FAQ
Frequently asked
Frequently asked
questions.
Answers to the questions B2B leaders ask before adopting Falora — the platform, the AI guardrails, the integrations, and the GTM engineering principles behind it.
About Falora
What is Falora and how does it work?
Falora is an AI-powered growth intelligence platform that connects to your marketing stack, CRM, and customer data to surface actionable growth opportunities. It combines proven methodologies from 6 years of hands-on growth experience with advanced AI to deliver strategic recommendations tailored to your business. Think of it as an operational brain that continuously analyzes your data, identifies patterns, and proposes experiments, so your team can focus on execution rather than analysis.
How long does it take to set up Falora?
Most teams are up and running in under 30 minutes. Falora connects directly to your existing tools like HubSpot, Salesforce, Meta, LinkedIn, and Google Ads, and starts ingesting data immediately. You'll receive your first strategic insights within the first day, and the system gets smarter the longer it runs. There's no complex onboarding process or months of setup. Just connect, configure your goals, and go.
Is my data secure with Falora?
Absolutely. We use enterprise-grade encryption both in transit and at rest, and we never share your data with third parties. Your data remains fully yours. You can export or delete it at any time. We're fully GDPR-compliant and undergo regular security audits to ensure your information stays protected. We also offer data residency options for enterprise clients with specific compliance requirements.
Can Falora integrate with my existing tools?
Yes. Falora integrates with 50+ tools across your entire growth stack, including HubSpot, Salesforce, Meta Ads, LinkedIn Ads, Google Ads, Google Analytics, and more. The platform is designed to work alongside your existing workflows, not replace them. For enterprise clients, we also offer custom integrations and API access to connect proprietary systems or niche tools specific to your industry.
Do I keep control over the AI suggestions?
100%. Falora operates on an 'AI proposes, you decide' model. Nothing goes live without your explicit approval. Every experiment, budget shift, content recommendation, and strategic pivot is presented as a suggestion with full context and reasoning. You stay in the driver's seat while Falora handles the heavy lifting of data analysis, pattern recognition, and opportunity identification.
What's the difference between Falora and a growth agency?
A growth agency provides people and execution on a retainer. You're paying for their time, waiting for weekly reports, and working on their timeline. Falora is an always-on AI platform that works inside your existing stack 24/7, continuously learning from your data and surfacing opportunities in real time. It doesn't replace your team. It amplifies them by automating the analysis and strategic thinking that would normally take consultants weeks. The result: faster insights, lower cost, and a system that gets smarter with every data point.
What is a growth engine?
A growth engine is a repeatable, data-driven system that continuously identifies, tests, and scales the most effective ways to acquire and retain customers. Instead of relying on one-off campaigns or gut feelings, a growth engine combines strategy, experimentation, and real-time analytics into a structured process. Falora builds this engine for you by connecting your data, generating experiments, analysing results around the clock, and adapting your roadmap based on what actually works.
Is Falora only for agencies?
No. While agencies can certainly use Falora to serve their clients more efficiently, Falora is primarily built for in-house teams. It gives CMOs, commercial leaders, and founders the knowledge and tools to create a profitable go-to-market strategy and implement it directly, without depending on external consultants or agencies. Falora puts the strategic intelligence inside your company, so your team stays in control.
MCP for GTM: the new integration layer for autonomous revenue
From our research →
What is MCP and why does it matter for GTM?
MCP (Model Context Protocol) is an open standard that defines how AI agents request and receive context from enterprise systems. For GTM teams it matters because it changes the integration pattern from custom API wiring to a standardised protocol. When properly implemented, agents inherit existing authentication, authorization and policy controls rather than bypassing them.
Is MCP safe to deploy in production?
Selectively. Help Net Security reported in May 2026 that 1 in 4 MCP servers opens AI agents to code-execution risk, with multiple CVSS 9.0+ vulnerabilities disclosed in the first half of 2026. The standard is in the brittle phase where adoption has outpaced governance. Treat MCP servers as third-party supply-chain components with the same due diligence as any subprocessor.
How should B2B GTM teams adopt MCP in 2026?
Through a vetted vendor-managed layer rather than raw self-hosted MCP servers. The platform that abstracts MCP behind a security perimeter is the safe deployment path. Direct MCP wiring is only appropriate for teams with mature security engineering capability and a clear use case.
What changes about my GTM stack when I adopt MCP?
Three things. One, integration cost between agents and systems drops by an order of magnitude. Two, your existing identity and access management extends to agent actions automatically. Three, vendor selection criteria shift; pick vendors that expose data via MCP, not vendors that require custom integrations.
Will MCP replace REST APIs?
No. MCP and REST coexist. MCP is the protocol agents use to consume context; REST is the protocol systems use to expose data. The two operate at different layers. In practice, MCP servers wrap existing REST APIs to make them agent-consumable, not replace them.
Self-reported attribution: the only B2B attribution that survives 2026
From our research →
What is self-reported attribution?
Self-reported attribution is a measurement method in which the buyer answers a required free-text or open-ended question (typically How did you hear about us) on a high-intent form. The response is captured as the primary attribution source, often weighted alongside (rather than instead of) tracked attribution data.
Why is multi-touch attribution failing in B2B?
Three reasons. One, 70% of the B2B buyer journey is anonymous (6sense, 2025), so tracking cannot see the deciding touchpoints. Two, 84% of content is shared through dark channels (DMs, Slack, podcasts) that produce no intent data. Three, AI-engine referrals from ChatGPT and Perplexity often arrive without a referrer header, defaulting to direct or unknown in analytics. Multi-touch attribution measures a shrinking minority of the actual journey.
How do I implement self-reported attribution?
Add a required free-text How did you hear about us field on every high-intent form (demo request, pricing inquiry, contact sales). Do not use a dropdown; dropdowns bias the answer. Capture the response in your CRM as a first-class attribution field. Categorise the responses weekly using simple keyword rules. Compare reported channels with tracked channels and weight budget toward what buyers actually report.
Does self-reported attribution work for product-led growth?
Yes, with a small adjustment. Ask the question at the moment of conversion to a paid plan rather than at signup, when the answer is more likely to be honest and considered. The PLG free-tier signup is often a vague answer (Google search, Twitter); the paid-plan conversion is where the real discovery answer surfaces.
What is the difference between self-reported attribution and survey-based attribution?
Self-reported attribution captures the answer at the moment of conversion, in the natural form flow, on 100% of conversions. Survey-based attribution sends a separate survey post-purchase, typically capturing 15 to 30% of customers with selection bias. The self-reported method is more reliable, higher volume and harder to ignore.
The signal-based selling playbook: Clay + Common Room + orchestrator
From our research →
What is signal-based selling?
Signal-based selling is the practice of triggering outbound contact on observed buying behaviour (a hire, a funding round, a product-usage spike, a competitor switch) rather than on a static cadence. The signal is the trigger, the message is the response. It replaces the spray-and-pray motion of legacy outbound with a behaviour-driven one.
What is the best signal-based selling tech stack in 2026?
The reference stack is Common Room for product, community and dark social signals; Clay as the orchestration workbench that enriches the signal into a workable record; an autonomous platform (Falora or equivalent) for activation across email, LinkedIn and voice; and a measurement layer that closes the loop. Best-of-breed at the signal and enrichment layers, single platform at orchestration and governance.
Which signals predict revenue best?
In our portfolio of 18 GTM rebuilds, three signal classes outperform: hiring signals (new VP-level role at a target account), product-usage signals (engagement spike or feature adoption), and switch signals (competitor adoption or sunset). Intent data from 6sense and Bombora is useful as a baseline; the three above are the leading indicators.
Do I need both Clay and Common Room?
If your buyer lives in communities, free-tier products or open-source repos, you need Common Room. If your motion depends on custom enrichment across multiple data sources, you need Clay. Mature signal-based teams almost always run both because they solve adjacent problems; Common Room observes the signal, Clay orchestrates the response.
What is the ROI of signal-based selling?
Across the deployments we have measured, signal-triggered outreach produces 3-5x the reply rate of cadence-triggered outreach (typically 3-6% vs 0.8-1.2%) and 2-3x the qualified-meeting conversion. The catch is operational complexity; the stack is more sophisticated and requires named ownership of signal definitions.
The Governance Mirage: why 74% of enterprises rolled back AI agents
From our research →
Why are enterprises rolling back AI agents in 2026?
VentureBeat's Q1 2026 research found three primary causes: 43% had no clear owner for AI governance, 23% could not agree who owned it, and 31% named vendor opacity as the single biggest obstacle. The model is rarely the problem; the runtime, the data context and the human-in-the-loop architecture are.
What is the pilot-to-production gap for AI agents?
67% of organisations report measurable gains from AI agent pilots, but only 10% successfully scale to production (Gartner, 2026). The 57-point delta lives in hallucination and reasoning failures that only emerge at real-world data volume and edge-case diversity, plus governance gaps that block scale even when the technology works.
Will half of AI agent deployments really fail by 2030?
Yes, according to Gartner's 2026 Data and Analytics Predictions. The cited failure modes are governance gaps and broken interoperability between systems, not model quality. The implication is that the buying decision in 2026 should weight governance and integration over benchmark performance.
Is MCP the solution to AI agent governance?
MCP is necessary but not sufficient. When properly implemented, MCP lets agents inherit existing authentication, authorization and policy controls rather than bypassing them. But Help Net Security reported in May 2026 that 1 in 4 MCP servers opens agents to code-execution risk. MCP is in the brittle phase where adoption has outpaced governance maturity.
What separates the 10% of agents that scale from the 90% that fail?
Three factors. One, a single named owner for AI governance accountable to the CFO, not a shared committee. Two, retrieval-augmented architecture so the agent acts on verified context, not free-form generation. Three, asynchronous human-on-exception design rather than synchronous human-in-every-step, which preserves scale while keeping the brand and compliance perimeter intact.
GEO for B2B SaaS: how to get cited by ChatGPT in 2026
From our research →
What is Generative Engine Optimization (GEO)?
GEO is the practice of structuring your brand's content and technical infrastructure so that AI engines like ChatGPT, Gemini, Perplexity and Claude cite and recommend you in their generated answers. Where SEO optimises for ranking in a list of blue links, GEO optimises for inclusion in a single synthesised answer.
How is GEO different from SEO and AEO?
SEO targets ranking in classic search engine result pages. AEO (Answer Engine Optimization) targets AI-powered search features like Google's AI Overviews and Bing Chat. GEO targets pure generative answer engines like ChatGPT and Claude that do not surface a SERP. The three overlap on technical fundamentals but diverge on signal weighting; GEO weights brand mentions and statistics more heavily, SEO weights backlinks and on-page keywords more heavily.
How do I get cited by ChatGPT?
Five things move the needle. One, publish original research, surveys and benchmarks with named statistics. Two, get quoted by named experts in your field. Three, structure content with self-contained 130 to 160 word answer capsules. Four, earn brand mentions across third-party sources (podcasts, communities, review sites). Five, keep content fresh; ChatGPT cites pages updated within 30 days at 76.4x the rate of older content.
Does GEO replace SEO for B2B SaaS?
Not yet, but the centre of gravity has shifted. Classic SEO still drives a meaningful share of high-intent search traffic, especially for category-defining keywords. But for early research, comparison and short-list questions, AI engines have replaced Google as the first stop for 51 percent of software buyers (G2, 2026). A B2B SaaS team that invests only in SEO is now competing for half the discovery market.
What is the ROI of GEO for B2B SaaS?
AI-referred visitors convert at 14.2 percent on average versus 2.8 percent for Google organic, because the buyer arrives pre-informed and recommendation-primed. The investment pays back at lower traffic volumes than classic SEO. Most B2B SaaS teams in our portfolio see GEO-attributable pipeline within 90 days of implementing the checklist.
Signal-based demand capture: 7 buying signals that make your outbound 4× more effective
From our research →
What is signal-based selling?
Signal-based selling triggers outreach the moment an observable buying signal fires (job change, funding round, hiring surge, tech stack switch, website intent), within hours, instead of blasting the entire TAM on a fixed calendar cadence. The result is dramatically higher reply rates because the timing and context match the buyer's actual situation.
Which buying signals convert best in B2B?
In our work with 18+ scale-up GTM rebuilds the 7 highest-leverage signals are: (1) past customer or champion job changes, (2) new executives in buyer roles, (3) funding rounds and M&A, (4) hiring surges, (5) tech stack changes and competitor churn, (6) de-anonymised website intent, and (7) community and dark social engagement. Past champion moves are the highest-converting single signal.
How fast do you need to act on a buying signal?
Speed is the lever. Aim for outreach within 4 hours for high-priority signals like past-champion moves and exec changes, within 24 hours for funding and hiring, and within 48 hours for tech stack changes. The first vendor to reach a buyer after a trigger event is roughly 5× more likely to win the deal.
Is signal-based outbound GDPR compliant in the EU?
Signal-based outbound is the most legally defensible form of B2B cold outreach in Europe. GDPR Article 6(1)(f) legitimate interest is much stronger when the contact's role makes business communication reasonably expected and the message is genuinely relevant — and a fired signal is direct evidence of relevance. Document your Legitimate Interest Assessment, comply with Article 14 source disclosure, and consult your DPO.
What tools do I need for signal-based selling?
Minimum stack: UserGems or LinkedIn Sales Navigator alerts for job changes and exec moves; Crunchbase, Specter or Signalbase for funding and M&A; Common Room, RB2B or Warmly for first-party and community signals; BuiltWith or HG Insights for tech stack signals. For EU coverage, prioritise GDPR-compliant providers like Cognism, Apollo (with EU data residency) and Salesflare.
Demand creation vs demand capture: why most B2B scale-ups have the ratio wrong
From our research →
What is the difference between demand creation and demand capture?
Demand creation builds future buying intent through brand, point-of-view content, founder-led social, podcasts and community. Demand capture intercepts buyers who already know they have a problem through paid search, retargeting, outbound to handraisers, and review-site presence. Both matter; the issue is the ratio between them.
What is the right demand creation vs demand capture ratio for B2B?
LinkedIn B2B Institute and Ehrenberg-Bass research, applied through Les Binet and Peter Field, points to roughly 60% creation and 40% capture for optimal long-term B2B growth. Most scale-ups today run closer to 10/90 the other way, which is the structural cause of CAC inflation.
What is the 95:5 rule and why does it matter?
Professor John Dawes of the Ehrenberg-Bass Institute showed that at any given moment roughly 95% of B2B buyers are not in the market. Only 5% are buying now. Capture-only marketing competes for the 5%; demand creation builds memory in the 95% so you are on their shortlist when they enter the market.
How do I measure demand creation if attribution software cannot track it?
Use a hybrid scorecard. Capture metrics: software-attributed pipeline, MQLs, meeting volume, CAC. Creation metrics: branded search volume, podcast downloads, founder-content engagement, share of voice, self-reported attribution from won deals, and win rate when buyers cite a creation touchpoint.
Where should an early-stage B2B SaaS company invest first?
Pre-product-market-fit, capture should dominate because closed-won this quarter matters more than memory built for next year. Once you cross €5M ARR with 100+ customers, the broken ratio becomes the binding constraint and creation needs a named, funded engine alongside capture.
AI SDRs vs human SDRs: 7 conditions where AI loses
From our research →
Do AI SDRs actually work?
Sometimes. In our benchmark of 11 Belgian and Dutch B2B deployments, 4 produced positive ROI within 90 days and 7 did not. The four that worked all shared three traits: an offer with a clear cost-of-inaction, a TAM above 5,000 accounts, and an existing brand presence in dark social. The seven that failed shared the inverse.
What is the difference between an AI SDR and an AI BDR?
Functionally none. Both terms describe an AI-driven outbound system that prospects, drafts and (in some configurations) sends outbound messages. SDR (sales development rep) and BDR (business development rep) are interchangeable in most North American and European usage.
Why do AI SDRs fail in Germany?
Germany's UWG and BGH case law impose stricter B2B consent and transparency requirements than most member states. The combination of inbox-saturation, AI-disclosure obligations under the EU AI Act Article 50 (effective August 2026), and German enforcement makes AI SDR deployments structurally harder in DACH than in BENELUX or France.
When should I keep my human SDR team instead of switching to AI?
Keep human SDRs when your ACV is above €100K, your buying group has 10+ stakeholders, your TAM is below 2,000 accounts, or your offer requires research and relationship work that no agent can compress. Multi-threading enterprise deals is a human-relationship problem, not a throughput problem.
How do I run an AI SDR diagnostic on my pipeline in 30 days?
Pick one ICP segment of 500–2,000 accounts. Deploy the AI SDR for 30 days with weekly review. Measure cost per qualified meeting, cycle time from signal to first reply, and percentage of conversations escalated to a human. If all three are improving by day 30, scale. If two are flat or worse, kill or rebuild.
The outbound agency cost autopsy: why software wins in 18 months
From our research →
How much does an outbound agency cost in 2026?
Outbound agencies in Europe charge €8K–€25K per month for retainer engagements, €350–€800 per booked meeting on per-meeting models, and €1,200–€2,800 per qualified opportunity on per-SQL models. Across 12 anonymised B2B scale-ups in our data, median total cost was €18K per month for 4 qualified meetings.
Are outbound agencies still worth it in 2026?
For most mid-market B2B SaaS scale-ups with €3M–€20M ARR, the unit economics no longer work. The agency model is structurally squeezed by sub-1% reply rates, 75% buyer preference for rep-free buying, and the deflationary pressure of AI tooling on the per-message cost. Some specialised use cases (enterprise outbound to named accounts, complex compliance markets) still favour an agency.
What is a typical AI SDR ROI?
In our 12-company benchmark, AI-platform-driven outbound delivered a median cost per qualified meeting of €380 versus €4,500 for agency-managed motions; a 12× reduction. Time to first booked meeting dropped from 6 weeks (agency onboarding plus ramp) to 11 days (platform deployment).
When should I replace my outbound agency with software?
When three signals fire: reply rates below 1% for two consecutive quarters, agency cost per qualified meeting above €2,500, and an internal team capable of running an AI GTM platform with 20% of one FTE's time. Below those thresholds the agency may still be the right answer; above them, the migration pays back within 18 months.
What does a 90-day agency-to-platform migration look like?
Weeks 1–2: data and CRM audit. Weeks 3–6: signal sources migrated and enriched. Weeks 7–10: workflows rebuilt on the platform. Weeks 11–12: agency contract exit and internal handover. By month 4, the platform is responsible for the full outbound motion under internal supervision.
GTM engineering vs growth marketing: the 2026 CMO guide
From our research →
What does a GTM engineer do day-to-day?
A GTM engineer owns the data, signals, agents and workflows that connect marketing, sales and product. Day-to-day this looks like building enrichment in Clay, automating signal capture from product usage and dark social, deploying AI agents for outreach, and instrumenting attribution as code rather than as dashboards.
What is the salary range for a GTM engineer in 2026?
In Europe, GTM engineers earn €75K–€140K base depending on stack ownership and seniority; in the US the range is $130K–$220K base. The role compounds value because one engineer typically replaces the workload of 1.4 marketing roles plus a junior SDR.
Do small B2B teams need a GTM engineer?
Below €2M ARR most teams don't need a dedicated GTM engineer. They need a marketer who is fluent in Clay, n8n and one autonomous platform. From €5M ARR upwards the dedicated role typically pays back within two quarters.
Is GTM engineering just a Clay marketing campaign?
Clay accelerated the term but did not invent the discipline. The role exists because of three independent forces: AI agents collapsing software cost, the death of MQL-based pipeline, and dark-funnel buying behaviour. Clay is one tool inside a much larger stack.
Will the GTM engineer role survive the AI agent wave?
The role survives precisely because of the AI agent wave. Agents need a system designer; someone who decides what signals matter, which workflows trigger which agent, and how human intervention is routed. That is what a GTM engineer does.
The Autonomous GTM Maturity Model: from copilot to self-driving
From our research →
What is the difference between an AI copilot and an AI agent?
A copilot suggests; an agent acts. A copilot generates a draft email for a human to send. An agent executes a multi-step workflow under defined rules and only escalates to a human on flagged exceptions. Most B2B GTM tools that claim 'agent' functionality in 2026 are still copilots.
What level is HubSpot Breeze, Salesforce Agentforce or Apollo's AI SDR?
On the Falora maturity scale, HubSpot Breeze, Salesforce Agentforce and Apollo's AI SDR sit at Level 2 (partial autonomy: agent executes single tasks under human supervision). They generate, draft and act on individual atomic steps but do not orchestrate end-to-end motions.
Is Level 5 self-driving GTM legal under the EU AI Act?
Level 5 self-optimisation involving solely-automated decisions about individuals likely falls under Article 22 GDPR (right not to be subject to fully automated decisions with legal or similarly significant effect) and may trigger Annex III high-risk classification under the EU AI Act. In practice, EU-deployed Level 5 systems require meaningful human oversight at the strategic layer.
How do I move my GTM stack up one maturity level?
Three changes consistently move teams up a level: (1) consolidate tooling so one orchestrator owns workflow state; (2) move from synchronous human-in-every-step to asynchronous human-on-exceptions; (3) instrument every action with feedback data so the agent can be evaluated and re-trained.
Why does the maturity model matter if I just want pipeline?
Because vendors price at one level and deliver at another. Knowing your current level; and the realistic next level for your stack; is the difference between a productive 12-month rebuild and a Frankenstack of half-finished automations that produce no measurable lift.
The anatomy of a GTM engineering system: 7-layer reference
From our research →
What is a GTM engineering system?
A GTM engineering system is the integrated stack of seven layers; signal, identity, reasoning, orchestration, activation, measurement and governance; that turns market and product data into qualified revenue conversations under autonomous or semi-autonomous control. It is the architectural counterpart to the GTM engineer role.
What does a modern B2B GTM tech stack look like?
A modern B2B GTM tech stack composes seven layers: a signal layer (Common Room, 6sense, RB2B), an identity layer (Clay, Cognism, Apollo), a reasoning layer (Claude, GPT, in-house models), an orchestration layer (Clay, n8n, Default), an activation layer (HubSpot, Lemlist, Outreach), a measurement layer (Dreamdata, HockeyStack), and a governance layer (compliance, audit, EU AI Act conformance).
How many tools should a B2B GTM stack have in 2026?
Top-quartile autonomous GTM teams in our portfolio use 8 tools across the seven layers; the median scale-up uses 23. The number that matters is not how many tools you have, but how many of them an orchestrator can read and write to in a single workflow.
What is a Frankenstack and how do I avoid it?
A Frankenstack is a stack of 20+ point tools, none of which talk to the others, all of which produce dashboards nobody reads. Avoid it by choosing one orchestrator that owns workflow state and treating every other tool as a service the orchestrator calls. Consolidation typically reduces tool count by 50–70% in our rebuild engagements.
Where does Falora fit in the 7-layer architecture?
Falora sits primarily in the orchestration and governance layers, with native integrations into the signal, identity, reasoning, activation and measurement layers. Falora's role is to be the single orchestrator that reads from your existing best-of-breed tools, executes the autonomous workflow, and exposes the audit trail for EU AI Act compliance.
The EU AI Act + your AI GTM stack: a 2027 audit map
From our research →
Is using an AI SDR illegal in Germany?
Not categorically illegal, but Germany's UWG and BGH case law impose stricter B2B consent and transparency requirements than most member states, and the EU AI Act Article 50 disclosure obligation (in force August 2026) requires recipients to know they are interacting with AI. Combined, this makes many current AI SDR deployments non-compliant in Germany without explicit AI disclosure and B2B legitimate-interest documentation.
Do I need to disclose AI in cold emails?
From August 2026, EU AI Act Article 50 requires that AI-generated content (including text generated and sent by AI agents) be detectable as AI-generated, with limited exceptions. For B2B cold email this practically means including an AI-disclosure line and ensuring the technical means of detection (e.g. watermarking metadata) are in place.
Is lead scoring a high-risk AI system under the EU AI Act?
Pure lead-scoring on firmographic and behavioural signals is generally not Annex III high-risk. It can become high-risk if it materially determines access to a service or has legal/significant effect on the individual (e.g. denying credit, employment, or essential services). The boundary is set by the deployer's purpose, not by the technology itself.
What are the EU AI Act fines?
Article 99 of the EU AI Act sets fines up to €35M or 7% of global annual turnover for prohibited-AI violations, up to €15M or 3% for high-risk and transparency violations, and up to €7.5M or 1% for incorrect, incomplete or misleading information to authorities. National competent authorities apply these fines proportionally.
When does the EU AI Act apply to my AI GTM tools?
Prohibited AI practices have been in force since February 2025. AI literacy obligations (Article 4) since the same date. General-Purpose AI obligations since August 2025. Transparency obligations including AI-content and chatbot disclosure (Article 50) from August 2026. Annex III high-risk obligations from August 2026, with the Digital Omnibus likely shifting some categories to December 2027.