Your revenue system leaks where your tools meet.
Every tool in your stack works. The handoffs between them don't — so people become the integration layer. I find where that's costing you most, then build the smallest system that stops it.
Research by handSomeone opens LinkedIn and the company site to decide if a lead is worth working.
Assignment by SlackOwnership is decided in a thread, not a rule. Leads sit while people are asleep.
Reports rebuilt weeklySomeone exports, pivots, and pastes the same numbers into the same deck.
Four problems that knock over everything downstream.
These aren't independent. Each one causes the next. Fix them in the wrong order and you spend budget without moving a number.
Leads go cold before anyone touches them
A form fills at 11pm. It's enriched by hand the next morning, assigned after standup, and worked in the afternoon. The buyer has already booked with whoever replied first.
- Conversion rate drops, so CAC rises
- Paid budget is judged on a broken funnel
- Reps blame marketing for lead quality
Nobody fully trusts the CRM
Duplicate accounts, blank fields, stages that mean different things to different reps. So people keep private spreadsheets, and the real pipeline lives outside the system of record.
- Forecast becomes an opinion, not a number
- Automation built on this data misfires
- Every meeting starts by debating whose report is right
Your closers are doing data entry
Call notes, field updates, follow-up scheduling, list building. An hour here, an hour there — until the person you hired to sell is spending half their week not selling.
- Effective headcount is far below actual headcount
- Good reps leave; admin isn't the job they took
- You hire to fix capacity you already paid for
No single source of truth across teams
Marketing reports one number, sales another, finance a third. Each is defended, none reconcile, and the discrepancy is rediscovered every quarter.
- Decisions get made on the loudest number
- Board and investor reporting becomes stressful
- Nobody can prove what actually worked
And underneath, the constant friction
None of these will sink you on their own. Together they eat a day a week, every week, and nobody logs the hours.
The problems above are symptoms. These are the causes.
Treat the symptoms and they return in a new form within two quarters. This is what actually has to change.
The stack was bought, never designed
Each tool was added to solve one urgent problem — a CRM here, an enrichment tool there, a scheduler, a dashboard. Every tool works. Nobody ever designed how work moves between them, so the seams became human. Your team is the integration layer.
Process lives in people's heads
How a lead gets qualified, who owns what, when a deal moves stage — it's known, not written, and definitely not encoded. That works until someone is on leave, until you hire, until volume doubles. Then it doesn't.
Nobody owns the whole pipe
Marketing owns the top, sales the middle, customer success the end. Each optimises their section. The losses happen at the boundaries, which is exactly where nobody's targets are measured — so the leaks stay invisible and unfunded.
Automation was attempted tool-first
The step that's easiest to automate gets automated. Not the step that's costing the most. You end up with a dozen small workflows that save minutes each, while the expensive break sits untouched because it's harder to reach.
Nothing was measured before it was changed
Without a baseline, improvement is invisible. You can't prove the automation worked, so it never gets extended, and the next budget conversation starts from zero. Measurement isn't reporting overhead — it's what makes the work defensible.
The cost compounds quietly, and never appears as a line item.
None of this shows up in a P&L as "revenue operations debt." It shows up as targets missed for reasons nobody can quite name.
Revenue leaks you never see
Lost deals don't get logged as "we replied too late." They get logged as lost to a competitor, or not logged at all.
A forecast you can't defend
When the underlying data is soft, every number you present is one question away from falling apart in the room.
Scaling multiplies the mess
Hiring into an undesigned process doesn't fix it. It adds people whose job is to absorb the friction manually.
Your best people leave
Strong operators and closers don't stay in roles that are 50% admin. The ones who tolerate it usually aren't the ones you want to keep.
Spend optimised against bad data
If attribution is broken, you scale the channel that reports well rather than the one that actually converts.
The founder becomes the bottleneck
When the process isn't encoded, the person who remembers how it works gets pulled into every exception. That's usually you.
What most teams try first — and why it doesn't hold.
These are reasonable instincts. They're also the reason the same problems come back next quarter.
"We need more leads."
You're losing a meaningful share of the leads you already pay for — to delay, bad routing, and follow-up that never happened. More volume into a leaking pipe raises spend, not revenue.
"AI will sort this out."
AI applied to an undefined process automates the mess faster and with more confidence. The model isn't the hard part — deciding what should happen, and when, is.
"We'll clean up the CRM later."
Data rot compounds. Every week of delay adds records to fix and decisions made on bad numbers. "Later" arrives as a migration project nobody has time for.
"We're too small for RevOps."
Small teams pay the highest tax per person, because the manual work is spread across people who have no slack. The cost is real; it's just distributed enough to stay invisible.
"We just need the right tool."
Adding a tool adds two more seams. Tools are rarely the constraint — the undesigned flow between them is. New software on top of that makes the diagram more complicated, not the work less manual.
"We'll hire an ops person to own it."
That hire inherits an undesigned system with no documentation. Without a redesign first, you've hired a very expensive human API — and created a single point of failure.
The everyday versions of the same mistake
Automating before auditing. Building the workflow that's obvious rather than the one that's expensive.
Measuring activity instead of outcomes. Counting emails sent tells you nothing about whether the pipe is leaking.
Undocumented automations. They break silently, months later, and nobody knows they were ever running.
One person holding every workflow. When they leave, the system leaves with them.
No baseline before "improving." If you didn't measure it first, you can't prove it worked or justify doing more.
Copying someone else's stack. Their constraints aren't yours. A template built for a 200-person team breaks a 15-person one.
Diagnose first. Then fix the cause, not the symptom.
I don't sell automations. I find where your revenue system is losing the most, and build the smallest thing that stops it.
That means the first engagement is diagnostic, the recommendation might be "don't automate this yet," and nothing gets built until we both know what it's worth. When the root cause is fixed, most of the symptoms above resolve without being addressed individually — which is the point.
Diagnosis before prescription
Every engagement starts by mapping how work actually moves through your stack — not how the org chart says it should. The expensive break is usually not the one people complain about.
Built for your constraints
Your team size, your tools, your budget, your data quality, who'll own it after I leave. No templates. A system your team can't maintain is a liability, not an asset.
Measured, or it didn't happen
Baseline before the build, measurement after. You get a number you can defend, and a clear basis for deciding what to do next.
Six stages. The order is the method.
Building before diagnosing is how teams end up with automations that solve the wrong problem elegantly.
Map
Every tool, every handoff, every place a human moves data between systems. Short interviews with the people actually doing the work — because the documented process and the real one are rarely the same.
- Full stack and data-flow map
- Every manual handoff, listed
- Where the process differs from the documentation
Quantify
Put hours and money against each break. How long does this take, how often, done by whom, at what cost. This is the baseline — without it, nothing that follows can be proven.
- Cost per break, in hours and currency
- Baseline metrics, recorded
- Ranked list by cost, not by ease
Prioritise
Which single fix knocks over the most downstream problems. Sequenced against your real constraints — team capacity, data quality, budget, and what has to keep running while we work.
- A written roadmap, sequenced
- What to fix now, next, and never
- Honest scope: what isn't worth automating
Build
Built in your environment, not a sandbox. Failure modes handled deliberately — expired credentials, rate limits, empty states, bad input. An automation that works only on the happy path is a future incident.
- The working system, live
- Error handling and alerting
- An audit trail of what it did
Hand over
Your team has to be able to change it without me. Written documentation, a recorded walkthrough, and a live training session with whoever will own it day to day.
- Documentation: what it does, how to change it
- Recorded walkthrough
- Training session and a defined support window
Operate
Optional. Systems drift as your process changes — new fields, new stages, new tools. A retainer keeps things monitored and extends the system as the business moves.
- Monitoring and fixes
- New automations as needs emerge
- Results measured against the baseline
The systems that usually come out of this.
Each one replaces something a person on your team is doing by hand, on a schedule, right now.
Routing & enrichment
A form fires. The lead is enriched, scored, assigned by rule, and logged — before anyone opens their inbox. Nights and weekends included.
CRM hygiene agent
Dedupes, normalises, fills gaps and flags rot on a schedule — so the number you take to a board meeting is one you can defend.
Call transcript → CRM
Transcripts become structured notes, updated fields, and next steps, written back automatically. No post-call admin.
Pipeline & spend digests
Weekly pipeline summaries and ad-spend anomaly alerts, assembled and delivered without anyone rebuilding a deck.
I came to automation from GTM, not from software.
I'm an independent consultant providing revenue operations and go-to-market automation services internationally — primarily to B2B SaaS companies and marketing and growth agencies.
The work is auditing GTM and CRM systems, designing and building AI-powered workflow automation, and advising teams on adopting AI agents and no-code tooling to remove manual operational work.
Before automation, my background was paid media, revenue operations, and go-to-market strategy. That's the relevant part: I've owned the funnel metrics and defended them to stakeholders. The systems I build are shaped around how a pipeline behaves commercially, not just what a workflow tool can technically do.
Three ways to start.
Fixed scope, 50% upfront. Most engagements begin with an audit, then move to a build and a retainer.
GTM Stack Audit
Stages 01–03. Your stack mapped, every manual break costed, and a sequenced roadmap — with a walkthrough. Stands alone; you can take it and build in-house.
Start an auditAutomation Sprint
Audit included, then the top one to three fixes built, instrumented, documented, and handed over. Measured against the baseline set in week one.
Scope a sprintMonthly Retainer
Stage 06. Monitoring, maintenance, and new automations as your process changes. Available after a sprint, not before.
Discuss a retainerCase studies, as engagements close.
First engagements are currently underway. Walkthroughs and results will be published here as they're delivered — measured against the baseline, with real numbers rather than projections. If you'd like to see a system in progress before then, ask on a call.
Start with the diagnosis.
Thirty minutes. I'll ask how a lead moves through your business and where people are filling gaps by hand. If there's nothing worth fixing, I'll tell you that.