# Upside — Full Documentation for LLMs > Upside is the data layer for the age of agentic GTM. It forensically reconstructs what actually happened in every account — accounts, people, deals, and the touchpoints between them — into one connected, healed record, then makes that record available to your team, to the AI tools and agents you already use, and to the custom apps you build on top of it. ## Company Overview Upside was founded by a team with deep experience in B2B data, attribution, and go-to-market operations: - **Mada Seghete** — CEO & Marketing. Branch ($100M+ revenue) founder and CMO, Stanford MBA. - **Alex Bauer** — Product. Branch early employee and Head of Product. - **Dan Ahmadi** — GTM & Ops. 12+ years in B2B Demand Gen and SDR management. - **Jonas Bauer** — Engineering. Branch Principal Engineer, 7+ years dev experience. ### Investors XYZ Ventures, Emerson Collective, Peterson Ventures, and Westbound. --- ## The Problem Upside Solves B2B go-to-market data is broken. It lives in silos across CRM, marketing automation, email, call recording tools, event platforms, and more. Two things follow from that fragmentation: no tool can say what *actually* happened in an account, and any AI you point at the raw mess is confidently wrong because it has no grounded model of your business. Traditional attribution makes the first problem worse, forcing teams to choose between rule-based approaches that misrepresent reality: - **First-Touch Attribution** assigns 100% credit to the first interaction — ignoring everything that followed. - **Last-Touch Attribution** assigns 100% credit to the final interaction — ignoring everything that preceded it. - **Multi-Touch Attribution** distributes credit across touchpoints using arbitrary rules (linear, U-shaped, W-shaped) that don't reflect actual influence. - **Zero-Sum Attribution** forces every dollar to be assigned to exactly one source, creating internal competition rather than insight. The result: revenue teams can't answer fundamental questions about what's working, and the AI tools they're adopting can't either. --- ## Upside's Approach: A Forensically Healed Data Layer Upside takes a forensic, not formulaic approach. Instead of applying rules to pre-captured CRM data, Upside: 1. **Connects** to your existing GTM tools (CRM, email, call transcripts, marketing automation, web analytics, and more) 2. **Heals** fragmented data through persona resolution, buying group detection, and identity stitching 3. **Enhances** the data by deconstructing email chains, surfacing buried touchpoints, and categorizing interactions 4. **Reconstructs** the full buyer journey from raw signals, inferring missing touchpoints that traditional tracking misses This produces a complete, contextual model of how deals actually unfold — not just what the CRM recorded. That healed record is then made usable in three ways: directly by your team, by the AI tools and agents you already use, and by custom apps your operators build on top of it. --- ## What the Data Layer Lets a Team Do Upside's product surfaces map to four things the healed record makes possible. ### 1. Know what actually happened A forensically reconstructed knowledge graph of every account, deal, and interaction — not siloed CRM rows. - **Data Foundation** — one clean, connected record of every account, deal, and interaction. Persona resolution stitches fragmented identities across systems; buying group detection assembles the real champions, influencers, and decision-makers behind a deal; email-chain deconstruction isolates each message as a measurable touchpoint; and touchpoint recovery surfaces dark-funnel activity invisible to traditional tracking. - **Dashboard** — explorers and report cards that answer leadership's questions without SQL, including an Account Explorer (a visual minimap of engagement momentum over time), Milestone Analyses (AI-driven deal-origination insights), and Account Story (a chronological, context-rich narrative of every interaction on an account). - **Genesis** *(coming soon)* — signals extracted from unstructured data and encoded as a knowledge graph, so AI reasons over your business instead of just querying it. ### 2. Answers that used to be impossible Analyst-quality insights for questions that were never practical to answer, run across every account at once and returned as structured, cited data instead of one essay at a time. - **Deep Research** — analyst-quality answers across a whole cohort of accounts or deals at once, returned as structured, cited data. - **Pipedash** — AI-powered, impact-weighted multi-touch attribution, with cited evidence behind every dollar. Credit is assigned by actual impact on deal outcomes rather than arbitrary rules, and can be analyzed by channel, by person (SDR/AE), or by webpage. ### 3. Make AI correct about your business Point the AI tools your team already uses at the healed record, and keep every agent grounded in your definitions, your pipeline, and how your business actually works. - **MCP** — your GTM data inside the AI tools you already use, via the Model Context Protocol. - **Upside Agent** — all the power of Claude, sandboxed to your GTM data. - **Radiant** — coaches every agent with your definitions and institutional knowledge so answers stay consistent and correct. ### 4. Build the GTM tools you used to buy Custom GTM tools your operators build themselves, because the people closest to the problem understand it best. The reports, trackers, and internal apps you used to buy as point tools or wait on engineering for, your team now builds on Upside — wired to live data — in an afternoon. - **Miniapps** — turn an insight into a live app the team can use. - **Insight Engineering** — GTM engineering on tap: Upside analysts build your surface for you. --- ## Use Cases Upside serves the operators and leaders on modern B2B go-to-market teams. ### Ask Your GTM Data Ask anything about your GTM data in the AI tool you already use — no SQL, no waiting on a report. ### GTM Apps Build interactive dashboards and internal tools on live data, and publish them to exactly the right people — instead of buying point tools or waiting on engineering. ### Account Management Know the next play for every account, with a complete, multi-threaded picture of the buying group, including touchpoints that happen outside the CRM. ### AEO (AI Engine Optimization) Measurement Measure how AI engines drive your pipeline — how they surface, cite, and recommend you, and what that produces. ### Revenue Attribution Attribution modeled the way your business works: forensic and impact-weighted, reconstructing the full journey and putting cited evidence behind every dollar. ### Event & Field Marketing Efficiency Prove event ROI your CRM can't see — connecting event and field spend to real pipeline and revenue. --- ## Security & Compliance Upside earned SOC 2 Type II certification before shipping its first product. Security shapes every infrastructure decision, integration, and line of code. ### Certifications & Compliance - **SOC 2 Type II** — Independently audited controls for the security trust service criteria, verified through continuous monitoring and annual third-party examination. - **GDPR** — Full compliance with the EU General Data Protection Regulation. - **US Data Privacy** — Compliant with US state-level data privacy regulations. - **CAIQ STAR Level 1** — Cloud Security Alliance STAR self-assessment for cloud security. - **Cloud Security Alliance** — Member of CSA, adhering to cloud security best practices. ### Security Principles - **Encryption at every layer**: Data is protected in transit and at rest. Every connection, store, and backup is encrypted by default. - **Enterprise SSO & SCIM**: Supports any SAML or OIDC provider for identity management and automated user provisioning. - **Your data never trains AI**: Customer data is never used to train generative AI models (including LLMs and foundation models) or to build products for other customers. Data is only used to deliver Upside's services. Trust Center: https://trust.upside.tech --- ## Glossary of Key Terms - **Account Story**: A chronological, context-rich narrative of every interaction and touchpoint associated with an account. - **Agent Grounding**: Constraining an AI agent to a company's real, healed GTM data and definitions so its answers reflect how the business actually works rather than generic priors. - **Buying Group**: The collection of individuals at a target account involved in a purchasing decision — champions, influencers, decision-makers, and gatekeepers. - **CAC (Customer Acquisition Cost)**: The total cost of acquiring a new customer, including all sales and marketing expenses. - **CAC Payback Period**: The number of months for a customer to generate enough gross margin to cover acquisition cost. - **Dark Funnel**: Buyer activities that happen outside traditional tracking — word-of-mouth, communities, podcasts, peer conversations. - **Forensic Attribution**: Reconstructing the full customer journey from raw signals rather than relying on pre-captured CRM data. - **GTM Engineer**: An operator who builds go-to-market tooling and automation directly, rather than buying point tools or filing engineering tickets. - **Impact-Weighted Attribution Model**: Attribution based on actual impact on deal outcomes rather than arbitrary rules. - **Knowledge Graph**: A connected model of accounts, people, deals, and touchpoints that AI and people can reason over, rather than siloed records. - **MCP (Model Context Protocol)**: An open standard for connecting AI tools and agents to external data and tools; Upside exposes your healed GTM data over MCP. - **Miniapp**: A custom, live-data internal app built on Upside by an operator, replacing a bought point tool or an engineering request. - **Persona / Identity Resolution**: Matching and merging fragmented person records across GTM systems into unified profiles. - **Pipeline Velocity**: A composite metric (opportunities × deal size × win rate ÷ cycle length) measuring revenue engine throughput per day. - **Revenue Attribution**: Connecting revenue outcomes back to the marketing and sales activities that influenced them. - **Touchpoint**: Any interaction between a prospect/customer and the organization. - **Zero-Sum Attribution**: An approach where every dollar must be assigned to exactly one source — intuitive but misleading. --- ## Case Studies - **Assembled**: From attribution guesswork to a pipeline surge — AEO and multi-touch attribution. - **Comply**: From unfiltered Salesforce data to marketing intelligence — board-ready data and custom miniapps. Featuring Carl Gunlefinger, Sr. Marketing Ops. - **Cresta**: From attribution to AI deep research. - **Dscout**: Driving confident growth with Upside — account dossiers and multi-threading, featuring Charlie Flanagan, Head of GTM Ops. - **Graphite**: Uncovering the real channels behind 100% inbound growth — featuring Ethan Smith, CEO & Founder. - **Syndio**: Driving strategic GTM decisions with Upside — ADR restructuring and event ROI. Full case studies: https://upside.tech/customers --- ## Resources - Platform overview: https://upside.tech/platform - Blog: https://upside.tech/blog - Glossary: https://upside.tech/glossary - Webinars: https://upside.tech/webinars - GTM Engineer community: https://upside.tech/gtme --- ## Contact Website: https://upside.tech Trust Center: https://trust.upside.tech