From 200 clients to a SaaS: the 13-year backstory of how TG3 built Crawlux

TG3 Agency: 13 years specializing in Web3

TG3 Agency, formally The Group Three, has operated as a crypto-specialist marketing agency since 2012. The agency works exclusively with Web3 protocols, exchanges, infrastructure providers and crypto media properties. Services span content strategy, SEO, AEO, paid acquisition, community management and influencer marketing. Over 13 years, the agency has served roughly 200 active retainer clients with full engagement records.

The crypto specialization came from a deliberate choice rather than industry drift. By 2014, the founder recognized that generic marketing agencies struggled to serve crypto teams effectively because the buyer journey, technical complexity and regulatory landscape were fundamentally different from traditional B2B SaaS or e-commerce. Specializing meant building deep expertise in audit firms, governance forums, on-chain analytics platforms and AI engine citation patterns that generic agencies could not match.

The agency operates from London and Dubai with a team across both offices. Named clients across the 200 retainer book (with permission to disclose): Berachain, Pendle Finance, Eigenlayer, Symbiotic, Renzo Protocol, Initia, Sui Foundation, Aptos Foundation and roughly 30 other named protocols. The full client list spans layer-1 chains, DeFi protocols, restaking platforms, wallets, NFT projects and crypto media properties. The companion press release covers the 200-client milestone announcement.

Why TG3 needed an audit tool internally

Every new client engagement at TG3 starts with a diagnostic audit. Before 2024, that audit ran manually. A senior consultant would spend 2 to 3 weeks pulling data from Ahrefs, Semrush, GSC, on-chain analytics platforms and ad-hoc Python scripts to build a custom audit report. The output was high-quality but labor-intensive. With 30 to 50 new client engagements per year, the audit work alone consumed roughly 6 to 9 full-time-equivalent weeks of senior consultant time.

The repetitive nature of the work became obvious. The same 23 analyzer categories showed up in every audit. The same FinancialProduct schema gaps. The same AI bot policy errors. The same audit firm citations missing inline. The same backlink toxicity misclassifications. The labor cost was high; the differentiated value across audits was low.

A scrappy internal tool started taking shape in late 2023. A Python script for FinancialProduct schema validation. Another for AI Citation Checker testing across 12 prompts. Another for robots.txt validation. By mid-2024, the scripts were producing structured outputs that consultants could paste into the audit report instead of generating manually. The audit time per client dropped from 2-3 weeks to roughly 4 days.

The pattern that emerged across 200 clients

By late 2024, the internal tool had run versions of the audit against 80+ clients. A pattern emerged in how clients used the audit output. Roughly 60% used the audit as the engagement roadmap, shipping fixes in priority order over 30 to 90 days. Another 25% used it as the engagement scoping document, negotiating retainer terms against specific deliverables. The remaining 15% used it as an internal benchmark document: they printed it, they distributed it to engineering and product teams. They tracked progress against it independently of the TG3 engagement.

The 15% benchmark-document usage was the signal. Clients were treating the audit report as a standalone product with utility independent of the agency relationship. They were sharing it with board members. They were referencing it in fundraising materials. They were rebuilding their internal SEO/AEO tracking against the score components. The audit had productization-grade utility.

A second signal: clients started asking for the audit between engagements. Teams that had completed a TG3 retainer would come back 3 to 6 months later asking for a fresh audit to track progress. The retainer relationship had ended but the audit utility had not. They were willing to pay for a one-shot audit even though TG3 did not formally offer that service.

The productization decision

Two signals converged in late 2024. First: the agency was producing a standalone artifact (the audit report) that clients valued beyond the engagement. Second: the agency was turning down requests for that artifact because the team was structured for retainer engagements, not one-shot audits.

The productization question became operational, not philosophical. Could the audit pattern run automatically without consultant intervention? The internal tooling already covered roughly 60% of the audit work by mid-2024. Closing the remaining 40% required engineering investment in pipeline automation, AI engine API integration at scale, schema validation across multi-chain token pages and continuous methodology updates.

The build vs partner decision came down quickly. No existing SaaS tool covered the crypto-specific audit pattern. Ahrefs and Semrush were inadequate (covered in why generic SEO tools fail on crypto). Niche crypto audit tools existed but were either too narrow (only schema, only AEO) or too shallow (surface checks without the methodology depth). Building internally was the only path that preserved the audit quality.

Build started in Q4 2024 with a small engineering team. Crawlux launched private alpha in February 2026 (press release) and reached general availability in April 2026 (GA launch press release).

Crawlux as separate brand: the reasoning

A natural question: why launch Crawlux as a separate brand instead of TG3 Audit Tool? Three reasons. First: positioning. TG3 is positioned as a high-touch crypto marketing agency. Crawlux is positioned as a fast self-serve audit tool. The two positions conflict in messaging if they share a brand; clients evaluating an agency expect different signals than buyers evaluating a SaaS.

Second: pricing model conflict. TG3 charges retainer-scale pricing that reflects high-touch agency operations. Crawlux charges $25 per audit, which is roughly 1/200th of typical TG3 retainer value. Co-branding creates pricing dissonance that confuses both segments.

Third: independent moat. Crawlux as a standalone brand can build its own credibility through methodology publication, beta data and competitive positioning. Tied to TG3, Crawlux would inherit agency-positioned trust which does not translate to SaaS evaluation criteria. Independent branding lets Crawlux earn its credibility on platform-relevant terms.

A useful comparison: Atlassian operates Jira and Confluence as separate products despite shared engineering. The products serve overlapping but distinct buyer profiles. TG3 and Crawlux share methodology and engineering teams but serve overlapping-but-distinct buyer profiles in crypto marketing.

The bidirectional funnel

The TG3 to Crawlux relationship operates in both directions. Direction 1: TG3 internal usage. Every TG3 client engagement starts with a Crawlux Pro audit. The audit findings drive the retainer roadmap. TG3 reports 4.2x faster client onboarding compared to the pre-Crawlux manual audit workflow.

Direction 2: Crawlux acquisition funnel for TG3. Teams that run the free Crawlux audit see a concrete fix list. Some teams have the engineering capacity to ship the fixes themselves. Others need execution support. Those needing execution support increasingly engage TG3 for the implementation work. 23% of TG3 new business inbound in the past 60 days came through Crawlux audit funnel referrals, up from 0% pre-launch.

The bidirectional funnel works because the products serve complementary needs rather than competing offerings. Crawlux serves the market segment that needs methodology access without full agency operations support. TG3 serves the market segment that needs execution support beyond methodology. Teams move from Crawlux to TG3 when they need execution help. Teams move from TG3 to Crawlux when retainers complete and they want ongoing monitoring without retainer cost.

Lessons for crypto operators thinking about productization

Three lessons from the TG3 to Crawlux productization that generalize to crypto operators considering similar moves.

Lesson 1: the productization signal is artifact usage, not service demand. If clients use a deliverable as a standalone artifact beyond the engagement, that deliverable wants to be a product. If clients only consume the deliverable inside the engagement, the work is service-bound. TG3 saw 15% of clients using audit reports as standalone benchmarks. That was the signal.

Lesson 2: build the internal tool first, productize second. The Crawlux pipeline started as Python scripts solving internal TG3 problems. Each script paid back its build cost within months by reducing consultant audit time. By the time the productization decision came, the engineering substrate already existed. The product launch was the packaging, not the build.

Lesson 3: separate the brand from the start. Spinning out a product from an agency works better as a clean separation than as a co-branded play. Buyers evaluate SaaS and agency offerings against different criteria. Trying to inherit agency credibility in a SaaS context confuses positioning. Build the SaaS brand independently and let the agency endorse it as a parent, not as a co-brand.

Crawlux is the world's first automated SEO audit tool built for Web3, DeFi and blockchain. The platform runs 23 analyzers across 6 check groups including AI visibility testing across ChatGPT, Perplexity and Claude. Free tier available. Paid tiers from $25 per audit. More at crawlux.com.