I'm Lewis Fielding.

There's a particular kind of marketing advice that only exists because the person giving it never has to be there when it doesn't work.

My day job is head of digital and marketing for a wholesale food supplier. Trade customers, independent retailers, caterers, a sales team on the phones, real stock going out on real vans. Everything I do there gets judged on whether orders went up — not on whether the campaign looked nice in a deck.

That's an unusual place to be advising from, and it's the reason this site reads the way it does. When I tell you a percentage-of-spend ad fee is a bad incentive, or that your follow-up process is worth more than your ad budget, or that a beautiful site with no service pages will not rank — that isn't a position I read about. It's a position I've had to defend to a managing director looking at a P&L.

What I actually do all day

The unglamorous version: Google Ads and Meta Ads, search visibility, email campaigns, promotional calendars, brand and artwork, photography and video, sales enablement material, and the reporting that tells everyone whether any of it worked. Then the systems underneath — CRM structure, automation, and increasingly AI tooling to take the repetitive half of the job off my desk.

The part people find surprising is that I build as well as market. React, TypeScript, Supabase, APIs, AI integrations. Not because a marketing consultant needs to write code, but because being able to build the thing rather than brief someone else to build it collapses the time between "we should try this" and "we tried it". Most of the leverage in a small business comes from that gap being short.

How I work with clients

Deliberately few of them at once. I'm not building an agency, I'm not hiring a team, and I have no interest in a model where you meet the person who sells the work and then never see them again. You get me for the scoping, the building and the awkward conversations, or you get someone else — those are the options, and the second one is genuinely fine.

In practice that means I turn down work fairly regularly. Sometimes because the timing doesn't fit, sometimes because what someone needs is a £600 fix rather than a £6,000 project and it would be dishonest to pretend otherwise. If that's your situation I'd rather tell you on the first call and have you remember it in two years.

What I believe, briefly

• Enquiries beat impressions. If a report can't tell you how many people got in touch and what they were worth, it's decoration.
• Speed is a feature. Of the website, and of your reply to the person who filled the form in. Both are cheaper to fix than almost anything else and both move the number.
• You should own everything. Domain, hosting, code, ad accounts, data. Lock-in is a business model, not a service.
• Most businesses don't need more leads. They need to stop losing the ones they already generate.
• AI is a lever, not a strategy. It's genuinely transformative for the repetitive half of marketing work. It's very bad at knowing what your customers care about.
• If it can't be measured, say so. Some things are worth doing anyway. Pretending they're measurable is where trust goes.

Outside the work

I build things — AI tools, small SaaS products, automation systems, agent workflows. Several at once, usually more than is sensible. It's how I stay current, and it's the reason clients get approaches that most consultancies won't have tried yet, because I've already broken them on my own time first.

If any of this sounds like the way you'd want your marketing handled, the next step is a twenty-minute call. If it doesn't, the work page and the process page will tell you more than another paragraph here would.