From founder hustle to a repeatable growth engine
Inside my engagement as strategic growth partner to CortexCart, an AI ecommerce analytics platform: how I diagnosed the real bottleneck, challenged the "we need more AI" assumption, and designed and built the operating system the business now runs on.
The transformation in one page
CortexCart is an AI-powered ecommerce analytics platform that unifies Shopify, WooCommerce, Meta and GA4 data into one dashboard and explains — in plain English — why a store's numbers are moving. The product was strong and differentiated. What surrounded it was typical of an early-stage, founder-led SaaS: growth activity that lived in the founder's head, ad-hoc outreach, and no single view of pipeline, marketing or finances.
Founder Jonathan S brought me in as strategic growth partner — not to "add AI", but to work out why growth wasn't compounding and fix it. I ran the discovery, diagnosed the bottleneck, and challenged the starting assumption that more AI was the answer. Then I designed and built what the diagnosis called for: CortexCart OS (an internal dashboard spanning sales, growth, finance and marketing), a repeatable outbound lead generation system with AI qualification and CRM workflow, a marketing asset library including a brand film and campaign kit — while advising Jonathan directly on positioning and go-to-market strategy.
The result: growth work that was previously manual and untracked now runs through one shared system, with documented processes and a staged roadmap. The engagement grew into an ongoing partnership.
What actually changed
The rest of this page explains how. This is the shape of the change itself.
Before
- Growth lived inside the founder's head
- No repeatable lead generation
- Four disconnected tools, no source of truth
- Outreach and follow-up run from memory
- Reactive, asset-by-asset marketing
- No shared view of pipeline or business health
After
- One centralised operating system, shared live
- Repeatable, signal-driven lead engine
- Structured 7-stage CRM workflow
- AI-assisted research, scoring and outreach drafting
- Positioning backbone and a reusable asset library
- Shared operational visibility and decision support
A differentiated product at the start of its commercial journey
CortexCart, founded by Jonathan S, gives ecommerce operators the analytics capability of an enterprise data team at a price an independent store owner can justify. It answers the question dashboards usually leave open: why did the numbers move?
One dashboard, plain-English answers
Unifies Shopify, WooCommerce, Meta and GA4 into a single view, with AI explanations of what changed and what to do about it — including revenue arriving from AI assistants like ChatGPT and Perplexity, a channel most analytics tools can't see.
A crowded, expensive category
Incumbent ecommerce analytics tools price for scale-ups, not independents. CortexCart's opportunity: growing store owners who need answers, not another dashboard subscription that costs more than their apps combined.
Founder-led and early
Strong engineering, real product differentiation, early commercial traction — and a founder personally carrying product, sales and marketing at the same time. The classic early-SaaS constraint: not ideas, capacity.
The product wasn't the bottleneck. The operation around it was.
My discovery work surfaced a consistent pattern: every growth activity worked when Jonathan did it personally, and stopped when he didn't. Nothing was broken — but nothing was repeatable either.
Objectives I agreed with Jonathan
Repeatable lead generation
A system — not a heroic effort — that finds, qualifies and prioritises the right prospects every week.
Structured sales workflow
Defined pipeline stages, a CRM that reflects reality, and outreach that starts from research rather than templates.
Operational visibility
One place where sales, growth, marketing and finance can be seen and questioned — by the founder, in plain English.
Marketing foundations
Positioning, messaging and a reusable asset library that make every future campaign faster to produce.
Sharper go-to-market
A clear ICP, a lead-with-strength narrative, and acquisition channels chosen deliberately rather than opportunistically.
Founder leverage
Move the founder's time up the value chain: from doing the growth work to directing it.
I didn't start by building. I started by understanding.
The first phase of the engagement produced no software at all. It produced a diagnosis — and the diagnosis is what made everything built afterwards worth building. Eight strands of discovery work, run before a single line of code:
Founder interviews
Working sessions with Jonathan on where his week actually goes. Revealed: growth activity stopped every time product work spiked.
Business interviews
Commercial deep-dive: pricing, ICP, competitors, sales conversations to date. Revealed: the strongest differentiator barely featured in the messaging.
Current-systems audit
Inventoried every tool the business ran on. Revealed: four disconnected tools, and none of them the source of truth.
Workflow mapping
Traced the prospect-to-customer journey end to end. Revealed: every path ran through the founder personally.
Process mapping
Separated the repeatable steps from the judgement calls in each workflow. Became: the blueprint for what to automate and what to keep human.
Pain-point identification
Ranked the friction Jonathan felt against what the business was actually paying for. Revealed: they weren't the same list.
Bottleneck analysis
Followed the constraint through the whole operation. Revealed: the bottleneck wasn't leads, product or budget — it was founder hours.
Opportunity prioritisation
Scored every opportunity by impact on that constraint. Became: the build order — positioning first, then the lead engine, then the operating system.
The diagnosis: CortexCart didn't have an AI problem. It had a clarity problem — and the founder was the system.
His deep expertise in data analytics allowed us to pinpoint specific areas of lost revenue, providing an excellent complement to our existing business model.
Operations first. Systems second. AI where it earns its place.
My method is deliberately unglamorous: understand how the business actually works before building anything, prototype fast, and only automate a process once I've mapped and simplified it. I apply AI where judgement is needed — not sprinkled everywhere.
The methodology: diagnose before designing, prototype before building, measure after deploying.
I started with the business, not the tech stack
The eight discovery strands above, run as working sessions with Jonathan. The goal: find the constraint, not confirm a solution I'd already decided on.
- Founder interviews
- Product deep-dive
- ICP & market review
I mapped how growth work really happened
I traced the full prospect-to-customer journey and the operational workflows behind it — where information lived, where handoffs failed, and which tasks consumed founder time without needing founder judgement.
- Workflow mapping
- Tooling audit
- Time & effort analysis
I prioritised by leverage, not novelty
I ranked opportunities by impact on the constraint — and challenged the instinct to automate first. The order that survived scrutiny: sharpen positioning around CortexCart's genuinely unique capability, systematise outbound lead generation, then centralise internal operations.
- Opportunity scoring
- Positioning review
- Roadmap sequencing
I designed the operating system around the workflow
I architected CortexCart OS and the lead engine as one connected system: deterministic scoring where rules are known, AI where qualification needs judgement, and a CRM that mirrors the real sales conversation.
- System architecture
- Scoring model design
- Data model & pipeline stages
I put working software in front of Jonathan, not slide decks
Working prototypes in days, reviewed together against real prospect data. Features earned their place by changing behaviour; anything that didn't, I cut before it accumulated cost.
- Days-not-months iterations
- Real-data testing
- Founder feedback loops
I deployed it, loaded real data, and handed it over working
I deployed CortexCart OS to cloud hosting with a shared database, so Jonathan and I work from one live system. It launched with 29 researched, signal-scored prospects — not an empty shell.
- Cloud deployment
- Real leads loaded
- Excel import for existing lists
I stayed to operate, measure and extend
The engagement continues as an operating partnership: I refine scoring signals, extend data sources and build the next acquisition channels on top of a documented, staged roadmap.
- Documented processes
- Staged V2 roadmap
- Ongoing strategy sessions
CortexCart OS — the company's growth operating system
I designed and built an internal dashboard that became the single place where the business runs: pipeline, prospect intelligence, outreach, marketing, finance and strategy — organised into departments, searchable in plain English, shared live between Jonathan and me.
One system replaced four disconnected tools
Before CortexCart OS, prospect lists, outreach drafts, deal notes and finance figures lived in separate spreadsheets and apps. Now a single cloud-hosted system holds all of it — every module reads and writes the same data, so the pipeline the founder sees is the pipeline that actually exists.
The twelve modules
Lead Pipeline
Every prospect from first discovery to closed, in one flow — no more list-in-a-spreadsheet, notes-in-an-inbox.
AI Prospect Scoring
Deterministic buying-signal scoring plus AI fit analysis: likelihood to buy, best sales angle and likely objections per lead.
Built-in CRM
Stages from contacted to won/lost with hot/warm/cold priority tags — a CRM shaped around the real sales conversation.
Sales Dashboard
Deal flow, stage conversion and activity at a glance — what's moving, what's stuck and what needs a follow-up today.
Growth Dashboard
Campaign activity, lead volume and channel performance in one view, so growth decisions start from data.
Finance Dashboard
Business-health figures alongside the pipeline that drives them — one login for the numbers that matter.
Outreach Draft Generator
Personalised first-touch drafts in CortexCart's brand voice, grounded in each prospect's research — reviewed by a human before anything sends.
Company Intelligence
Each prospect's store platform, tech stack, marketing signals and research notes in one profile — context before contact.
AI Search
Ask questions across the whole prospect base and get ranked answers — search that understands intent, not just keywords.
Natural-Language Querying
"Show me hot UK leads running paid ads" — plain English in, a filtered, prioritised list out. No query language to learn.
Workflow Visibility
Every lead's stage, owner and next action is visible — so nothing depends on remembering, and nothing falls through.
Strategic Reporting
Decision-ready summaries for founder strategy sessions — what happened, why, and what to do next week.
System architecture
CortexCart OS data flow — deterministic rules handle the known, Claude handles the judgement, humans make the calls.
The Lead Generation System
I designed the complete outbound engine around one principle: research before outreach. Every message a prospect receives starts from what the system actually learned about their store — not a template with their first name pasted in.
Discover
Shopify stores matching the ICP
Enrich
Platform, tech stack, marketing footprint
Detect signals
Paid ads, email stack, analytics gap, hiring
Qualify
AI fit analysis & scoring
Prioritise
Hot / warm / cold ranking
Draft
Personalised outreach, human-reviewed
Manage
CRM stages through to won
A message worth systematising
Systems amplify whatever message you feed them — so I did the positioning work first. CortexCart's most defensible story: it sees revenue arriving from AI assistants that competitor analytics tools can't track, and it explains every number in plain English.
Brand film & campaign kit
A cinematic brand film built around the "Stop guessing. Start growing." narrative, blending real product footage with AI-generated cinematography — delivered in 16:9 and 9:16 for web and social, with a campaign brief and ready-to-send launch email.
- Hero film in two formats, walkthrough-ready
- Campaign brief with rollout plan
- Reusable production pipeline for future videos
Positioning & founder messaging
Sharpened the story from "another analytics dashboard" to a differentiated claim: enterprise-grade answers at an independent-store price, including the AI-referral revenue channel competitors can't see.
- Sales messaging & objection-ready talking points
- Founder voice for LinkedIn and outreach
- Educational content angles mapped to ICP pains
Channel strategy
Deliberate channel choices instead of scattered effort: outbound powered by the lead engine, LinkedIn as the founder's authority channel, and education-led acquisition through product walkthroughs.
- LinkedIn content strategy
- Outbound email sequencing
- Product walkthrough & demo assets
Webinar strategy
Designed a repeatable education-led channel: teach store owners to find lost revenue in their own data, and let the product demonstrate itself — a framework ready to launch as the next acquisition experiment.
- Webinar framework & topic map
- Registration-to-demo journey design
- Content repurposing plan per session
Advising the founder, not just building for him
Alongside the builds, I run a standing strategy partnership with Jonathan: regular working sessions where decisions about growth, product and operations get made with data on the table — and where my job includes challenging assumptions, not just agreeing with them.
Go-to-market & growth strategy
ICP definition, channel prioritisation and a sequenced growth plan — matched to the capacity of a founder-led team rather than a fantasy org chart.
Product strategy & positioning
Which capabilities to lead with, how to price against incumbents, and where the product roadmap and the sales narrative reinforce each other.
AI opportunities & operational improvement
A pragmatic map of where AI genuinely helps the business — and where a simpler process fix beats a model. Capacity first, complexity last.
What changed
No invented revenue figures — the honest measure of this engagement is structural: work that used to depend on the founder's memory and hustle now runs through documented, shared systems.
Operational visibility increased
The founder can see pipeline, marketing and finances in one place — and query it in plain English.
Lead generation became repeatable
Launched with 29 researched, signal-scored prospects; the same engine keeps producing without heroics.
Sales workflow became structured
Defined stages and priority tags replaced memory-driven follow-up.
Marketing foundations established
Brand film, campaign kit, messaging and channel strategy — assets that compound instead of expiring.
Founder gained decision support
Strategy sessions now start from dashboards and lead intelligence, not gut feel.
Growth foundations created
Documented processes and a staged roadmap mean the next hire — or the next system — has something to stand on.
I highly recommend Charlie Marshall to any business looking to accelerate growth through AI and operational systems. Charlie has been instrumental in helping shape the strategy behind CortexCart Insight Dashboard, developing practical automation ideas, marketing initiatives, and growth systems that complement our platform. His ability to identify opportunities, simplify complex operational challenges, and translate them into actionable solutions has made him a valuable strategic partner. I would confidently recommend Charlie to any business looking to improve efficiency, unlock growth, and modernise their operations.
Most businesses think they need AI. They don't — not first.
They need clarity. Then process. Then systems. Then automation — in that order. This engagement worked because I ran that order strictly, even when jumping straight to the shiny part would have been easier to sell.
Clarity
See the business as it actually runs — not as the org chart claims
Process
Simplify and document the workflow before touching it
Systems
One shared source of truth the whole business works from
Automation
AI only where judgement adds value a rule can't
If I had started with AI, I would have automated a broken workflow — faster chaos, at a monthly cost. Because I started with clarity, every system that followed was built on a process that had already been mapped, simplified and agreed with the founder. That's why the systems stuck: they matched how the business actually works, instead of asking the business to work like the software.
What this engagement proved
Start with the operation, not the AI
The biggest gains came from mapping and simplifying workflows before automating them. AI earned its place where judgement was needed — qualification, drafting, search — and deterministic rules handled everything else, cheaper and more predictably.
Decide with working software
Prototypes reviewed against real prospect data settled in days what slide decks would have debated for weeks. Features that didn't change behaviour were cut early.
One shared system beats four good tools
Consolidation was worth more than any individual feature. When founder and partner work from the same live data, alignment stops being a meeting and becomes a property of the system.
Founder time is the scarcest input — design for leverage
Every build was judged by one test: does this move the founder from doing the work to directing it? Systems that create capacity compound; systems that create admin don't survive.
Ship the smallest system that changes behaviour
A staged roadmap beat a big-bang rebuild. Each stage went live, proved itself in use, and funded confidence in the next — the same discipline now written into the V2 plan.
Where the partnership goes next
Deeper data sources
Paid store-intelligence feeds to widen discovery beyond the current research pipeline.
Webinar channel launch
Taking the education-led acquisition framework from strategy to a running channel.
Expanded automation
Follow-up sequencing and reporting automation on top of the proven CRM workflow.
Content engine
Systematising LinkedIn and educational content production around the positioning backbone.
Could your business benefit from a growth system?
Three honest questions:
- ✓Is growth dependent on one person?
- ✓Are repetitive tasks slowing your team down?
- ✓Are your sales and operational systems disconnected?
If you answered yes to any of these, a free Growth Systems Audit will show you — honestly — what's worth building and what isn't.
No pitch deck. No obligation. If there's no leak worth fixing, you'll hear that too.