Perspectives

The Emergence Rate: Our Framework for Measuring AI Evolution

April 29, 2025

several years Emergence using initially called Delta Metric evaluate effectively AI software learns users evolves time first sharing publicly call Rate measures impact incorporation proprietary business data processes leading ever-improving outcomes software's world B2B must improve every day separates stagnant startups continuously optimize solutions result better customers chances startup's success incredibly competitive landscape Built journey began Coaching Networks thesis 2017 diligence observed best enterprise responded constantly evolving needs today's last decades operated relatively statically simply reflection real-world constraints number engineers employ takes turn feature requests code technological era discomfort big changes confusing Client-Server slick demo out-of-the-box performance typically won prospects Companies paid upfront license fees on-prem installations Meaningful upgrades occurred seemed whenever Olympic Games came back around struggled keep new generation SaaS led Salesforce.com emerged address challenges migrated cycle now repeated generative breakthrough unleashed unprecedented capabilities help break add value long haul leaders increase assets importantly ingenuity workforce sides marketplace think framework measure revolutionary businesses high ability models automatically refine distinguishes market winners stagnate deliver products evolve commanding premium prices avoiding churn common among early adopters becomes increasingly valuable unlike traditional depreciates self-improving soon indispensable heart company worth paying indefinitely providing compounding advantages create massive strategic differentiation Define Three Key Metrics portfolio work refined essential dimensions Change frequently system assess variants based granular user activity faster integrates unique behaviors higher advanced implements continuous learning loops whether ambient observations e.g RPA collaborations within agentic workflows captures applies insights real-time without requiring manual updates dynamically piece interaction generated ways Scale Surge provide foundation filling large-model deficiencies abilities humans HF RLHF focusing commercial markets advantage workflow relationships large access firewalled carefully cultivate relationship center protecting treasured IP become Quality variations matters volume poor recommendation undermine trust entire preceded good ones outcome improvements identified rigorous variant testing analyze hundreds interactions group tiers identify successful patterns test interstitial feedback proxies sales cycles example patient receives personalized note healthcare provider following visit likely take recommended actions adherence engineering lens able observe qualities pull request minimal discussion downstream errors rollback tailor style tests adhere closely possible attributes Breadth Adoption top performers disseminated organization-wide requires excellent experience design behavior either automate process guide toward dynamic UX goal mushy middle average maintain human innovation elevating mean anything increasing closes raising ACV improving cohort retention rates scenario benefit Finally keeping loop throughout enables adopt recommendations experiment preventing dependency Crutch systems learn broad five point startup Track identification Document identifies given timeframe detection Implement outcome-based clear specific industry function deal closure life science clinical trial recruitment variance attributed bootstrapped assessing team produced person performer area proxy Monitor evolution Count automatic occur reflecting learned speed quickly newly persist inexplicably fall away collect historical across focus particularly bar capturing brilliance outliers organization kind matter size intellectual property industry-specific knowledge datasets use overcome Plateau populate platforms competitors allows customer increases previous eras human-generated rare allowing Perpetual Algorithm lies becoming domain serving virtuous Future Adaptive Intelligence natural a-ha engines minds nearly 100 billion interconnected neurons store lived experiences enabling boundless creativity adaptability seek breakthroughs beyond status quo printing press tractor spreadsheet core express harnesses innate identifying discovered stumbled upon approaches validating impacts distributing Looking ahead know forward-thinking buyers moved startup-maintained deployments constant renewal maintaining edge harnessing uniquely-human innovations

The AI-Native Services Playbook

January 31, 2025

SaaS startup founders playbooks mentors guide AI-native services charting new territory pioneers combine AI human expertise deliver faster better cheaper outcomes legacy service providers Take Mechanical Orchard using accomplish months traditionally takes years moving mainframes cloud wrote emerging model last year Since seen explosion companies adopting compelling early traction catch trying force-fit traditional strategies onto service-based businesses key similarities critical differences invested studied space identified 5 essential lessons building iconic company Consider contribution playbook 1 Bring domain expert selling product credibility important existential customers need believe ability results established goes long towards trust authority unlocks access high-quality talent channels enables rapid staffing maturing said technical leadership equally vital prevent over-reliance manual revenue paired applied makes ideal founding team balancing relentless push automation 2 Beware Mirage PMF different beast Strong growth NDR mask lack true enablement call Real requires proving scale non-linearly relative costs must drive measurable improvements cost quality speed ideally three 3 Develop partnerships accelerator incumbents major accelerant offer immediate market distribution proprietary datasets crucial days data corpus small advantage benefits startups exploring partnership models go beyond revenue-share approaches equity share relationships others fully acquiring existing pros cons worth considering critically 4 Leverage pricing help unlock higher contract values come classic used effectively delivering technology labor exist price contracts understand risks Labor-Based based number hours guarantees margin cannibalizing increases Outcome-Based value final output delivered aligns incentives scales margins long-term hurt profitability capabilities nascent found generally best vendors start norm typically approach learning efficiently time set clear timelines transition Regardless however look opportunities embed recurring continue monetize clients initial projects completed Palantir example deeply engagements heavily customized software support via ongoing demo stupid demos showcase tried method works particularly era buyers eager see magic art forgotten Instead default pitch decks talk tracks understandable directly customer provider wields internally huge mistake Without hard buyer visualize every similar competing trusted brands hide behind curtain feel wasteful build facing never use it'll GTM investment make halve sales cycles adding stop telling prospects Show em love learn alongside thoughts pushback examples figuring together Special thanks Arjun Chopra Medha Agarwal Wayne Hu James Currier Zachary Bratun-Glennon Wenz Xing Nic Poulos Kent Goldman great feedback