth at that velocity is what gets meetings. ## What This Means for Sales Teams If you are evaluating an offer from a vertical SaaS company or anything outside AI infrastructure, ask different questions. What is the funding runway? Who are the existing investors, and do they have reserves to carry the company if it cannot raise again? What does the cap table look like, and how realistic is the exit scenario? Companies that would have closed Series B rounds in 2021 are now building 100-investor lists and budgeting two to three months for fundraising processes that used to take two weeks. That changes quota relief, ramp expectations, and how long your equity will remain illiquid. The advice for founders also applies to sales hires: understand the fund math of whoever controls your company's cap table. If the outcome your startup is building toward does not meet their minimum return threshold, your equity is priced for an exit that may not happen on the timeline you were sold.

25 days ago
News

Canva ships home-grown AI model after cutting revenue forecast, valuation

Canva launched Craft, its first image generation model built in-house, positioning it against OpenAI, Adobe Firefly, Google and Microsoft Designer. The move comes three months after CEO Melanie Perkins cut the company's revenue growth forecast by a third, from 30% to 20%, citing AI costs that were "too high" when relying on frontier models. Two weeks after that announcement, Canva's valuation dropped $11 billion to US$31 billion in an employee share sale, down 20% from the prior year. Craft is designed to handle composition, layout and text before generating images, targeting legible lettering and fewer required edits. The company says advanced model distillation cuts the number of inference steps to a fraction of what competing models need. Translation: faster generation, lower compute cost. For sales and marketing teams already using Canva for presentations, social content and brand assets, Craft slots into existing workflows rather than requiring a separate tool. That is the enterprise play: AI features that increase adoption and drive Pro, Teams and Enterprise upsells across Canva's 31 million paying customers. ## Why this matters for ANZ sales teams Canva remains one of Australia's largest tech companies, employing more than 5,000 globally with Sydney headquarters. The company hit roughly US$3.5 billion in 2025 revenue, with third-party estimates putting annual recurring revenue near US$4 billion. For sales professionals using Canva to create pitch decks, one-pagers and campaign assets, the in-house model means faster turnaround without switching to ChatGPT or Midjourney. Whether that translates to material workflow improvement depends on how Craft performs against established tools sales teams already use. Canva is developing faster, cheaper Craft variants and planning integrations that let AI assistants tap the model. The company did not disclose specifics on pricing or how Craft impacts existing subscription tiers. ## Market context Canva competes with Adobe Express and Firefly, Microsoft Designer, Google image tools, OpenAI products and specialist platforms like Midjourney. The core opportunity is less about selling a standalone image model and more about converting AI capability into higher-value bundles. With 265 million monthly active users, Canva has distribution scale that newer AI-image startups cannot match. The risk: if sales teams can generate acceptable marketing collateral directly in ChatGPT or Google tools, Canva's freemium funnel weakens. Building Craft in-house gives the company cost control and product differentiation, but execution will determine whether it defends market position or becomes another feature in a crowded AI landscape.

26 days ago
News

SaaStr runs sales with 1.25 humans, 21 AI agents, claims $2.4M closed revenue

SaaStr, the SaaS media and events business founded by Jason Lemkin, says it runs go-to-market with 1.25 human sales staff and 21 AI agents. The company attributes $4.8 million in pipeline and $2.4 million in closed revenue to the setup by early 2026. The core is '10K', a custom agent built on Replit and connected to more than 30 GTM tools: Salesforce, HubSpot, QuickBooks, enrichment sources, and outbound platforms. It runs inbound, outbound, renewals, support tickets, and quote-to-cash in one centralised queue called the Prospector. **How the Prospector works:** Every lead, reply, renewal notice, or support case lands in one surface. The agent filters noise, assigns a goal per record type (booked meeting for inbound, closed case for support), and works each until resolution. First-party data runs first: event attendance, newsletter subscription, open opps, downloaded assets. Third-party enrichment (Exa, Clay, Apollo) comes second. The agent generates personalised prospectuses and books meetings without human handoff unless it cannot answer a question. SaaStr started 10K 16 months ago as a Notion helper. Tool connections expanded over time, the way you would grant a new hire system access as they prove out. **Context for ANZ sales teams:** This is agent-led growth taken to an extreme operating model. Most companies testing AI sales agents run them alongside SDR teams or use them for list building and research, not as primary booking infrastructure. SaaStr's claim is that the agent handles the full motion, inbound through close, with humans approving deals and intervening on edge cases. The setup assumes strong first-party data (CRM history, event attendance, content engagement) and tool integration across the stack. Teams without centralised account data or with fragmented systems would need to consolidate before replicating this. SaaStr competes for SaaS founder attention, event sponsorships, and marketing budgets globally. No dedicated ANZ team or local headcount is evident. For Australian and New Zealand sales professionals, the takeaway is operational: whether agent-led GTM at this scale works outside a founder-led media business with strong inbound remains an open question. The comp and hiring implications are obvious: if the model proves out, traditional SDR and AE headcount shrinks further. Lemkin has published the full playbook at saastr.ai/agents, updated regularly with case studies and tool connections.

27 days ago
News

SaaStr runs 21 AI agents with 3 humans, agents now pick vendors

## The new buyer is not a person SaaStr is running 21 production AI agents with three humans. Founder Jason Lemkin and Chief AI Officer Amelia Lerutte say the agents now pick vendors, kill vendors, and tell them to skip buying anything at all. The company says its inbound AI agent handled hundreds of thousands of prospect chats and booked 614 meetings. Average sponsor ACV: $85,000. SaaStr says this replaced the work of roughly three BDRs. Broader automation across the stack reportedly covers work previously done by about 10 SDRs and AEs combined. These are company-reported figures, not audited metrics. But the implication is concrete: buyers are asking their agents before they talk to your reps. ## What this means for vendors Lemkin's rule: "We'll try it if 10K, our AI VP Revenue, wants to. It's up to 10K now." SaaStr's Connect agent recently picked Render as its platform choice. The agent knew the app requirements (always-on server, background workers, long-running jobs, zero-downtime deploys) and ranked five options. Render landed first. Railway second. AWS and Google Cloud were overkill. Vercel came fifth, marked not suitable for minutes-long jobs. Lemkin had never used Render. It would not have occurred to him for this project. The agent picked it anyway. ## The pitch has shifted In 2021, developers became the real buyers of infrastructure software. You won the engineer, the engineer brought the budget. Now the same thing is happening one layer down. More vendor decisions start with a question to an agent. A growing share end there too. For sales teams: your content, your docs, your pricing transparency, your technical specs all need to work when an agent is doing discovery. The human buyer might never see your deck if the agent filters you out first. SaaStr is unusually vocal about its stack, which makes it a useful case study. But the shift is broader. Agents are compressing traditional go-to-market functions. The vendors who win will be the ones whose material lands with the agent, not just the AE.

27 days ago
News

Clay, Figma pricing shift: 1% of users drive 75% of AI costs

## The Problem AI costs follow a power law. Across dozens of high-growth pricing projects, Willingness to Pay founder Ulrik Lehrskov-Schmidt sees the same pattern: 1% of users drive 50-75% of total AI spend. In extreme cases, 0.1% of users generate three-quarters of the inference bill. One client had a customer spending 380x more in tokens than it paid in revenue. That is not a pricing problem. That is a unit economics problem. ## What Changed Traditional B2B had flat costs. Every customer cost roughly the same to serve, so leaving value on the table was an opportunity cost that never hit the P&L. AI added a second curve. Now you have spiky value (one customer gets 1,000x more than another) and spiky cost (one customer burns 380x more than they pay). The ratio between those curves keeps moving. Model prices shift, capabilities improve, customers find new workflows. Lehrskov-Schmidt calls this "the sticks dance." ## What Clay, Figma, and Higgsfield Are Doing Clay (data enrichment, AI workflows, $100m+ ARR) has spiky value and spiky cost. It needs usage-based pricing that tracks customer outcomes, not token spend. Figma (collaborative design, $1.3bn revenue estimate) sits in the same quadrant. Broad enterprise user base, AI features expanding usage, compute costs requiring guardrails through credits or tiers. Higgsfield (generative video, $1bn annualised revenue reported) has flat value and spiky cost. Its $19-$99 plans included unlimited AI generation. Fixed revenue, uncapped cost. That structure produces the 380x customer. ## What This Means for Sales Teams If you are selling AI-enabled platforms, deal qualification just got harder. A logo that looks like $500k ACV can turn into a loss if usage spikes. One client's largest enterprise accounts ran $20k/year in AI costs on $500k contracts. That works. The 380x customer does not. Quota planning needs to account for this. So does territory assignment. The biggest users are not always the most profitable accounts. Ramp periods assume predictable unit economics: if 1% of your book can burn through margin in three months, your comp plan has a problem. Revenue teams need to understand customer usage patterns before the deal closes, not after. That means sales engineering, implementation, and success functions need to be aligned on what sustainable usage looks like. If your product has unlimited AI features and variable compute costs, you are carrying risk the comp plan does not price in. ## The Takeaway Flat pricing worked when costs were flat. AI broke that. If your platform lets one customer use 380x more resources than another for the same price, you are not leaving money on the table. You are losing it.

28 days ago
News

Getty Images Chapter 11: $910M revenue, 27% margins, lost 81,000 subscribers to AI

Getty Images filed Chapter 11 in Delaware. The company is running at $910M annual revenue with 27% EBITDA margins, but it carried $2.1B in debt and lost 81,000 annual subscribers in 12 months. The Getty family and Koch Equity Development are putting in up to $533M in new equity. Public shareholders, who bought in via the 2022 SPAC at $26, get nothing. The stock hit $0.083 before delisting. ## What Killed It AI took the low end. Buyers who wanted generic stock photos moved to generative tools. Getty lost 25% of its subscriber base year over year, dropping from 321,000 to 240,000 annual subscribers. Total purchasing customers fell 10%. Subscription retention dropped from 93.4% to 88.4%. That is not a pricing cleanup. That is customers leaving. The high end held. Editorial revenue grew 9%, and corporate accounts (over 60% of revenue) were up 6%. Enterprises paying for rights-cleared content, news coverage, and legal indemnity stayed. The microstock customers left. ## The Failed Shutterstock Merger Getty and Shutterstock signed a merger in January 2025 to create scale and service the debt. The UK Competition and Markets Authority wanted Shutterstock to sell its editorial business. Getty's board declined. The merger died in July 2026 after 18 months and $100M in costs. Three months later, Getty filed Chapter 11. ## The Restructuring Noteholders holding 83.5% of unsecured bonds agreed to retire $270M of debt. Secured lenders agreed to lower interest and extend maturities. Koch is providing $143M to fund operations through the process. Getty says customer operations continue normally. Suppliers get paid in full. The company expects to exit Chapter 11 by year end. ## What It Means for ANZ Sales Teams Getty sells into ANZ through enterprise licensing for brands, publishers, agencies, and broadcasters. The restructuring creates procurement uncertainty: will contracts hold, will pricing change, will product investment continue? For sales teams selling visual content or competing with Getty, this is a signal. AI is taking the transactional, price-sensitive segment. The enterprise segment that values legal indemnity and editorial coverage is stable, but retention pressure is real. Getty's subscriber decline and retention drop are leading indicators. If your book includes transactional customers who could switch to free or cheaper AI alternatives, watch your renewal rates. The high end is defensible if you deliver something AI cannot replicate. The middle is getting squeezed. Getty changed ownership three times in 14 years: Carlyle bought it in 2012, it went public via SPAC in 2022, now Koch and the Getty family are taking it private again. Public shareholders lost 99.7%. That SPAC math is worth remembering next time someone pitches you equity in a pre-IPO startup.

28 days ago
News

Modal hits $1B run rate, raises at $15.75B: infrastructure spend concentrates

Modal raised a $750 million Series D at a $15.75 billion post-money valuation, according to reports discussed on the 20VC podcast. The AI infrastructure company closed a $355 million Series C at $4.65 billion only months earlier. Modal says it has exceeded $1 billion in annualised run rate, with usage tripling since the Series C. These are company-reported run-rate figures, not audited revenue. The round illustrates how capital is concentrating into AI infrastructure while other categories struggle. Harry Stebbings cited a $20M ARR company growing 600-700% and profitable that could not raise. Jason Lemkin confirmed similar dynamics: top-decile SaaS businesses outside AI are getting shut out. Rory O'Driscoll framed it as vertical software facing an unstated fear that AI makes the category obsolete. Modal's investor base includes Accel, Kleiner Perkins, Menlo Ventures, General Catalyst, Redpoint, Lux Capital, Amplify Partners and Bain Capital Ventures. Customers include DoorDash, Ramp, Linear, Cognition and Runway. The company's Sandboxes product, which executes AI-generated code, is reportedly a $300 million annualised business with nearly 3 billion instances launched since 2023. For sales teams, the implication is clear: if you are selling infrastructure into AI workloads, budgets are open. If you are selling vertical SaaS, you are fighting an uphill battle regardless of metrics. Lemkin's advice for companies with strong numbers outside the hot categories: meet 100 VCs. Top decile used to mean an automatic term sheet. It does not anymore. Modal was founded in 2021 by CEO Erik Bernhardsson, previously at Spotify and Better.com, and CTO Akshat Bubna, an early engineer at Scale AI. No public data is available on total headcount, sales team size or ANZ presence. The company's growth model appears developer-led and product-led, with enterprise adoption driven by infrastructure spend rather than traditional sales motions. The funding environment confirms what quota-carrying teams already know: if your product is not in the AI chain, expect longer cycles and tighter scrutiny. If it is, expect competitors with war chests.