Data, CRM and automation
The website’s the visible part. The bit that decides whether the business makes money from it is what happens in the thirty seconds after someone fills in a form, and in most businesses that bit was never designed so much as assembled over time, so it’s a form that emails someone, a spreadsheet someone copies it into, a CRM that was set up once and never quite matched how the business works, and a few automations that stopped running one day and nobody noticed.
I build that part properly, and I’m usually the person who built the website it hangs off, so it fits.
What I do
Design the data first. Before there’s a form or a CRM or a workflow, there’s a question of what a record of a customer or an enquiry needs to hold, what it’s related to, and what’s allowed to be missing. Most CRM problems are that question being skipped. On the biggest job of this kind I’ve done, four Airtable databases holding 65,525 claims and 65,123 customers, 39% of the customer rows were duplicates, because nobody had decided what made two records the same person.
Move the data without losing any of it. That job ended with 305,713 records backed up and reconciled into a new four-table structure, 2,512 claims and 1,108 people migrated with zero rejections, and the duplicate rate cut from 39% to 18%. Along the way a 506-row loss turned up, which I traced to an outage and a rate limit and, after a three-way comparison of the old bases, reduced to a real gap of 58 rows. I’d rather find that than assume it wasn’t there.
Build the automations so failures show up. The old system on that job had webhooks that tried for about seven seconds and then gave up silently, with no queue and no record. That’s the normal way automations fail, and it’s why a business can lose a month of enquiries and not know. Everything I build now either succeeds, retries, or tells somebody, and I test what happens when the other end doesn’t answer.
Make the CRM match the business. HubSpot for Express Encapsulation, set up so the team could follow up leads and segment their trade customers for email, which is the list the business used in 2020 to sell PPE through its shop, £312,000 of it that year. Airtable and Make.com for Claims Bible, where the volume is higher and the rules about consent are stricter.
Do the boring infrastructure when it’s in the way. For Express that meant the router, shared files instead of files on one person’s machine, and a phone system moved to the cloud so it wasn’t tied to the building. None of it’s marketing and all of it’s the difference between a business that can grow and one that can’t.
The one I’m most pleased with
A claims system for a firm of solicitors, built in 2026. Most lead generation stops at a name and a phone number, and then someone has to ring, ask the real questions, get the paperwork signed and type it all into the firm’s system. This one carries a person from first enquiry through screening, consent, signature and documents, and delivers a structured case straight into the firm’s onboarding system through their API, with the source and a shared reference attached, and then hears back about what happened to it.
I designed it backwards from what the firm’s case handlers need, rather than forwards from a contact form. Between April and September 2026 it recorded 333 submissions, every one of them handed over, and 206 of them, which is 61.9%, reached an accepted status at the firm. An acceptance is a processing status rather than a win, and I don’t have a baseline to claim an uplift against, so I won’t. What I can say is that the firm gets a case, not a callback.
How I work
Airtable, Make.com, HubSpot, Zoho, Postgres and Python, depending on the size of the job and what the business can afford to run. The migration tooling is written so it can be run again, with a dry run first, because the first attempt is never the last. And I document the reconciliation, so that when a number’s questioned in a year’s time there’s a method to point at rather than a memory.
The proof
The two case studies are the Claims Bible data migration and Armed Forces Claims, with the small-business version inside Express Encapsulation.
“Gavin’s been working with us for several years now, covering the big three: website, marketing, and systems. He is hands on, quick to respond, organised, efficient and never needs a nudge. Before Gavin we had a very weak online presence; now we have a properly functioning, maintained and updated website, plus systems support. Gavin is very easy to work with, a great problem solver and he doesn’t overcomplicate anything.”
Deepak Wadhwani ACAa director at Forces Compare Limited, the company behind Claims Bible, which I co-founded and have worked with since
“Gavin at Fluxify has implemented our marketing strategy from the ground up, including: website development, CRM system, e-mail marketing, IT infrastructure and sales automation, with fantastic results.”
David BrantonProduction Manager at Express Encapsulation, in a Google review
Who this is for
Businesses where enquiries arrive and then something goes wrong with them, businesses with a CRM nobody uses because it doesn’t match how they work, and anyone with data in more than one place who’s stopped trusting any of it.
Tell me where the enquiries go after the form, and if you’re not sure, that’s usually the answer to what’s wrong.
Tell me what’s not working.
I’ll tell you whether I can help and roughly what it would involve.